University of Ghana http://ugspace.ug.edu.gh SCHOOL OF PUBLIC HEALTH COLLEGE OF HEALTH SCIENCES UNIVERSITY OF GHANA, LEGON ESTABLISHING REFERENCE INTERVALS FOR HAEMATOLOGICAL, BIOCHEMICAL, AND IMMUNOLOGICAL ANALYTES AMONG URBAN GHANAIAN ADULT POPULATION BY SERWAA AKOTO BAWUA (10552622) THIS THESIS IS SUBMITTED TO UNIVERSITY OF GHANA, LEGON, IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN PUBLIC HEALTH JULY, 2019 University of Ghana http://ugspace.ug.edu.gh DECLARATION I hereby declare that this thesis is the result of research undertaken by Serwaa Akoto Bawua toward the award of Doctor of Philosophy in Public Health and that, to the best of my knowledge it neither contains material previously published by another person nor material which has been accepted for the award of degree in any University, except where due acknowledgement has been made in the text and the reference section. This thesis write-up was done under the joint supervision of Prof. Julius Fobil, Prof. Patrick F. Ayeh- Kumi and Dr. John Arko-Mensah. …………………………….. ……………………… Serwaa Akoto Bawua Date (Student) …………………………….. ……………………... Dr. John Arko-Mensah Date (Co-Supervisor) …………………………….. ……………………... Prof. Patrick F. Ayeh-Kumi Date (Co-Supervisor) …………………………….. ……………………... Prof. Julius Fobil Date (Principal Supervisor) i University of Ghana http://ugspace.ug.edu.gh DEDICATION I dedicate this work to my Father in Heaven for taking me through my academic journey. His grace has been my source of strength. To my wonderful mother, Madam Margaret Oteng, my lovely sons Deyn Ennin and Yoni Ennin for enduring the many times I had to neglect them because of this research work. ii University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT I wish to express my profound gratitude to God almighty for the gift of life as well as his sufficient grace and mercy towards a successful academic journey. I also extend my appreciation to my supervisors Prof. Julius Fobil, Prof. Patrick F. Ayeh-Kumi and Dr. John Arko-Mensah for their guidance and encouragement throughout the research work. I wish to acknowledge Medlab Ghana Limited, Dr. Rosemary Keatley, Prof. Kiyoshi Ichihara and Prof. Erasmus for financially sponsoring this research. I am thankful to Prof. Ichihara for his support and contribution towards the statistical analysis section of the thesis. My sincere gratitude to all the marketing team and clinical laboratory professionals of Medlab Ghana Limited for their assistance in volunteers recruitment and the laboratory analysis of all the analytes. I am grateful to all the volunteers in the research, especially the Prison Officers, Retired Police Officers, GN bank staff as well as the Muslim community at Tamale central. My special thanks to my colleague and brother Richmond Owusu for his endless support towards this research work as well as his encouragement. I wish to express my sincere gratitude to my mother for her financial support and my husband and lovely children for their love and prayers. Finally, to my uncle Prof. Martin Oteng-Ababio, Dad Osei Akoto Bawua and siblings (Elfreda Bawua and Kwasi Bawua) you all cheered me up unto greater heights. I am forever grateful. iii University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ................................................................................................................. i DEDICATION ................................................................................................................... ii ACKNOWLEDGEMENT ................................................................................................ iii TABLE OF CONTENTS .................................................................................................. iv LIST OF TABLES ........................................................................................................... xvi LIST OF FIGURES ...................................................................................................... xviii LIST OF ABBREVIATIONS..........................................................................................xix ABSTRACT ................................................................................................................. xxiii CHAPTER ONE ................................................................................................................. 1 INTRODUCTION .............................................................................................................. 1 1.0 Introduction ............................................................................................................... 1 1.1 Background of the Study .......................................................................................... 1 1.2 Problem Statement .................................................................................................... 4 1.3 Justification of the Study .......................................................................................... 8 1.4 General Objective ................................................................................................... 10 1.5 Specific Objectives ................................................................................................. 10 1.6 Research Questions ................................................................................................. 10 1.7 Conceptual Framework ........................................................................................... 11 CHAPTER TWO .............................................................................................................. 16 LITERATURE REVIEW ................................................................................................. 16 2.0 Introduction ............................................................................................................. 16 2.1 Concept of Reference Intervals .............................................................................. 16 2.2 Concept of Multicenter Reference Interval Studies ................................................ 18 2.3 Factors Affecting RI Derivation ............................................................................. 21 2.3.1 Recruitment of the Reference Population. ....................................................... 22 2.3.2 Pre-Analytical Variables .................................................................................. 23 2.3.2.1 Controllable Factors Before the Sample Collections ................................ 24 2.3.2.2 Uncontrollable Factors before the Sample Collections ............................ 26 2.3.3 Analytical Variation. ........................................................................................ 27 2.4 Statistical Methods in Calculating Reference Interval ........................................... 28 iv University of Ghana http://ugspace.ug.edu.gh 2.4.1 Non-parametric Method. .................................................................................. 28 2.4.2 Parametric Method. .......................................................................................... 28 2.5 Common Analytes in Human Medicine ................................................................. 30 2.5.1 Haematological Analytes (Complete Blood Count). ....................................... 30 2.5.1.1 Neutrophil. ................................................................................................ 31 2.5.1.2 Lymphocytes. ............................................................................................ 31 2.5.1.3 Monocytes. ................................................................................................ 32 2.5.1.4 Basophils................................................................................................... 32 2.5.1.5 Eosinophils. .............................................................................................. 32 2.5.1.6 Platelets. .................................................................................................... 32 2.5.1.7 Haemoglobin (Hb). ................................................................................... 33 2.5.1.8 Other Red Blood Cell Indices (RDW, Red cell count, MCH, MCHC). ... 34 2.5.2 Biochemical Analytes. ..................................................................................... 34 2.5.2.1 Calcium (Ca). ............................................................................................ 34 2.5.2.2 Creatine Kinase (CK). .............................................................................. 35 2.5.2.3 Lactate Dehydrogenase (LDH). ................................................................ 35 2.5.2.4 Amylase. ................................................................................................... 36 2.5.2.5 Uric Acid (UA). ........................................................................................ 37 2.5.2.6 C-Reactive Protein (CRP)......................................................................... 37 2.5.2.7 Liver Function Tests. ................................................................................ 37 2.5.2.8 Total Protein. ............................................................................................ 38 2.5.2.9 Albumin. ................................................................................................... 38 2.5.2.10 Bilirubin. ................................................................................................. 39 2.5.2.11 Aspartate aminotransferase (AST). ........................................................ 39 2.5.2.12 Alanine aminotransferase (ALT). ........................................................... 40 2.5.2.13 Alkaline phosphatase (ALP). .................................................................. 40 2.5.2.14 Gamma-glutamyl Transferase (GGT). .................................................... 41 2.5.2.15 Urea. ........................................................................................................ 41 2.5.2.16 Creatinine. ............................................................................................... 42 2.5.2.17 Glomerular Filtration Rate (GFR). ......................................................... 42 2.5.2.18 Total Carbon dioxide (TCO2). ............................................................... 42 2.5.2.19 Anion Gap. .............................................................................................. 43 2.5.2.20 Diabetes test. ........................................................................................... 43 2.5.2.21 Lipid Studies. .......................................................................................... 44 v University of Ghana http://ugspace.ug.edu.gh 2.5.2.22 Immunoglobulin A, M, G (IgA, IgM, IgG). ........................................... 44 2.5.2.23 Iron Studies. ............................................................................................ 45 2.5.2.24 Vitamins Studies (B12, D 25-OH, folate). ............................................. 45 2.5.2.25 Endocrine Studies. .................................................................................. 46 2.5.2.26 Thyroid Studies. ...................................................................................... 47 2.5.2.27 Thyroid-Stimulating Hormone (TSH). ................................................... 48 2.6 Empirical Studies .................................................................................................... 49 2.7 Conclusion .............................................................................................................. 60 CHAPTER THREE .......................................................................................................... 61 METHODS AND MATERIALS ..................................................................................... 61 3.0 Introduction ............................................................................................................. 61 3.1 Study Design ........................................................................................................... 61 3.2 Study Area/ Study Sites .......................................................................................... 62 3.3 Study Population ..................................................................................................... 63 3.4 Sample Size ............................................................................................................ 63 3.5 Sampling Procedure ................................................................................................ 64 3.5.1 Inclusion criteria. ............................................................................................. 64 3.5.2 Exclusion criteria. ............................................................................................ 64 3.5.3 Selection of Study Participants. ....................................................................... 65 3.6 Data Collection Techniques and Tools ................................................................... 66 3.6.1 Source of Data ................................................................................................. 66 3.6.2 Sociodemographic and lifestyle characteristics ............................................... 67 3.7 Variables ................................................................................................................. 67 3.7.1 Independent Variables. .................................................................................... 67 3.7.2 Dependent Variables. ....................................................................................... 67 3.8 Blood pressure and anthropometric measurement .................................................. 68 3.8.1 Blood Pressure Measurement. ......................................................................... 68 3.8.2 Weight Measurement. ...................................................................................... 68 3.8.3 Height Measurement ........................................................................................ 68 3.8.4 Body mass index (BMI) Measurement. ........................................................... 69 3.8.5 Waist Circumference. ...................................................................................... 69 3.8.6 Physical Activity (Exercise level). .................................................................. 69 3.8.7 Alcohol. ........................................................................................................... 70 vi University of Ghana http://ugspace.ug.edu.gh 3.9 Laboratory analysis ................................................................................................. 70 3.9.1 Sample Collection, Storage, and Measurements. ............................................ 70 3.9.2 Pre-laboratory Analytical Process. .................................................................. 71 3.9.3 Sample Transporting Procedure. ..................................................................... 72 3.9.4 Target Analytes and Measurements. ................................................................ 72 3.9.5 Haematological Analytical Procedure. ............................................................ 73 3.9.5.1 Reagents and Equipment. ......................................................................... 73 3.9.5.2 Determination of Haematological analytes. ............................................. 73 3.9.6 Biochemical Analytical Procedure. ................................................................. 75 3.9.7 Gamma-Glutamyl Transferase Analytes (GGT). ........................................... 76 3.9.7.1 Required Materials, Reagents and Equipment. ......................................... 76 3.9.7.2 Determination of serum GGT concentration by enzymatic rate method. . 76 3.9.7.3 Reaction. ................................................................................................... 76 3.9.8 Alkaline phosphatase (ALP). ........................................................................... 77 3.9.8.1 Required Materials, Reagents and Equipment. ......................................... 77 3.9.8.2 Determination of serum ALP concentration by kinetic rate method. ....... 77 3.9.8.3 Reactions. .................................................................................................. 77 3.9.9 Aspartate Aminotransferase (AST). ................................................................ 78 3.9.9.1 Required Materials, Reagents and Equipment. ......................................... 78 3.9.9.2 Determination of serum AST concentration by IFCC method. ................ 78 3.9.9.3 Reaction. ................................................................................................... 78 3.9.10 Alanine Aminotransferase (ALT). ................................................................. 79 3.9.10.1 Required Materials, Reagents and Equipment ........................................ 79 3.9.10.2 Determination of serum ALT concentration by IFCC method. .............. 79 3.9.10.3 Reaction .................................................................................................. 79 3.9.11 Lactate Dehydrogenase (LDH). ..................................................................... 79 3.9.11.1 Required Materials, Reagents and Equipment. ....................................... 79 3.9.11.2 Determination of serum LDH concentration by enzymatic rate method. .............................................................................................................................. 80 3.9.12 Amylase (AMY) ............................................................................................ 80 3.9.12.1 Required Materials, Reagents and Equipment ........................................ 80 3.9.12.2 Determination of serum Amylase concentration by Beckman Olympus- blocked CNPG3 method. ...................................................................................... 80 3.9.12.3 Chemical Reaction. ................................................................................. 81 vii University of Ghana http://ugspace.ug.edu.gh 3.9.13 Creatine Kinase (CK). ................................................................................... 81 3.9.10.13.1 Required Materials, Reagents and Equipment.................................. 81 3.9.13.2 Determination of serum CK concentration by Immune-Inhibition (IFCC) method. ................................................................................................................. 81 3.9.13.3 Chemical Reaction. ................................................................................. 82 3.9.14 Creatinine (CRE). .......................................................................................... 82 3.9.13.1 Required Materials, Reagents and Equipment ........................................ 82 3.9.14.2 Determination of serum Creatinine concentration by kinetic method. ... 82 3.9.14.3 Chemical Reaction. ................................................................................. 82 3.9.15 Principle Procedure for Total Protein (TP). ................................................... 82 3.9.15.1 Required Materials, Reagents and Equipment. ....................................... 82 3.9.15.2 Determination of serum Total Protein concentration by timed rate biuret method. ................................................................................................................. 83 3.9.16. Principle Procedures of Albumin (ALB). ..................................................... 83 3.9.16.1 Required Materials, Reagents and Equipment. ....................................... 83 3.9.16.2 Determination of serum Albumin concentration by Bromocresol Green (BCG) method. ..................................................................................................... 83 3.9.16.3 Chemical Reaction. ................................................................................. 84 3.9.17 Principle Procedures of Urea. ........................................................................ 84 3.9.17.1 Required Materials, Reagents and Equipment. ....................................... 84 3.9.17.2 Determination of serum Urea concentration by kinetic method. ............ 84 3.9.17.3 Chemical Reaction. ................................................................................. 84 3.9.18 Principle Procedures of Total Bilirubin (TBIL). ........................................... 84 3.9.18.1 Required Materials, Reagents and Equipment. ....................................... 84 3.9.18.2 Determination of serum Total Bilirubin concentration by Dichlorophenyl Diazonium method. ............................................................................................... 85 3.9.18.3 Chemical Reaction. ................................................................................. 85 3.9.19 Principle Procedures of Direct Bilirubin (DBIL). ......................................... 85 3.10.19.1 Required Materials, Reagents and Equipment. ..................................... 85 3.9.19.2 Determination of serum Direct Bilirubin concentration by Dichlorophenyl Diazonium method. .................................................................... 85 3.9.19.3 Chemical Reaction. ................................................................................. 86 3.9.20 Principle Procedures of Glucose (GLU). ....................................................... 86 3.9.20.1 Required Materials, Reagents and Equipment. ....................................... 86 viii University of Ghana http://ugspace.ug.edu.gh 3.9.20.2 Determination of serum Glucose concentration by hexokinase method.86 3.9.20.3 Chemical Reaction. ................................................................................. 87 3.9.21 Principle Procedures of Uric acid (UA). ........................................................ 87 3.9.21.1 Required Materials, Reagents and Equipment. ....................................... 87 3.9.21.2 Determination of serum Uric acid concentration. ................................... 87 3.9.21.3 Chemical Reaction .................................................................................. 87 3.9.22 Principle Procedure of Total Cholesterol . .................................................... 88 3.9.22.1 Required Materials, Reagents and Equipment. ....................................... 88 3.9.22.2 Determination of serum Total Cholesterol concentration enzymatic timed-point method. .............................................................................................. 88 3.10.22.3 Chemical Reaction. ............................................................................... 88 3.9.23 Principle Procedures of High-Density Lipoprotein Cholesterol (HDL- C). .. 89 3.9.23.1 Required Materials, Reagents and Equipment. ....................................... 89 3.9.23.1 Determination of serum HDL-Cholesterol concentration by Enzymatic Immunoinhibition method. ................................................................................... 89 3.10.23.3 Chemical Reaction. ............................................................................... 89 3.9.24 Principle Procedures of Low-Density Lipoprotein Cholesterol (LDL-C). .... 90 3.10.24.1 Required Materials, Reagents and Equipment. ..................................... 90 3.9.24.2 Determination of serum LDL--Cholesterol concentration by Enzymatic selective method. .................................................................................................. 90 3.10.24.3 Chemical Reaction. ............................................................................... 91 3.9.25 Principle Procedure of Triglycerides (TG). ................................................... 91 3.9.25.1 Required Materials, Reagents and Equipment. ....................................... 91 3.9.25.2 Determination of serum Triglycerides concentration by glycerol phosphate oxidase – Peroxidase method. ............................................................. 91 3.10.24.3 Chemical Reaction. ............................................................................... 92 3.9.26 Principle Procedure of Electrolytes (Na+ K+, Cl-). ........................................ 92 3.9.26.1 Required Materials, Reagents and Equipment. ....................................... 92 3.9.27 Principle Procedure of Sodium (Na+). ........................................................... 93 3.9.27.1 Required Materials, Reagents and Equipment. ....................................... 93 3.9.27.2 Determination of serum Sodium (Na+) concentration by ion selective electrode method . ................................................................................................. 93 3.9.28 Principle Procedure of Potassium (K+). ......................................................... 93 3.9.28.1 Required Materials, Reagents and Equipment. ....................................... 93 ix University of Ghana http://ugspace.ug.edu.gh 3.9.28.2 Determination of serum Potassium concentration by ion selective electrode method. .................................................................................................. 94 3.9.29 Principle Procedure of Chloride (Cl-). ........................................................... 94 3.9.29.1 Required Materials, Reagents and Equipment. ....................................... 94 3.9.29.2 Determination of serum Chloride concentration by ion selective electrode method. .................................................................................................. 94 3.9.30 Principle Procedure of Phosphorous. ............................................................. 95 3.10.29.1 Required Materials, Reagents and Equipment. ..................................... 95 3.9.30.2 Determination of serum Phosphorous concentration. ............................. 95 3.10.30.3 Chemical Reaction. ............................................................................... 95 3.9.31 Principle Procedure of Calcium. .................................................................... 95 3.9.31.1 Required Materials, Reagents and Equipment. ....................................... 95 3.9.31.2 Determination of serum Calcium concentration by Arsenazo method. .. 96 3.10.31.3 Chemical Reaction. ............................................................................... 96 3.9.32 Principle Procedure of Magnesium. .............................................................. 96 3.9.32.1 Required Materials, Reagents and Equipment. ....................................... 96 3.9.32.2 Determination of serum Magnesium concentration by Xylidyl Blue and Tris-Buffer method. .............................................................................................. 96 3.10.32.3 Chemical Reaction. ............................................................................... 97 3.9.33 Principle Procedure of C – Reactive Protein (CRP). ..................................... 97 3.9.33.1 Required Materials, Reagents and Equipment. ....................................... 97 3.9.33.2 Determination of serum C – Reactive Protein (CRP) concentration by Immunoturbidimetric method. .............................................................................. 97 3.10.33.3 Chemical Reaction. ............................................................................... 98 3.9.34 Principle Procedure of Carbon Dioxide. ........................................................ 98 3.9.34.1 Required Materials, Reagents and Equipment. ....................................... 98 3.9.34.2 Determination of serum Carbon Dioxide concentration by Enzymatic method. ................................................................................................................. 98 3.10.33.3 Chemical Reaction. ............................................................................... 98 3.9.35 Hormones Analytical Procedure. ................................................................... 99 3.9.36 Principle Procedure of Estradiol. ................................................................... 99 3.9.36.1 Required Materials, Reagents and Equipment. ....................................... 99 3.9.36.2 Determination of serum Estradiol concentration by competitive assay method. ................................................................................................................. 99 x University of Ghana http://ugspace.ug.edu.gh 3.9.37 Principle Procedure of Progesterone. .......................................................... 100 3.9.37.1 Required Materials, Reagents and Equipment. ..................................... 100 3.9.37..2 Determination of serum Progesterone concentration by competitive assay method. ...................................................................................................... 100 3.9.38 Principle Procedure of Luteinizing Hormone (LH). .................................... 101 3.9.38.1 Required Materials, Reagents and Equipment. ..................................... 101 3.9.38.2 Determination of serum Luteinizing Hormone concentration by sandwich assay method. ...................................................................................... 101 3.9.39 Principle Procedure of follicle‐stimulating hormone (FSH). ...................... 102 3.9.39.1 Required Materials, Reagents and Equipment. ..................................... 102 3.9.39.2 Determination of serum Follicle‐stimulating hormone concentration by sandwich assay method. ...................................................................................... 102 3.9.40 Principle Procedure of Prolactin. ................................................................. 103 3.9.40.1 Required Materials, Reagents and Equipment. ..................................... 103 3.9.40.2 Determination of serum Prolactin hormone concentration by sandwich assay method. ...................................................................................................... 103 3.9.41 Principle Procedure of Testosterone. ........................................................... 104 3.9.41.1 Required Materials, Reagents and Equipment. ..................................... 104 3.9.41.2 Determination of serum Testosterone concentration by competition assay method. ...................................................................................................... 104 3.9.42 Principle Procedure of Thyroid‐stimulating hormone (TSH). ..................... 105 3.9.42.1 Required Materials, Reagents and Equipment. ..................................... 105 3.9.42.2 Determination of serum TSH concentration by sandwich assay method. ............................................................................................................................ 105 3.9.43 Principle Procedure of Ferritin. ................................................................... 106 3.9.43.1 Required Materials, Reagents and Equipment. ..................................... 106 3.9.43.2 Determination of serum Ferritin concentration by sandwich assay method. ............................................................................................................... 106 3.9.44 Principle Procedure of Prostate‐Specific Antigen (PSA). ........................... 107 3.9.44.1 Required Materials, Reagents and Equipment. ..................................... 107 3.9.44.2 Determination of serum PSA concentration by sandwich assay method. ............................................................................................................................ 107 3.10 Quality Control ................................................................................................... 108 xi University of Ghana http://ugspace.ug.edu.gh 3.10.1 Laboratory Quality Control. ........................................................................ 109 3.10.2 Mini-panel. ................................................................................................... 109 3.11 Data Management ............................................................................................... 109 3.12 Statistical Analysis .............................................................................................. 110 3.12.1 Descriptive Analysis. ................................................................................... 110 3.12.2 Data Regression. .......................................................................................... 111 3.12.3 Partitioning Criteria. .................................................................................... 112 3.12.4 Latent abnormal values exclusion (LAVE) method. ................................... 112 3.12.5 Reference Interval Derivation. ..................................................................... 114 3.13 Ethical Approval ................................................................................................. 114 CHAPTER FOUR .......................................................................................................... 115 RESULTS ....................................................................................................................... 115 4.0 Introduction ........................................................................................................... 115 4.1 Participant Demographics ..................................................................................... 115 4.2 Haematological Analytes ...................................................................................... 117 4.3 Biochemical Analytes ........................................................................................... 119 4.3.1 Kidney Function Tests. .................................................................................. 121 4.3.2 Liver Function Tests. ..................................................................................... 122 4.3.3 Diabetes and Lipids Analytes ........................................................................ 123 4.3.4 Immunoglobulins. .......................................................................................... 124 4.3.5 Iron and Vitamins Analytes. .......................................................................... 125 4.3.6 Endocrine Parameters. ................................................................................... 126 4.3.7 Thyroid Analytes. .......................................................................................... 127 4.3.8 Tumour Markers. ........................................................................................... 128 4.4 Relationship between Age, Ethnicity, BMI, SBP, DBP, Alcohol, Hours of Standing and Haematological analytes. ...................................................................... 129 4.4.1 Age, BMI and Haematology Analytes. .......................................................... 129 4.4.2 Relationship between Ethnicity and Haematology Analytes. ........................ 129 4.5 Demographic and Physical Characteristics and Clinical Chemistry .................... 132 4.5.1 Relationship between Age and Clinical Chemistry ....................................... 132 4.5.2 Ethnicity and Clinical Chemistry ................................................................... 132 4.5.3 Relationship between BMI and Clinical Chemistry ...................................... 132 4.5.4 Blood Pressure (SBP and DBP) and Clinical Chemistry............................... 133 xii University of Ghana http://ugspace.ug.edu.gh 4.5.5 Exercise level, Standing Hours and Clinical Chemistry ................................ 133 4.5.6 Age, BMI and Kidney Function Tests ........................................................... 136 4.5.7 Relationship between Blood Pressure (SBP and DBP) and Kidney Function ................................................................................................................................ 136 4.5.8 Age and Liver Function Parameters .............................................................. 138 4.5.9 BMI and Liver Function ................................................................................ 138 4.5.10 Age and Lipids Analytes. ............................................................................ 140 4.5.11 BMI and Lipids Analytes. ............................................................................ 140 4.5.12 Age and Immunoglobulins. ......................................................................... 140 4.5.13 Age,Thyroid and Endocrine Parameters. ..................................................... 143 4.5.14 BMI, Thyroid and Endocrine Parameters .................................................... 143 4.5.15 Age, BMI and Iron and Vitamin Metabolism .............................................. 146 4.5.16 Age and Tumour Markers. ........................................................................... 146 4.6 Age, Sex and Haematology .................................................................................. 149 4.6.1 Age, Sex and Haematology Analytes. ........................................................... 149 4.7 Clinical Chemistry, Kidney and Liver function ................................................... 150 4.7.1 Age, Sex and Liver Function ......................................................................... 150 4.7.2 Age, Sex and Kidney Function ...................................................................... 151 4.7.3 Age, Sex, Lipids, Immunoglobulins, Iron and Vitamins ............................... 152 4.7.4 Age, Sex, Tumour Markers and Endocrinology ............................................ 153 4.8 Graphical Presentation of SDRage and SDRsex in Selected Analytes. ............... 154 4.8.1 Clinical Chemistry. ........................................................................................ 155 4.8.1.2 Kidney Function ..................................................................................... 155 4.8.1.3 Liver Function ........................................................................................ 155 4.8.1.4 Lipids Profile. ......................................................................................... 156 4.8.1.5 Immunoglobulins. ................................................................................... 156 4.9 Derivation and Comparison between LAVE (+) and LAVE (-) methods ............ 163 CHAPTER FIVE ............................................................................................................ 171 DISCUSSION ................................................................................................................. 171 5.0 Introduction ........................................................................................................... 171 5.1 Reference Intervals for Haematological Analytes ............................................... 171 5.2 Reference Intervals for Clinical Chemistry Analytes ........................................... 176 5.2.1 Reference Intervals for Kidney Function Analytes. ...................................... 178 xiii University of Ghana http://ugspace.ug.edu.gh 5.2.2 Reference Intervals for Liver Function Analytes. ......................................... 180 5.2.3 Reference Intervals for Diabetes Screen Analytes. ....................................... 182 5.2.4 Reference Intervals for Lipids Profile. .......................................................... 183 5.2.5 Reference Intervals for Immunoglobulins. .................................................... 185 5.2.6 Reference Intervals for Iron Studies. ............................................................. 187 5.2.7 Reference Intervals for Vitamins Analytes. ................................................... 188 5.2.8 Reference Intervals for Endocrine Analytes. ................................................ 190 5.2.9 Reference Intervals for Tumour Markers. .................................................... 192 5.2.9.1 Alpha 1-Fetoprotein (AFP). .................................................................... 192 5.2.9.2 Carcinoembryonic Antigen (CEA). ........................................................ 193 5.2.9.3 Carbohydrate Antigen 125 (CA125). ..................................................... 194 5.2.9.4 Prostate-specific antigen (PSA) .............................................................. 194 5.3 Determination of Sources of Variation ................................................................. 195 5.3.1 Age and Clinical chemistry Analytes. ........................................................... 195 5.3.1.1 Relationship between Age and Kidney Function Analytes. ................... 195 5.3.1.2 Relationship between Age and Liver Function Analytes. ...................... 197 5.3.1.3 Relationship between Age and Lipids Analytes. .................................... 198 5.3.1.4 Relationship between Age and Endocrine Analytes. .............................. 198 5.3.1.5 Relationship between Age and Tumour markers. ................................... 200 5.3.2 Body Mass Index and Clinical Chemistry Analytes. ..................................... 202 5.3.2.1 Body Mass Index and Liver Function .................................................... 203 5.3.2.2 Body Mass Index and Lipid Analytes. ................................................... 203 5.3.2.3 Body Mass Index and Endocrine Analytes (Hormones). ....................... 204 5.3.3 Systolic Blood Pressure and Kidney Function .............................................. 205 5.3.4 Hours of Standing and Lipids Profile. ........................................................... 205 5.4 Standard Deviation Ratios (Sex, Age, Ethnicity). ................................................ 206 5.4.1 Standard Deviation Ratios and Haematology. ............................................... 206 5.4.2 Standard Deviation Ratios and Chemistry Analytes. .................................... 207 5.4.2.1 Standard Deviation Ratios and Kidney Function Analytes. ................... 207 5.4.2.2 Standard Deviation Ratios andLiver Function Analytes. ....................... 209 5.4.2.3 Standard Deviation Ratios and Lipids Profile. ....................................... 210 5.4.2.4 Standard Deviation Ratios and Iron Studies. .......................................... 210 5.4.2.5 Standard Deviation Ratios and Tumour Markers. .................................. 211 5.4.2.6 Standard Deviation Ratios and Endocrinoly.......................................... 211 xiv University of Ghana http://ugspace.ug.edu.gh 5.4.2.7 Standard Deviation Ratios (Sex, Age, Ethnicity) on Immunoglobulins. 213 5.4.2.8 Standard Deviation Ratios (Sex, Age, Ethnicity) on Vitamins Studies. . 213 5.5 Latent Abnormal Values Exclusion Method on both parametric and non- parametric approaches ................................................................................................ 213 CHAPTER FIVE ............................................................................................................ 217 CONCLUSIONS AND RECOMMENDATIONS ......................................................... 217 6.0 Introduction ........................................................................................................... 217 6.1 Conclusions........................................................................................................... 217 6.2 Recommendations ................................................................................................. 220 6.2.1 Ministry of Health, Ghana Health Service, Ghana Standards Authority (Medium-term) ....................................................................................................... 220 6.2.2 Researchers and Clinical Research Laboratories (Short-term) ...................... 221 6.2.3 Healthcare Providers and Medical Practitioners (Short-term) ....................... 221 6.2.4 General Population/Ministry of Health (Long-term) ..................................... 222 6.3 Limitation of Study ............................................................................................... 222 6.4 Future Research Directions ................................................................................... 223 6.5 Contribution to Knowledge .................................................................................. 223 REFERENCES ............................................................................................................... 225 APPENDICES ................................................................................................................ 272 APPENDIX I: QUESTIONNAIRE ............................................................................ 272 APPENDIX II: CONSENT FORM ............................................................................ 276 APPENDIX III: COMPARISON OF GHANAIAN RIs VS MANUFACTURER’S INSERT ...................................................................................................................... 280 xv University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 2.1:Empirical Studies of Reference Interval .......................................................... 49 Table 2.2:Multicenter Studies on Reference Intervals .................................................... 57 Table 4.1:Characteristics of age mean, gender, BMI, SBP and DBP of study participants ................................................................................................... 116 Table 4.2:Distribution of age category with gender, ethnicity and exercise of study participants ................................................................................................... 116 Table 4.3;Reference intervals for haematological analytes derived parametrically ...... 118 Table 4.4:Reference intervals derived parametrically for clinical chemistry analytes .. 120 Table 4.5:Reference intervals derived parametrically for Kidney function Tests ......... 121 Table 4.6:Reference intervals derived parametrically for Liver function Tests ............ 123 Table 4.7:Reference intervals for Diabetes and Lipids Analytes .................................. 124 Table 4.8:Reference intervals for Immunoglobulins ..................................................... 125 Table 4.9:Reference intervals for Iron and Vitamins .................................................... 126 Table 4.10:Reference intervals for Endocrine (Hormones) ........................................... 127 Table 4.11: Reference intervals for Thyroid Analytes .................................................. 128 Table 4.12: Reference intervals for Tumour markers Test derived parametrically. ...... 128 Table 4.13: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Haematology Analytes ............................................................................ 130 Table 4.14: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Clinical Chemistry Analytes ................................................................... 134 Table 4.15: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Kidney Function Test Analytes .............................................................. 137 Table 4.16: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Liver Function Test Analytes ................................................................. 139 xvi University of Ghana http://ugspace.ug.edu.gh Table 4.17: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Lipids Profile Analytes ........................................................................... 141 Table 4.18: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, Exercise Level and Hours of Standing on Immunoglobulin Analytes ........................................ 142 Table 4.19: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Thyroid and Endocrine Analytes ................................................................. 144 Table 4.20: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Tumour Markers .......................................................................................... 147 Table 4.21: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Tumour Markers .......................................................................................... 148 Table 4.22: Standard Deviation Ratios for Sex, Age and Ethnicity in Haematological Analytes ....................................................................................................... 149 Table 4.23: Standard Deviation Ratios for Sex, Age and Ethnicity in Clinical Chemistry Analytes ....................................................................................................... 150 Table 4.24: Standard Deviation Ratios for Sex, Age and Ethnicity in Liver function analytes Analytes ......................................................................................... 150 Table 4.25: Standard Deviation Ratios for Sex, Age and Ethnicity in Kidney Function Analytes ....................................................................................................... 151 xvii University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 1.1:Conceptual Framework of some factors influencing the derivation of reference intervals of analytes ....................................................................... 11 Figure 3.1:Flow diagram for recruitment of Participants ................................................ 66 Figure 4.1: Graphical representation of SDRage and SDRsex in Selected Clinical Chemistry Analytes ..................................................................................... 157 Figure 4.2: Graphical representation of SDRage and SDRsex in Selected Kidney Function Tests .............................................................................................. 158 Figure 4.3: Graphical representation of SDRage and SDRsex in Selected Kidney Function Tests .............................................................................................. 159 Figure 4.4: Graphical representation of SDRage and SDRsex in Selected Liver Function Tests ............................................................................................................. 160 Figure 4.5: Graphical representation of SDRage and SDRsex in Selected lipids and Diabetes Tests .............................................................................................. 161 Figure 4.6: Graphical representation of SDRage and SDRsex in Selected Immunoglobulins. ........................................................................................ 162 Figure 4.7: Comparison between LAVE (+) and LAVE (-) Method in Hb and Ht ....... 163 Figure 4.8: Comparison between LAVE (+) and LAVE (-) Method in MCV, MCH, RBC and Folate ..................................................................................................... 164 Figure 4.9: Comparison between LAVE (+) and LAVE (-) Method in TF, RDW, Fe and Ferr. .............................................................................................................. 165 Figure 4.10: Comparison between LAVE (+) and LAVE (-) Method in Neutrophil# and CRP. ............................................................................................................. 166 Figure 4.11: Comparison between LAVE (+) and LAVE (-) Method in HDL-C and LDL-C .......................................................................................................... 167 Figure 4.12: Comparison between LAVE (+) and LAVE (-) Method in TG and TC ... 167 Figure 4.13: Comparison between LAVE (+) and LAVE (-) Method in ALP, GGT, AST, and ALT ....................................................................................................... 168 Figure 4.14: Comparison between LAVE (+) and LAVE (-) Method in Cre, UA, Dbil, and Tbil. ....................................................................................................... 169 xviii University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS Alb - Albumin ALT - Alanine Aminotransferase ALP - Alkaline Phosphatase AMY - Amylase ANOVA - Analysis of variance AST - Aspartate Aminotransferase BMI - Body Mass Index BASO# - Basophils Ca - Calcium CA125 - Carcinoembryonic Antigen Cre - Creatinine CK - Creatine Kinase CLSI - Clinical Laboratory Standard Institute C-RIDL - Committee on Reference Intervals and Decision Limits CRP - C- Reactive Protein C3- C3 - Complement C4- C4 - Complement EOS# - Eosinophils Counts FDA - Food and Drugs Authority Fe - Iron FSH - Follicle‐Stimulating Hormone FT3 - Free Triiodothyronine FT4 - Free Thyroxine (Ft4) GGT - Gamma-Glutamyl transferase xix University of Ghana http://ugspace.ug.edu.gh GLU - Glucose GSA - Ghana Standards Authority GSS - Ghana Statistical Service Hb - Haemoglobin Ht - Hematocrit HDL-C - High-Density Lipoprotein Cholesterol HBV - Hepatitis B Virus HCV - Hepatitis C Virus HIV - Human Immunodeficiency Virus ICSH - International Committee for Standardization in Haematology IFCC - International Federation for Clinical Chemistry IgG – Immunoglobulin G IgA – Immunoglobulin A IgM – Immunoglobulin M LYMPH# - Lymphocytes Counts K – Potassium LDH - Lactate dehydrogenase LDL-C - Low-Density Lipoprotein Cholesterol LAVE – Latent Abnormal Value Exclusion MALRA - Major-Axis Linear Regression Analysis MCV - Mean Corpuscular Haemoglobin Volume MCHC - Mean Corpuscular Haemoglobin Concentration MPV - Mean Platelet Volume Mg - Magnesium xx University of Ghana http://ugspace.ug.edu.gh MONO# - Monocytes Counts MONO% - Monocytes % NEUT# - Neutrophils Counts PLT - Platelet PSA - Prostate‐Specific Antigen PTH - Parathyroid Hormone RI - Reference Interval RMP - Reference Measurement Procedures RDW - Red Cell Distribution Width RBC - Red Blood Cells SDR - Standard Deviation Ratio SV - Source of Variation SOPs - Standard Operation Procedures TBil - Total Bilirubin TC - Total Cholesterol TG - Triglycerides TF - Transferrin TSH - Thyroid‐Stimulating Hormone WBC - White Blood Cells WHO - World Health Organization xxi University of Ghana http://ugspace.ug.edu.gh OPERATIONAL DEFINITION OF TERMS Reference Interval - is the set of range obtained by either observation or quantitative measurement of an analyte in a selected group of healthy individuals based on well-defined criteria. Reference Value - is the value, or test result, obtained by the observation or measurement of a particular type of quantity on a reference individual. Standard Deviation Ratio - is the ratio of between-subgroup standard deviation (variation of the subgroup means from grand mean) to between -individual standard deviation. Latent Abnormal Value Exclusion - is an iterative optimization secondary exclusion procedure use to eliminate individuals with latent disease in order to refine the data. xxii University of Ghana http://ugspace.ug.edu.gh ABSTRACT Background Reference intervals (RIs) refer to the upper and lower reference limits of laboratory test derived from healthy individuals recruited with well-defined criteria. RI serves as a comparison tool and an important determinant of whether an individual is healthy or not, which apparently remains the most extensively used decision-making tool in clinical settings. Considering the importance of RIs for the interpretation of laboratory test, the International Federation of Clinical Chemistry (IFCC) has recommended that each laboratory obtains its own reference values and estimate the corresponding RIs within defined procedures. Even though the recommendation from IFCC/Clinical laboratory standard institute (CLSI) is required, yet majority of diagnostic laboratories in Ghana are unable to implement their own RIs due to the cost and challenges involved in recruiting the reference population. For that reason, the majority of laboratories work with reference intervals that are based on guidelines developed by manufacturers of analyzers. Thus, the continual use of such existing RIs from different manufacturer analyzers affects clinical decision making and might lead wrong interpretation of laboratory results, which is a public health concern. Aim This study aimed to establish reference intervals for haematological, biochemical and immunological analytes that would inform context-specific clinical decision-making and interpretation of laboratory results for healthcare practice in Ghana. Methods In a cross-sectional study design; healthy individuals, aged 18 – 60+ years were recruited using a simple random sampling technique. A total of 501 apparently healthy subjects; from the Tamale and Accra Metropolis which are capital cities of the Northern and Greater xxiii University of Ghana http://ugspace.ug.edu.gh Accra region respectively were recruited into the study. Structured questionnaires were administered to volunteers to collect data on demographics, lifestyles, dietary pattern and their clinical information. The selection of eligible participants was primarily based on well-defined inclusion and exclusion criteria, which were in accordance with the IFCC/C- RIDL protocol. A fasting blood sample of 24mL was drawn into two plastic evacuated tubes containing ethylene diamine tetra-acetic acid (EDTA), three serum separate tube (SST), one sodium fluoride (NaFl) tube and one lithium heparin tube for complete blood count (CBC) and clinical chemistry analysis. Whole blood samples were analyzed using Sysmex XN 1000 analyzer for haematology analytes while the clinical chemistry, immunoglobulin, hormones, and tumour marker analytes were analyzed using Beckman Coulter AU 480 analyzer, Roche Cobas e411 analyser and Centaur XP Siemens’ analyzer respectively. All laboratory investigations were carried out in accordance with the laboratory’s standard operating procedures (SOPs). Stata version 13 software (Stata Corp., College Station, Texas, United States) was used in analyzing the data. Multiple regression analysis was performed to determine the relationship between variables. Partitioning of reference values by sex and age was done by the computing standard deviation ratio (SDR) using 3-level nested ANOVA by StatFlex version 6.0 statistical software (Artech Inc., Osaka, Japan). Prior to the derivation of the RIs, the latent abnormal values exclusion (LAVE) method was conducted to exclude individuals with such latent diseases. The latent abnormal values exclusion (LAVE) method is a secondary exclusion procedure applied to refine the data. Using “Reference interval Master” software, RIs of each analyte were derived using the parametric method. xxiv University of Ghana http://ugspace.ug.edu.gh Results The participants’ mean age was 41.3 ± 13.5 years. RIs derived for major analytes were haemoglobin (males=12.8 - 17.2; females=10.7 - 14.3 g/dL), haematocrit (39.4 - 52.1; 34.0 - 44.2%), platelet (male 115–339; female 157–402 ×1010/L), uric acid (males=231- 487; females=149 – 377Umol/L), GGT (male = 17-87 IU/L; female = 12-49 IU/L). SDR ≥ 0.4 was used as a critical value for the partitioning of RVs by sex for all erythrocyte parameters (RBC, Hb, Ht, Fe, Ferr) and platelet counts. No sex-related difference was observed in any leukocyte parameter except eosinophil counts. Again, with clinical chemistry analytes’ sex variation was observed in Ca, CK, UA, Tbil, Dbil, AST, ALT, GGT, Cre, Cl, IgM, Fe, Ferr, CA125 and the hormone analytes as their SDR ≥ 0.4. Also, age-related changes with SDR ≥ 0.4 was noted only for RBC in males for haematology analytes. Application of LAVE had a conspicuous effect on RIs for the majority of erythrocyte parameters as well as folate, Fe, Ferr, CRP, LDL-C, TG, TC, CK, Glu, AST, ALT, and GGT. Some of the RIs for Ghanaian adults are RBC (male = 4.57–6.50×1012/L; female = 4.00–5.46 ×1012/L), Hb (male =12.8–17.2 g/dL; female=10.7–14.3 g/dL), eosinophils (male =0.5–10.3%; male= 0.4–6.5%), CK (male = 93-502U/L; female 52 -276 U/L) and amylase (male= 42 – 177 Umol/L; female = 43-158 Umol/L). Conclusion The findings from the present study therefore, indicate that adopting these haematological and clinical chemistry RIs for clinical use will be beneficial to healthcare systems in Ghana since most of the RIs derived were significantly different from the ones currently in use. Sex, age and BMI related differences were mainly the sole determinants of variations among most analytes. This indicates that sex and age-specific RIs are required for the effective interpretation of some haematological and clinical chemistry analytes. The robust statistical technique used in this study makes the RI derived more reliable for clinical decision making and patient care in the country. xxv University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.0 Introduction This chapter presents a general background of the study which emphasises the need for carrying out this research. It gives a brief background to the study by providing its contextual basis. It captures the problem statement that establishes the basis for the study and the general challenge of Reference Intervals (RI) derivation as well as the need for context-based generated reference intervals. Subsequently, the general and specific objectives are presented as well as the hypotheses. The conceptual framework and the significance of the study are also presented as part of this chapter. 1.1 Background of the Study Laboratory tests and accompanying results for patients may serve varying purposes. They are used to determine diagnoses for treatment, monitoring, predicting risk, and managing therapy. Generally speaking, for the critical roles that laboratory tests play in the health care environment and the practice of medicine such tests cannot be overlooked. Reference values remain very relevant in the practice of medicine and the generic health care environment. Reference values emerged as an important concept in the 1960s to replace the obsolete term “normal values” and standardized the use of laboratory results (Gräsbeck, 2004). They can be explained by the provision of values from subjects who are relevant controls for the patients under study or by comparison. Reference Interval (RI), according to the International Federation for Clinical Chemistry (IFCC) and Clinical Laboratory Standard Institute (CLSI), is the set of results obtained by either observation or quantitative measurement of an analyte in a selected group of healthy 1 University of Ghana http://ugspace.ug.edu.gh individuals based on well-defined criteria (Clinical and Laboratory Standards Institute, 2000). Thus, to interpret laboratory data, the recorded test values are matched with RI which is the range that contains reference or normal values (Solberg, 2006). It serves as a comparison tool, thus an important determinant of whether an individual is healthy or not (Horn & Pesce, 2003). Because with the predetermined RI, which represents a value recorded for healthy persons, any value that falls outside the reference interval is considered abnormal and must be subjected to further examination. Apparently, RI remains one of the most extensively used medical decision-making tool (Abebe et al., 2018; Geffré et al., 2009; Hallworth, 2011). Considering the relevance of RI for laboratory test interpretation, the IFCC has recommended that each laboratory obtains its own reference values and based upon which to estimate the corresponding RIs within defined procedures (Bakan et al., 2016). Thus; in this regard, the CLSI published the C28-A3 guideline in 2010 that describes recommendations on RI (CLSI, 2010). Even though the recommendation of CLSI is desirable, several clinical laboratories are unable to apply their own RIs due to factors such as cost and inadequate representative population and subpopulations (Bakan et al., 2016). Hence, the inability of many clinical laboratories to generate their own RIs (Xia et al., 2016) means they largely rely on alternatives such as, what is reported in extant literature and manufacturers’ instructions or instrument manuals. Of course, this has so many implications as there are existing structures that may contribute to differences in RIs depending on the context within which they were produced. For instance, age, sex, ethnicity, lifestyle, culture, nutrition, etc. have been identified to affect RIs as well as the instrument or methodology used (Ichihara, 2014). 2 University of Ghana http://ugspace.ug.edu.gh To overcome the challenge of individual laboratories’ inability to produce their own RIs, the concept of multicenter study for derivation of global (more general purpose) RIs emerged, (Ceriotti, Hinzmann, & Panteghini, 2009) as well as the use of hospital-based data by the application of varying criteria for RI derivation (Concordet, Geffré, Braun, & Trumel, 2009). Consequently, in a more recent development relating to RIs, two papers have been published by the IFCC Committee on Reference Intervals and Decision Limits (the C-RIDL) which included a protocol and standard operating procedures (SOPs) for multicenter RI studies (Ichihara, 2014; Ozarda, Ichihara, Barth, & Klee, 2013). This has the intent of traceability such that a serum panel with assigned values that can be traced to the Reference Measurement Procedures (RMP) is a part of the multicenter study (Xia et al., 2016). Overall, such development is expected to bring about the alignment of test results among laboratories and for promoting common use of RIs through standardization (Bakan et al., 2016; Ichihara et al., 2013b; Xia et al., 2016). The main purpose of the multicenter RI study is to compare the reference values through harmonization of procedures and to investigate the regionality and ethnicity of reference values on a global scale to contribute to globalized medical practices. In the guidelines of IFCC (C28-A3), some defined criteria are to be met before a multicenter RI study can be performed. Thus, a priori selection of reference subjects (i.e. inclusion and exclusion criteria, etc.), and clear definition of the pre-analytical phases (i.e. collection of blood sample, processing, etc.) (CLSI and IFCC, 2008; Ozarda et al., 2013). Additionally, there is the need for well defined analytical phase, standardization and traceability of results, quality control, and statistical procedures for analysis of data and 3 University of Ghana http://ugspace.ug.edu.gh results reporting (CLSI and IFCC, 2008; Ozarda et al., 2013). The popularity of multicenter study is fast growing, and several studies in various countries have used this approach (Ichihara, Ozarda, Barth, Klee, Shimizu, et al., 2017; Ozarda et al., 2014; Qin et al., 2015; Xia et al., 2016). Thus, the need for standardized and globally representative common RIs that will help effective clinical laboratory investigation is the main driver behind the current multicenter studies involving five African countries. 1.2 Problem Statement The processes involved in the determination of reference interval are long, challenging, and expensive, mainly because of the need to select appropriate numbers of reference individuals according to well-defined criteria (Borai et al., 2016; Concordet et al., 2009). Establishment of RIs is a necessity or at least their verification for each analyte and specimen source in every clinical laboratory is required (Yan, Hu, Yang, et al., 2018). However, very few laboratories or manufacturers carry out their own reference interval studies (Xia et al., 2016). Over the years, efforts by both local and international groups of clinical chemical and laboratory medicine to improve laboratory test results through standardization has seen significant progress with IFCC and CLSI being instrumental in this regard. Nonetheless, the RIs for the laboratory tests which are central to interpretation of laboratory test results remain largely discordant among clinical laboratories (Dosoo et al., 2012; Ichihara, 2014; Ruzagira, Abaasa, Mulenga, & Kilembe, 2014). The exclusion and inclusion criteria for reference individuals, methods used in the computation for the RIs on collected sample values as well as the steps followed for the standardization of the analytes all affect the RI results (Ichihara, 2014). 4 University of Ghana http://ugspace.ug.edu.gh Previous guidelines’ contribution to RIs development through appropriate scientific methods cannot be overemphasized. For instance, IFCC and CLSI’s document published in 2008 which provides important basis for the standardization of RIs has been extensive in use and has had an overwhelming impact; with its effect transcending into the increase in RI studies in Africa (Dosoo et al., 2012; Karita et al., 2009; Mulu et al., 2017; Segolodi et al., 2014; Siraj et al., 2018). However, selection of reference individuals using the guideline is limited because it does not give clear procedure for such selection (Ichihara, 2014). Moreover, the need to use healthy populations for RIs is apparent but the relativity of the concept of health among different culture makes RI studies more difficult. It is, thus, acknowledged that the problem facing the laboratory scientist would be resolved by deriving RIs for healthy populations (CLSI and IFCC, 2008). More so, there should be careful attention to the impact of factors such as diet, exercise, age, sex, and non-healthy status on RI estimation (Horn, Feng, Li, & Pesce, 2001; Horn & Pesce, 2003; Ichihara, 2014). Accordingly, the challenge of samples selection and inadequate sample size are issues to be considered (Horn et al., 2001), in addition to the controversy surrounding the proposed statistical method (Ichihara & Boyd, 2010). Despite the abilitiy of some laboratories to perform local studies to generate RIs for their own use, there has been an increase in multicenter studies (Ozarda, 2016). With the recognition of its importance for standardization and traceability of results, multicenter studies have been conducted in Australia (Koerbin et al., 2015), Asia (Qin et al., 2015; Xia et al., 2016), Spain (Fuentes-Arderiu et al., 2001), Nordic countries (Rustad et al., 2004), and Turkey (Ozarda et al., 2014). This points to the relentless commitment of the scientific community to improve RIs and its impact on the practice of laboratory medicine and diagnosis at large. 5 University of Ghana http://ugspace.ug.edu.gh An Asian multicenter study which was conducted after previous studies in the region have showed regional differences in test results (Ichihara et al., 2004; 2008) produced interesting results. In this wider study in terms of geographical coverage, regionality of test results was explored (Ichihara et al., 2014). The results of the study corroborated the previous findings that found regional variation in test values. Additionally, analytes such as parathyroid hormone (PTH), high-density lipoprotein cholesterol (HDL-C), several analytes that were studied showed regional variations (Ichihara et al., 2014). Consequently, this triggered a possible global multicenter study that could produce RI results which can be shared, taking into consideration sources of variation (Ichihara, 2014; Ichihara et al., 2013a; Xia et al., 2016). Interestingly, the existing RIs for use in African countries have largely been based on studies carried out in other continents (Zeh, 2012; Dosoo et al., 2012; Koram et al., 2007; Mulu et al., 2017; Segolodi et al., 2014). The demographic characteristics of these populations may well be different from those of populations in African continent. In that respect, the use of such RIs may not be appropriate as they may not result in the required accuracy or precision that will aid health care providers in making correct clinical decisions. In addition, the laboratory manufacturers’ insert on their products and existing literature used as RIs which have been mentioned by other authors (Bakan et al., 2016; Xia et al., 2016) may be more applicable in many African countries (Koram et al., 2007), but not in others. Even though there is acknowledgement of the increasing number of RI studies in Africa since 2008 upon the publication of the IFCC and CLSI revised guidelines; Ghana, like many other African countries does not use locally generated reference interval (Karita et al., 2009; Kueviakoe, Segbena, Jouault, Vovor, & Imbert, 2011; Koram et al., 2007, Ruzagira et al., 2014; Segolodi et al., 2014). 6 University of Ghana http://ugspace.ug.edu.gh Commonly, laboratories in Ghana use RIs from the extant literature, manufacturers’ standard/guide, or RIs of high income countries. Characteristically, the RIs from those population may not reflect that of the local Ghanaina population. The variations in lifestyle, ethnicity, environment, among other factors may not lead to the ultimate accuracy desired in clinical decision-making. This is confirmed in previous studies, where locally generated RIs were found to vary significantly from other RIs that were in current use (Dosoo et al., 2012; Karita et al., 2009; Kibaya, Bautista, Sawe, Shaffer, & Sateren, 2008). A literature search indicates that three main RI studies have been conducted in Ghana (Koram et al., 2007; Dosoo et al., 2012; Asare et al., 2012). These studies specifically targeted selected haematological analytes such as Hb, WBC, RBC, lymphocytes, neutrophils, etc. Biochemical analytes the authors considered included: Urea, sodium, potassium, direct biliribun, total biliribun, albumin, amylase, ALT, AST, GGT, uric acid, total protein, etc. Geographically, these studies covered the Akuapem North district in Eastern region, Greater Accra region, and the middle belt of Ghana. However, all these previous studies that have contributed somewhat to laboratory medicine in Ghana all used the non-parametric statistical approach with few selected analytes. Moreover, there is wide acknowledgement of the benefits of standardization and traceability of results of multicenter RI studies as a global trend to boost their use. Against this background, this research, being part of an African multicenter study seeks to generate a comprehensive local RI for selected analytes among healthy Ghanaian adults using the parametric statistical approach with the application of latent abnormal values exclusion (LAVE). Thus, this study aims at deriving RIs for haematological, biochemical, and immunological analytes and to examine the effect of BMI, sex, age, and ethnicity on the reference values of Ghanaian adults. The results are compared with the existing manufacturers’ RIs 7 University of Ghana http://ugspace.ug.edu.gh currently in use in Ghana as well as those of the other African countries. 1.3 Justification of the Study Traditionally, European and North American populations have served as sources of RI for clinical laboratory parameters (Dosoo et al., 2012; Koram et al., 2007; Segolodi et al., 2014). The use of reference intervals from these populations has shown discrepancies between values when compared with healthy African population values (Kibaya et al., 2008; Tembe et al., 2014; Zeh et al., 2011). There are already published reports of differences in some biochemical and haematological parameters based on race and ethnicity (Ichihara et al., 2013; Lawrie et al., 2009; Lim, Miyamura, & Chen, 2015). Yet, using “foreign-derived” RIs to make clinical decisions locally has persisted for several decades and until now. A good explanation may be that studies on RIs in Africa are limited. The differences in reference intervals that have been reported also found within group variances; including several varying indices among different ethnic groups in Africa (Lawrie et al., 2009; Ngowi, Mfinanga, Bruun, & Morkve, 2009). In addition, higher monocyte and eosinophil levels, lower Hb, RBC, hematocrit, MCV, platelets, and neutrophils are reported for African populations when compared with Western populations (Dosoo et al., 2012; Kibaya et al., 2008; Odhiambo et al., 2015). There are some conditions that have been reported to account for the variations for both between and within these groups. These factors include ethnicity, age, sex, dietary patterns, and environmental pathogens (Romeo et al., 2009). The establishment of RIs that is representative of local population is not only part of good laboratory practice but also a recommendation made by most pathology bodies and 8 University of Ghana http://ugspace.ug.edu.gh manufacturers concerned with the different chemistry analysers. In many African countries, such values are non-existent and when available, they have frequently been derived using either small samples or populations that have not been carefully selected. In some studies, the exclusion and inclusion criteria have not been strictly enforced according to IFCC and CLSI guidelines. However, it is noted that the inclusion or exclusion of specific people as reference individuals largely impact the RI results (Ichihara, 2014). Given the critical role of laboratory test and RI in clinical decision making, the need to have appropriate reference interval for a defined population cannot be overemphasized. Thus, the lack of such evidence-based values can lead to misdiagnosis and mismanagement, which might affect its importance as an effective decision tool. It may also be associated with unnecessary investigation and waste of scarce resources. At a time of ever-growing pressure on the need for better quality of care while reducing cost; means efficiency in clinical decision making can lead to enormous benefits if well-defined RIs are available for use. Specifically, resource poor settings can reduce resource wastage while enjoying improved population health. Derivation of RIs that are representative of local populations is desirable for broader benefit of the economy and population health. Improvement in accurate diagnosis and case management by physicians means a lot for cost-saving and improved health. In particular, Ghanaian households can benefit from avoidance of unnecessary misdiagnoses and poor case management by removing the economic burden they suffer for using foreign- generated RIs for their case management. Likewise, local RIs can improve policy decisions that can help save scarce economic resources in the healthcare sector. 9 University of Ghana http://ugspace.ug.edu.gh 1.4 General Objective This study aimed at establishing RIs for haematological, biochemical and immunological analytes so as to provide context-specific perspective toward improvement of clinical decision-making in clinical diagnosis and effective case management in Ghana. 1.5 Specific Objectives The study seeks to: 1. Determine the RIs of haematological analytes among healthy Ghanaian adults. 2. Determine the RIs of biochemical and immunological analytes among healthy Ghanaian adults. 3. Assess the relationship between, sex, ethnicity, BMI, alcohol, exercise, and blood pressure on analytes for haematology, lipids profile, kidney, and liver function tests, immunoglobulins, tumour markers, thyroid and hormones. 4. Examine the impact of the application of latent abnormal values exclusion (LAVE) and without LAVE methods on RIs. 5. To compare Ghanaian RIs with existing manufacturers’ RIs currently in use in Ghana. 1.6 Research Questions 1. What are the RIs of haematological analytes among healthy Ghanaian adults? 2. What are the RIs of biochemical and immunological analytes among healthy Ghanaian adults? 3. What are the relationship between, sex, ethnicity, BMI, alcohol, exercise, and blood pressure on analytes for haematology, lipids profile, kidney, and liver function tests, immunoglobulins, tumour markers, thyroid and hormones?. 10 University of Ghana http://ugspace.ug.edu.gh 4. What is the effect of latent abnormal values exclusion (LAVE) and without LAVE on RIs? 5. Are there differences between Ghanaian RIs and the existing manufacturers’ RIs currently in use in Ghana? 1.7 Conceptual Framework Figure 1.1:Conceptual Framework of some factors influencing the derivation of reference intervals of analytes The factors influencing the RIs of analytes are succinctly reviewed using the Figure 1.1 above. Reference intervals serve as a comparative tool with which laboratory test results are compared with to determine either an individual is healthy or not. RI as well as clinical presentation, imaging and other methods of diagnosis serve as medical decision-making 11 University of Ghana http://ugspace.ug.edu.gh tool for diseases diagnosis and management. Due to its importance, it is necessary for its derivation to be accurate. The factors affecting the RI of analytes shown in the framework are divided into four main subgroups. These are biological variation, statistical method, anthropometric measurement, and lifestyle. Under each subgroup the specific indicators that directly influence the reference intervals of analytes are indicated. There are certain uncontrollable biological variations that affect the reference intervals of analytes, such factors include age and sex. Age and sex are the most significant factors that determine health or disease (Ichihara et al., 2008). The physiological and hormonal changes associated with ageing means derivation of RIs must consider age as an important source of variation. Although some analytes’ RIs are age-adjusted, most of them are age- dependent. Several studies have confirmed that lipids tests such as total cholesterol, triglyceride, and low-density lipoprotein levels increase with age (Carroll et al., 2005; Kuzuya, et al., 2002; Abbott et al., 1998). Further, sex has a major impact on the reference intervals of analytes because males and females show significant differences in biological and physiological make up. These differences in biological constituents suggest there exist possible variations in the monitored analytes that are measured among these two groups. It is generally acknowledged that many parameters depend on the hormone set and physical make-up. For instance, studies have shown that men have higher levels of creatine kinase, ALT, AST, ALP, and haemoglobin than women (Dosoo et al., 2012; Eller et al., 2008; Neal, Ferdinand, Yčas, & Miller, 2009). Furthermore, ethnicity or race is a key determinant of the outcome of the RI results. Based on several genetic factors that come with race or ethnicity, RIs have been identified to vary significantly between different ethnic and racial groups. Studies conducted in some 12 University of Ghana http://ugspace.ug.edu.gh African settings affirm racial variations of some haematological analytes (haemoglobin, haematocrit, red blood cells, and white blood cells) between Africans and Caucasians (Kibaya et al., 2008; Tembe et al., 2014; Abebe et al., 2018). Again, previous studies have found variations among some clinical analytes such as AST, ALT, and GGT in African populations (Samaneka et al., 2016; Saathoff et al., 2008; Segolodi et al., 2014; Abebe et al., 2018; Kibaya et al., 2008). These show the important role ethnicity plays in the derivation of RIs, hence the use of Caucasian RIs in different ethnic groups are inappropriate. Therefore, all reference populations used for derivation of RIs of various analytes must represent the ethnicity of the respective populations. Derivation of reference interval involves various statistical methods. In reference interval derivation, after laboratory analysis of the blood samples from the reference population is obtained, statistical techniques are employed to compute the RIs. The statistical process involves the normal distribution of reference values to obtain the 95% confidence interval, detection of outliers, and partitioning computation. There are two main methods that are commonly used for calculation of the reference interval from study data. These are parametric and non-parametric methods. The parametric method is used when the data follows a Gaussian distribution or when it can be transformed to a normal distribution using appropriate transformation methods whiles the non-parametric method is used for skewed and non-normal data distributions. The lower reference limits are estimated as the th th 2.5 percentile and the upper limits as the 97.5 percentile of the distribution of test results for the reference population. Different statistical methods used in the derivation of the reference intervals have shown significant variations with the results of the RI. 13 University of Ghana http://ugspace.ug.edu.gh The RIs derived by the parametric method with transformation are narrower than those for the non-parametric method. This may be due to the ability of the parametric method to identify and exclude extreme values during computation of RIs. Also, one important factor to consider when deriving RIs is the need to identify and exclude outliers to improve the reliability of the results. Ichihara et al. (2010) suggested the latent abnormal value exclusion (LAVE) method which is a secondary exclusion method to eliminate possible abnormal results hidden within the reference values (Ozarda, 2016). The application of the LAVE method to exclude individuals with latent diseases from the reference values would affect the RIs of the analytes. In addition, lifestyles which consist of alcohol, smoking, and exercise are also factors that influence the outcome of the reference interval of analytes. The effect of alcohol may vary subject to the extent and duration of use. Transient effects of alcohol consumption may increase plasma lactate and serum glucose, with lower excretion of uric acid because of the intake of hepatic gluconeogenesis. Also, excessive intake of alcohol affects the levels of gamma-glutamyl transferase (GGT) as well as mean cell volume (MCV) which are widely used to test for high alcohol consumption. Likewise, smoking can have both chronic and acute effects on investigations. Within an hour after smoking up to five cigarettes, serum/plasma concentrations in fatty acids, glycerol, adrenaline, cortisol, and aldosterone may rise. Chronic smokers may have significant rises in leukocyte, lipoproteins, hematocrit, and tumour markers, and reduction in the activities of enzymes. The intensity of exercise may also affect the RIs of analytes. For instance, the most common cause of increasing creatine kinase (CK) levels is exercise. CK levels can elevate due to intense exercise for several days to a week. Thyroid function 14 University of Ghana http://ugspace.ug.edu.gh is also supposed to change among people going through high-intensity workout. For instance, FT4 and TSH levels can increase in anaerobic exercise but it reduces FT3 (Ciloglu et al., 2005). Lastly, anthropometric indicators including body mass index (BMI), waist circumference and blood pressure are widely used to predict increased chronic disease risk for conditions such as hypertension and diabetes. These anthropometric indicators play an important role in recruiting the healthy population for RI derivation. BMI, waist circumference, and blood pressure may have effect on the RIs of some analytes such as lipids (Shen et al., 2012). 15 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.0 Introduction This chapter provides an overview of previous research on Reference Intervals (RIs). In a broader scope, it covers theoretical, conceptual and empirical literature. Specifically, it introduces the concept and framework for derivation of RI, focusing on the trend of change in the field of laboratory medicine associated with RI derivation, factors and challenges associated with RI derivation, the common analytes used in human medicine, empirical literature, and the different statistical approaches involved in the derivation of RI. Moreover, the methodological weaknesses of previous studies and other gaps in the literature are discussed. 2.1 Concept of Reference Intervals There is a long stock of studies that has affirmed the variations of the human system to physiological processes, genetic difference, and diseases (Brodin et al., 2015; Emilsson et al., 2008; Jirtle & Skinner, 2007; Nicholson & Wilson, 2003; Varki, Geschwind, & Eichler, 2008; Weiss, 1993; Whitney et al., 2002). Hence the rational interpretation of laboratory results demands understanding of the variation of these components of the individual under study. The importance of RIs is heightened, with an estimated 4 out of 5 clinical decisions dependent on laboratory reports (Katayev, Balciza, & Seccombe, 2010). Usually, it is the comparison of test results with RIs that give meaningful interpretation for decision making by physicians (Fontes, Coeli, Aguiar, & Vaisman, 2013; Katayev et al., 2010; Ozarda, 2016). The traditional term “normal values” was previously used as a basis of comparison for 16 University of Ghana http://ugspace.ug.edu.gh laboratory tests. However, Grasbeck & Saris (1969) asserted that the concept was scientifically flawed; they therefore suggested a well-defined nomenclature together with recommended procedures be established (Gräsbeck, 2004). This therefore necessitated the replacement of “normal values” with a more appropriate term that takes care of biological and pathological processes. Thus, the demand for a more precise and comprehensive interpretation became necessary as “normal values” was viewed as inadequately defined (IFCC & International Committee for Standardization in Haematology (ICSH, 1987). Subsequently, the alternative concept “Reference Values” was proposed in 1965. Since then, the corresponding theory and terminology have been elaborated upon by international bodies, with the IFCC spearheading this course. Reference values can be explained by the provision of values from subjects who are relevant controls for the patients under study, or by comparison (IFCC & ICSH, 1987). The “RI is simply by threshold values between which the test results of a specified percentage (usually 95%) of apparently healthy individuals would fall” (Boyd, 2010, p. 84). In essence, any person whose test value is within the RI has a lower chance of having a particular condition and vice versa (Ceriotti, 2007). The concept of RI seems dynamic albeit the extensive use and acceptance of IFCC’s recommendations on reference intervals. In fact, the primary ideas are kept as foundation, yet, there are ever growing developments that further the evolution of the concept. Henny et al. (2000), for instance, suggested a “need to revisit the concept of reference values”. In fact, there are ongoing debates about some of the principles surrounding RI derivation – reference population, statistical methods, etc. These continuous developments regarding RI are what Siest et al. (2013) refers to as “unfinished symphony”. 17 University of Ghana http://ugspace.ug.edu.gh The past three decades has witnessed significant developments and implementation concerning RIs (Fuentes-Arderiu, 2006; Harris, Wong, & Shaw, 1991; Ichihara et al., 2013; Ozarda et al., 2013). The C28-A3 guideline, dubbed “defining, establishing, and verifying reference intervals in the clinical laboratory” offers fundamental procedures that are necessary for RI derivation studies (CLSI and IFCC, 2008; Ozarda, 2016). Moreover, a revised document titled “EP28-A3C- defining, establishing and verifying reference interval in the clinical laboratory” by the CLSI in 2008 and published in 2010 improved the earlier guideline. It includes “recommendations for transference and validation of RI from other sources and the use of robust methods for verifying RI from small sample sizes of 20 healthy subjects” (Clinical and Laboratory Standards Institute, 2010). This recommendation was given against the backdrop that many laboratories faced the challenge of conducting a population-based reference interval study by meeting the minimum 120 or more reference population required in the C28-A3 guideline. Of course, this approach remains a challenge because the credibility is always under scrutiny since details of design, reference population selection and statistical techniques are usually deficient. 2.2 Concept of Multicenter Reference Interval Studies In recent times, it is the concept of multicenter RI studies that has grown in prominence and continues to gather momentum. Unsurprisingly, it has been branded as one of the most important developments in the area of RI derivation (Ozarda, 2016). One of the utmost objectives of multicentric studies is to achieve standardization, traceability and transferability. Against the backdrop of recent developments in RI derivation, a protocol and standard operating procedures (SOPs) for multicenter RI studies were produced by the IFCC/C-RIDL (Ichihara et al., 2013a; Ozarda et al., 2014). Transferability of test results 18 University of Ghana http://ugspace.ug.edu.gh for specific analytes is a desirable aim of the laboratory community (Boyd, 2008). However, to achieve this, standardization of methods that will ensure traceability to reference standards is said to be a requirement. Admittedly, transferability has remained elusive because of the variations between populations (Boyd, 2008). This means that test results cannot be easily used or compared with other populations that are heterogenous. If test results were transferable, then the use of common RIs will be easier even across countries and regions. Based on the above challenge the IFCC/C-RIDL resolved to develop common RIs for homogeneous groups independent of the laboratory that measures it. In effect, the global multicenter study on RI was launched in 2011 with the following objectives: ● to establish country-specific RIs from 500 or more healthy individuals in a harmonized manner using a common protocol (Ichihara et al., 2013c). ● to make test results comparable across participating countries through the common measurement of a specified panel of sera on the basis of linear regression analysis (Ichihara et al., 2013). ● to explore sources of variation of the aligned test results using information obtained from a questionnaire given to each individual in the pool (Ichihara et al., 2013c). For the purposes of global standardization and achievement of the above objectives, the IFCC/C-RIDL presented a proposed protocol and SOPs for common use in conducting multicenter RI studies on a national or international scale (Ozarda et al., 2013). Some countries such as Turkey, South Africa, China, etc. and the Scandinavia adopted the multicenter studies on the derivation of RIs, which has resulted in fascinating but 19 University of Ghana http://ugspace.ug.edu.gh contrasting findings. For example, while no between-country variations for RI was observed among the Scandinavian groups (Rustad et al., 2004), significant differences in RIs were observed between countries in Asia (Ichihara et al., 2004, 2008). These contrasting findings make further studies a necessity and determination of the sources of variation more imperative. Following from this, statistical methods employed in analyzing the between-laboratory differences and the appropriateness of the study protocol have become a focal point of discussion in the literature. Against this background, Ichihara et al. (2017b) conducted a landmark study that explored rational and harmonizable procedures for RIs derivation as well as assessment of the possibility of sharing RIs through assessment of sources of variation (SV) of RVs on a global scale. This is a global multicenter study that involves 17 countries across 5 continents with 13,386 apparently healthy adults. Lenient criteria to facilitate the recruitment was used against the backdrop that LAVE method will be employed. A clear exclusion criterion was also defined. In the data analysis, appropriate statistical techniques such as major-axis linear regression analysis (MALRA) was used for all pairwise method comparisons (Ichihara et al., 2013c). Again, Gaussian transformation using modified Box- Cox formula; ANOVA for measuring between-country differences and standard deviation ratio (SDR) for expressing the magnitude of variation between countries were used (Ichihara et al., 2017). The study found that the use of LAVE methods significantly impacted the results of analytes that are related to inflammation and nutrition. This pointed to the need for the use of the LAVE method for secondary exclusion. Moreover, the results indicated that no within-country variations were observed in China, Japan, Saudi Arabia, and Turkey (Xia 20 University of Ghana http://ugspace.ug.edu.gh et al., 2016; Ichihara et al., 2013a; Borai et al., 2016; Ozarda et al., 2014). Among South Africans, ethnic distinctions were defined broadly into Africans and non-Africans. BMI varied widely among the countries and was identified as a possible confounder. Other sources of variation like sex, age, alcohol intake, smoking coupled with BMI were found to affect some analytes. The study also found that there was major preference for the parametric method using the modified Box-Cox formula in almost all circumstances (Ichihara et al., 2017a). Moreover, a second part of the study explored sources of variation against the background that previous studies have found region-related variations in RIs of some analytes (IgG, C3 and CRP) (Ichihara et al., 2008). In addition, a study in East and South-East Asian countries which sampled 3500 healthy individuals and screened 72 analytes, show significant regional differences (Ichihara et al., 2013a). Based upon the above findings, the authors concluded derivation of RIs based on rational methods and comparison of RVs which are dependent on dataset obtained in a synchronized way is possible (Ichihara et al., 2017a; Ichihara et al., 2017b). Moreover, some researchers have confirmed that changes in RV that are sex, age, alcohol-related which are analogous globally is important for RVs and RI derivation (Ichihara et al., 2017b). Regardless, each analyte’s unique relationship with sex and age may be relevant in the practice of medicine than fixed RIs derived without cognizance of age and sex (Ichihara et al., 2017b). The benefits associated with the multicenter studies with regard to country-specific derivation of RIs has resulted in the widespread adoption of this approach. 2.3 Factors Affecting RI Derivation There are several factors considered in establishing a well-controlled and reliable reference interval. Some of such challenging factors include appropriate reference population 21 University of Ghana http://ugspace.ug.edu.gh selection, pre-analytical and analytical variables management. Additionally, application of correct statistical approach to derive the intervals is critical to the outcome of the RI results. These factors and how they are addressed in the research process are believed to affect the results significantly. In as much as these factors are critical to RI derivation, they are readily modifiable, which means the investigators can exercise great control over them to ensure accuracy of RI results. 2.3.1 Recruitment of the Reference Population. Arguably, recruiting the appropriate number of people and the right healthy population is critical in RI derivation studies. This is also an overwhelming task in RI derivation studies because recruitment of eligible subjects as reference population could be both cost and time intensive. Apparently, it is one of the main reasons why individual laboratories find it difficult to derive their own RIs. The sampling techniques may be simple random, direct or indirect, a priori or a posteriori sampling. Moreover, it may reflect a normal population of volunteered blood donors, or direct door-to-door recruitment. Intriguingly, whichever approach that may be used for the selection of the subjects, there is always a potential that the sample may not be representative of that population (Boyd, 2010). It is always a question of which clinical variable is of interest. A question which when answered or well- defined shapes the category of subjects to be considered healthy or “normal” for participation. Undoubtedly, in as much as efforts are made to streamline the subject of choice and eligibility, volunteering cannot guarantee that the sample will be entirely unbiased (Boyd, 2010). Often it becomes difficult to qualify people as non-disease or healthy since many diseases may be at the asymptomatic stage. This biases the selection of subjects as reference 22 University of Ghana http://ugspace.ug.edu.gh individuals (Miller et al., 2016). In some cases, however, it is the entirety of a disease condition which influences a laboratory test result that might not be readily known. In that respect, it is obvious that if secondary exclusion is not applied to the analysis the RI results would be affected. A fascinating situation is where volunteers may see the exercise as an opportunity for free medical evaluation and join even though they may not be healthy. Adequate sample size selection is another issue that is often experienced during RI derivation studies. Using the appropriate sample size makes the results certain and meaningful. In most cases, a small sample size of 20 is required when it is only to serve the purpose of validation. Often times, population-based studies require larger sample to serve as reference population. As mentioned earlier, the recruitment of subjects is very costly and time-intensive which means larger sample size makes it difficult to obtain. In fact, when there is a need for partitioning by specific parameters, a larger sample size is required for meaningful RIs (Miller et al., 2016). Moreover, the use of the word “healthy” is subjective (e.g., all the population must meet some specific criteria) therefore, there should be a well-defined criteria. Subsequently, the reference population should be randomly sampled from the reference individuals based on the defined criteria. 2.3.2 Pre-Analytical Variables Pre-analytical phase involves the preparation for collection of the biological material, the sample collection, storage and the transportation of the collected sample to the laboratory, and preparation of the sample for the assay. Over time, this stage has been identified as the main source of error in laboratory analysis (Bonini, Plebani, Ceriotti, & Rubboli, 2002; Hammerling, 2012; Hawkins, 2012; Plebani, 2010). Specifically, it is found that up to 46 – 68% of erroneous results are due to non-adherence to guidelines or procedures (Bonini 23 University of Ghana http://ugspace.ug.edu.gh et al., 2002). A typical example is serum versus plasma, sample collection time, and hemolysis for potassium, CK, AST and LDH. It is amidst this background that the necessary protocol and SOPs should be strictly followed so as to minimize the impact of this stage on RI results. Furthermore there are controllable and uncontrollable factors that may affect the results at this stage. These are discussed below: 2.3.2.1 Controllable Factors Before the Sample Collections Diet The food intake of subjects before their participation in the reference interval study is very important, especially the time immediately before the blood sample collection. This is one of the factors that is controllable. It is generally recommended that subjects should fast at least 8 – 12 hours before sample collection. In the absence of fasting by subjects, there is the possibility of observing increased levels of some metabolites because of ingested nutrient metabolism. For example, it is common to observe elevated levels of some analytes such as glucose, triacylglycerol, free fatty acid and lipoprotein after meals (Švagera & Šigutová, 2016). Similarly, elevated serum triacylglycerol concentration can be observed in subjects whose diet is rich in fats. It is also the case that increased ammonia and urea levels are found among protein-rich foods consumers; at same time there is release of postprandial hormones (e.g. insulin which decreases the level of potassium and phosphate) (Švagera & Šigutová, 2016). Moreover, the influence of diet on laboratory results is further confirmed by Lippi et al. (2010), who found that hours (e.g. 1, 2, 4) after meals, haematological parameters such as neutrophil and MCH increased significantly whereas lymphocyte, monocyte, RDW, HCT, 24 University of Ghana http://ugspace.ug.edu.gh MCV decrease significantly. The authors confirmed that some of the variations are clinically significant for neutrophils, eosinophils, RBC, hematocrit and MCH. Physical Activity In general, reports in the literature on the influence of exercise on laboratory test results are scattered. Yet, there is a common conclusion that exercise or physical activity has impact on the results of laboratory tests but this depends on the duration and intensity of the exercise (Sanchis-Gomar & Lippi, 2014). For instance, increases in exercise intensity might explain a substantial drop in red blood cell (RBC) and haemoglobin (Hb) concentration (Sanchis-Gomar, Banfi, Pareja-Galeano, Martinez-Bello, & Lippi, 2013). Elsewhere, ALT, CK, AST, and LDH are found to be affected by muscular activities (Brancaccio, Lippi, & Maffulli, 2010; Colombini et al., 2012; Lippi, Schena, Montagnana, Salvagno, & Guidi, 2008). Based on the foregoing, it is reasonable for investigators to always pay critical attention to these details and possibly measure the role of physical activities. This is particularly vital in RI derivation since it is going to be a critical tool for clinical decision making. In this regard, this study has as part of its objectives to measure the role of the type and intensity of exercise on the reference intervals to be generated. Medication Another factor that is vital when conducting reference interval study is the consideration for subjects who are on medication. Of course, the standard protocol and SOPs are categorical on which drugs or medication are permissible. However, there is evidence that several drugs have an impact on laboratory results (Švagera & Šigutová, 2016). Moreover, interference of analytical assay procedure may be caused by some drugs such as vitamin 25 University of Ghana http://ugspace.ug.edu.gh C because of the strong reduction properties it possesses; thereby causing a false decrease in the level of analytes detected using peroxide (Švagera & Šigutová, 2016). 2.3.2.2 Uncontrollable Factors before the Sample Collections Uncontrollable factors before biological material collection include age, gender, race and biological rhythms. Age In epidemiology, age is a good determinant of health or disease. Generally, old age and early years of life come along with different states of health. Similarly, in the derivation of RIs, attention must be given to the role of age in the derivation process. In fact, research evidence shows that most monitored analytes are age-related (Hoffmann, Nabbe, & van den Broek, 2015; Park et al., 2016; Qiao et al., 2014). For example, high level of cholesterol is recorded among older people than young people. This particular concern of age influencing RI values has led to some researchers recommending the need to partition RI results according to age (Ichihara et al., 2017a; Ozarda, 2016). An issue this study seeks to address by partitioning RI results with respect to age. Gender Males and females show significant differences in biological and physiological makeup. These differences in biological constituents mean there exist possible variations in the monitored analytes that are measured among these two groups. It is generally acknowledged that many parameters depend on the hormone set and physical make-up. For example, men have higher levels of CK, ALT, AST, ALP, uric acid, urea, haemoglobin, ferritin, iron and cholesterol than women (Dosoo et al., 2012b; Kone et al., 26 University of Ghana http://ugspace.ug.edu.gh 2017; Koram et al., 2007; Švagera & Šigutová, 2016). Consequently, several studies have confirmed that there is always the need to distinguish between females and males when using laboratory test results. This current study takes this into account and will establishes reference intervals for males and females separately. Ethnicity/Race Ethnicity or race is a fundamental key determinant of the outcome of the RI results. Based on several genetic factors that come with race or ethnicity, RIs have been identified to vary significantly between different ethnic and racial groups. Some empirical studies have confirmed this (Ichihara et al., 2013; Lawrie et al., 2009). It is also noteworthy that this PhD work seeks to examine the ethnic-specific RI variations in Ghana. 2.3.3 Analytical Variation. This also remains one of the most critical phases of the RI derivation process. Of course, variations during the analytical stage could be attributed to imperfections in the testing methods or equipment; which may cause values to differ slightly when measured over time. This however, has been acknowledged to have improved because of improved technology which means that other factors and sources of variation may be more critical than the analytical variation. In fact, for the purposes of traceability of test results, reference measurement systems and standard reference materials should be given the necessary attention in RI studies (Jones & Barker, 2008). This will allow for comparison of results from different laboratories. 27 University of Ghana http://ugspace.ug.edu.gh 2.4 Statistical Methods in Calculating Reference Interval Once a series of results are generated from the laboratory for the selected reference population, the RIs can be computed (Boyd, 2010). Generally, at this stage detection of outliers, partitioning and confidence interval computation are important. There are two main methods that are commonly used for calculation of the RI from study data – parametric and non-parametric methods. The parametric method is used when the data follows a Gaussian distribution or when it can be transformed to a normal distribution using appropriate transformation methods. 2.4.1 Non-parametric Method. The non-parametric method is used for skewed and non-normally distributed data. The lower reference limits are estimated as the 2.5th percentile and the upper limits as the 97.5th percentile of the distribution of test results for the reference population (Boyd, 2010). It is worthy of note that, 5% of all results from healthy individuals will fall outside of the reported RI and as such will be flagged as being ‘abnormal’ (Boyd, 2010; Ozarda, 2016). Over the years with advancement in computing, resampling methods, weighted percentile estimation and smoothing techniques have increased the precision of the non- parametric approach (Harris & Boyd, 1995; Shultz et al., 1985). The process invove in estimating reference intervals using the non-parametric procedure does not assume the probability distribution of the observed reference values (Solberg, 1987). 2.4.2 Parametric Method. The parametric approach has the assumption that the observed values, or some mathematical transformation of those values, follow the Gaussian or ‘‘normal’’ probability distribution (Ozarda, 2016). In parametric method, the central 95% boundaries are 28 University of Ghana http://ugspace.ug.edu.gh specified by the mean ± 2SD in the Gaussian data (Boyd, 2010). Data may not follow a normal distribution, so the parametric method can be applied after data transformation which makes it Gaussian by applying the appropriate transformation method (e.g. logarithm). The most suitable transformation method must be chosen and testing is then applied to determine whether the transformed reference values conform to Gaussian distribution (Ozarda, 2016). Though the C28-A3 approves the non-parametric calculation method, the RIs calculated by the parametric and non-parametric methods were compared in the recent IFCC, C- RIDL study which concluded that the results of the two methods are very close and parametric methods can also be used as a first choice (CLSI and IFCC, 2008). Subsequently, scholars such as Ichihara has been one of the advocates for the parametric method in RI derivation, adopting the method in several of his studies (Ichihara et al., 2008; Ichihara, Ozarda, Barth, Klee, Qiu, et al., 2017b; Ichihara, Ozarda, et al., 2013; Ichihara & Boyd, 2010). One of the main factors during the statistical analysis is the detection of outliers because they can significantly confound the results. Hence, whether it is parametric or non- parametric method there is an urgent need to identify and exclude outliers to improve the reliability of RI results. Ichihara et al. (2010) suggested the latent abnormal value exclusion (LAVE) method which is a secondary exclusion method to eliminate possible abnormal results hidden within the reference values (Ozarda, 2016). LAVE an iterative approach for the derivation of multiple reference RIs simultaneously, when no exclusion of values has been made in the initial calculation of the RIs (Ozarda, 2016). 29 University of Ghana http://ugspace.ug.edu.gh Partitioning has also become a common feature in RI derivation because of the widespread variations recorded in the literature. The suggested partitioning method by Harris & Boyd (1990), in which the means and standard deviations of the subgroups are considered as a separate standard deviation that may produce different limits is the most commonly used method. Though, this method is only appropriate for analytes with a normal distribution with subclasses, where the values are of similar size and standard deviation (Ozarda, 2016), Ichihara and Boyd (2010) proposed nested analysis of variance (ANOVA) as a method of choice for partitioning because of its ability to handle multiple groups and being able to adjust for multiple factors. The sensitivity of the population-based RIs can be increased and so, the worth of RIs is improved by stratification according to age, gender, race, ethnicity and lifestyle (Ozarda, 2016). 2.5 Common Analytes in Human Medicine This section presents brief overview of common analytes that are used in human medicine. 2.5.1 Haematological Analytes (Complete Blood Count). Complete blood count (CBC) is the most common investigation performed on patients. The wealth of information provided by this multiparameter analysis is essential for several purposes. For example, it represents an important basis for disease diagnosis, screening for blood donors, and assessment of overall health. Indeed, it can lead to the discovery of specific haematological condition that may trigger further investigations for disease management. Clinicians are often informed of a primary haematological disorder by deviations in the CBC in relation to existing reference intervals. Thus, generally, the primary focus is whether an individual is anaemic, whether the White Blood Cell (WBC) 30 University of Ghana http://ugspace.ug.edu.gh show evidence of infection, and whether the platelets level affects haemostasis. There are about 20 parameters that are considered during CBC which are broadly classified into three categories – White Blood Cell (WBC), red blood cells (RBC), and platelets. Total WBC is very useful yet, it is important to be guided by the absolute count of each cell type, as the total WBC may be misleading. 2.5.1.1 Neutrophil. In adults, neutrophil accounts for about 40% - 70% of all WBC, with a normal range of 2.0 – 8.0 x 109/L. Neutropenia is a granulocyte disorder characterized by a dramatically low number of neutrophils, the most important type of white blood cell (Tirali, Yalçınkaya Erdemci, & Çehreli, 2013). On the other hand, high neutrophil count may result from acute inflammation – heart attack, infarction and necrosis, stressor (e.g. cigarette smoking or heavy exercise) can cause it too. Drugs like lithium, clozapine, and adrenalin can increase the neutrophil count as well. However, when the elevation is persistent it is an indication of chronic myeloid leukaemia (CML). 2.5.1.2 Lymphocytes. Circulating WBC is made up of about 20-40% lymphocytes with a normal concentration of 1.0 – 4.0 x 109/L. Although lymphocyte counts outside the range is considered abnormal, low lymphocytes are usually negligible, a condition known as lymphocytopenia. HIV patients usually experience a characteristically low count in the later part of the infection when the CD4+ T cells are destroyed. On the other hand, high lymphocyte counts (lymphocytosis) may occur in the presence of viral hepatitis or other acute infections. 31 University of Ghana http://ugspace.ug.edu.gh 2.5.1.3 Monocytes. Monocytes make up about 3 – 8% of WBC with normal concentration between 0 – 1.0 x 109/L. In the event of low monocyte count, the condition is called monocytopenia which is also an unusual finding especially when the decrease is in isolation. It is however not uncommon to witness this condition in an extremely overwhelming bacterial infection. It is noteworthy that isolated increase in the monocyte count alone is rare, yet such finding may be associated with dialysis, tuberculosis, chronic inflammatory conditions and a precursor of myelomonocytic leukaemia. 2.5.1.4 Basophils. Basophils are the least common among the WBC constituting only about 0.01 – 0.3% of WBC; with a normal reference interval of 0 – 0.2 x 109/L. Basophils are capable of the production of histamine and phagocytosis. Basophilia which indicates high basophil count is rare, yet when the level is elevated significantly it presents an important indication of myeloproliferative disorder. 2.5.1.5 Eosinophils. Eosinophils constitute about 1-6% of WBC, with a normal concentration of 0 – 0.5 x 109/L. High counts of eosinophil also known as eosinophilia is usually caused by allergic conditions such as asthma and hay fever and parasitic infection. 2.5.1.6 Platelets. Platelets are produced by budding off from megakaryocytes in the bone marrow. Each megakaryocyte produces between 5000 to 10000 platelets. Platelets circulate for about 7 to 14 days and are destroyed by the spleen and liver. The reference interval for a normal 32 University of Ghana http://ugspace.ug.edu.gh platelet count is 150 – 450 x 109/L. Significantly low platelet count is referred to as thrombocytopenia (<100 x 109/L). When it is severe, it is associated with increased risk of bleeding. The condition is particularly noteworthy if complemented with other changes in the CBC. Isolated cases of thrombocytopenia may be caused by viral infection, liver disease, autoimmune disease, HIV infection, medications etc. Conversely, thrombocytosis (elevated platelet count) may not necessarily be a clinical issue but only a reactive change. Thus, only chronic persistent increase over six months should trigger further clinical investigation. 2.5.1.7 Haemoglobin (Hb). Haemoglobin (HB) is a protein in RBC that carries iron. This iron holds oxygen, making haemoglobin a vital component of the human blood. The normal concentration of Hb is between 13 – 17.5 g/L for males and 11.5 – 16.0 g/L for females. The extant literature show differences according to gender and ethnicity (Dosoo et al., 2012). In most cases, low levels of Hb have been associated with anaemia. Nevertheless, wide differential diagnosis narrows significantly when Hb level is checked with mean cell volume (MCV). It is also commonly observed that iron deficiency, thalassaemia and anaemia of chronic disease are associated with low levels of Hb. As iron deficiency is the most common cause of microcytic anaemia, serum ferritin value serves as important additional information for clinical decisions and usually appropriate to be considered as the first step. If thalassaemia trait is suspected in the presence of low ferritin it is important to correct the iron deficiency before requesting a haemoglobinopathy screen. Additionally, bleeding, haemolysis, chronic inflammation, etc. may also cause normocytic anaemia, whereas alcoholism, liver disease, folate deficiency, thyroid disease may cause macrocytic anaemia. Elevated Hb, 33 University of Ghana http://ugspace.ug.edu.gh and packed cell volume or “haematocrit” levels, it may reflect decreased plasma volume or increased red cell mass (polycythaemia). 2.5.1.8 Other Red Blood Cell Indices (RDW, Red cell count, MCH, MCHC). The combined red cell indices reflected in red cell distribution width (RDW), red cell count, mean corpuscular haemoglobin (MCH) and mean corpuscular haemoglobin concentration (MCHC) are useful for building a picture of many patterns suggested by haemoglobin level. In most cases, isolated abnormalities in any of these parameters are usually insignificant and negligible. Moreover, not all the indices are reported by all laboratories. 2.5.2 Biochemical Analytes. 2.5.2.1 Calcium (Ca). The calcium in human body is mainly stored in bones. Calcium is largely accessible through food and dietary supplements as well as some medicines. Milk, and milk products, green leafy vegetables, beef, seafoods, bony fish, etc. are important dietary sources of calcium. A low blood level of calcium is referred to as hypocalcaemia. This condition makes the nervous system highly irritable, leading to tetany, abdominal cramps, etc. When the deficiency of calcium is chronic, it contributes to poor mineralization of bones, osteomalacia and osteoporosis. On the other hand, hypercalcaemia happens mostly in hyperparathyroidism and cancer patients. It is interesting to know that females are more affected by hypocalcaemia than males (Egbuna & Brown, 2011). 34 University of Ghana http://ugspace.ug.edu.gh 2.5.2.2 Creatine Kinase (CK). Measurement of serum CK is a vital part of assessing myalgia, myopathies or rhabdomyolysis patients. There are several non-modifiable factors that have been identified as critical to CK levels including sex and race (Brewster, Coronel, Sluiter, Clark, & van Montfrans, 2012; Prince, Abbott, Lee, Oliver, & Olson, 2015; Yen et al., 2017). In addition, CK levels experience small reduction as people age (Lilleng et al., 2011). Elsewhere, significant elevation of CK levels is observed immediately after physical activity or exercise. Creatine Kinase is one of the trickiest parameters to measure and use for clinical decisions because elevated concentration is long and complex. As a result, it has been recommended that elevated CK level should be redefined as 1.5 times the upper limit of normal interval by the European Federation of Neurological Societies (Kyriakides et al., 2010). Elevated CK levels can also be caused by drugs and supplements which means in the measurement of CK levels, patients on medications should be noted. For instance, statin (lipids lowering medication) can cause myalgia, muscle weakness, and rhabdomyolysis with one out of 20 users experiencing about 2 to 10 times CK levels elevation of the upper limit of normal (Mancini et al., 2013). Hyper-CK-emia is the condition of elevated CK level. CK is an important diagnostic parameter for myocardial infarction. 2.5.2.3 Lactate Dehydrogenase (LDH). It is a glycolytic enzyme present in the liver, heart, kidney, and pancreas, in the human body. When the cells of human body become extracellular upon cell death, LDH becomes detectable in the cell cytoplasm. Consequently, the extracellular presence of LDH is usually related to cell necrosis and tissue damage. Hence, in various pathological 35 University of Ghana http://ugspace.ug.edu.gh conditions, it is common to observe elevated levels of serum LDH. This explains why generally the serum level of LDH may provide only non-specific measure of cellular destruction, yet very useful if specific LDH isoenzyme can be determined as it aids differential diagnosis of certain pathologic conditions. Interestingly, several studies allude to the fact that increased serum LDH may be a poor predictive factor in malignancies like renal cell cancer, nasopharyngeal carcinoma, pancreatic cancer, lymphoma, breast cancer, prostate cancer, colourectal cancer, lung cancer, and oesophageal squamous cell carcinoma (Mekenkamp et al., 2012; Motzer et al., 2013; Powles, Bascoul-Mollevi, Kramar, Lorch, & Beyer, 2013; Wei et al., 2016; Weide et al., 2012; Yu et al., 2017; Zhang et al., 2015). The serum LDH level test is convenient and cheap which makes it suitable for daily clinical use. 2.5.2.4 Amylase. Amylase is simply a group of enzymes that hydrolyze glucosidic bonds present in starch, namely, α-amylase, β-amylase, and glucoamylase (Taniguchi & Honnda, 2009). Amylase levels greater than three-fold normal are usually diagnostic of acute pancreatitis. On the other hand, lower levels of amylase are found in alcoholic pancreatitis. It is important to note that lipase, used as a biomarker alongside amylase is critical because about up to 19% of pancreatitic patients have normal serum amylase (Lott et al., 1992). Similarly, it is common to find elevated levels of amylase and lipase in gastrointestinal diseases, renal failure, pulmonary failure, non-malignant hepatobilary and subdural bleeding (Fody, 2010). 36 University of Ghana http://ugspace.ug.edu.gh 2.5.2.5 Uric Acid (UA). Kidney stones can be formed in the human body through deposition of sodium urate microcrystal or formation of ammonium acid urate. In addition, diabetes and gout are found to be associated with uric acid. Elevated levels of uric acid can result from poor excretion, fasting, rapid weight loss, uric acid dense diet intake, and possibly heredity. Chronic hyperuricemia is found to be risk factor for cardiovascular diseases, CKD, and hypertension (Kansui et al. 2018; Ameijeiras et al. 2016; Kallistratos et al. 2018; Borghi et al. 2015). 2.5.2.6 C-Reactive Protein (CRP). Serum CRP varies among healthy individuals with the mean concentration generally found to be < 10 mg/L in adults. The elevated levels of serum CRP concentration is always an alarming situation. Indeed, such situations are related to increased risk of colorectal cancer, lung cancer, and even death (Kaptoge et al., 2010; Zhou et al., 2014; Zhou, Liu, Wang, & Xi, 2012). Additionally, it is fundamentally one of the most important indicators for a high chance of inflammatory diseases (Thiele et al., 2015). 2.5.2.7 Liver Function Tests. One of the key organs in the human system is the liver. Some of its functions include synthesis of hormones, metabolism of drugs, storage of vitamins and iron, and excretion of waste products. Some of the biochemical markers for liver screening includes serum bilirubin, albumin, total protein, alanine aminotransferase (ALT), aspartate aminotransferase (AST), ratio of aminotransferases, alkaline phosphatase (ALP), and Gamma-Glutamyl transferase (GGT) (Lala & Minter, 2018). The presence of the liver enzymes mostly indicates the integrity of the cells in the liver. High levels of both ALP 37 University of Ghana http://ugspace.ug.edu.gh and GGT in liver assessment points towards the obstruction of bile ducts whiles the high levels of ALT and AST usually indicate liver cells are damaged (hepatitis). 2.5.2.8 Total Protein. Total protein measures the concentration of all proteins present in the serum. Usually body fluids such as urine and cerebral spinal fluid are where total protein is found. Largely, albumin and globulins make up total protein even though there are other proteins that are included in the measurement, they are considered to be largely insignificant. According to the Association for Clinical Biochemistry (2012), measurement of total protein in the serum gives an indication of the total immunoglobulin concentration as globulins are the outcome of (total protein- albumin). Similarly, it is occasionally captured as part of ‘liver function tests’ given that chronic liver diseases cause increases in immunoglobulins (Association for Clinical Biochemistry, 2012). Evidence indicates that total protein is less useful when measured singularly without measuring it simultaneously with albumin. 2.5.2.9 Albumin. One of the major serum protein is albumin. It is specifically produced by the liver, constituting 60% of total protein and globulins. In chronic liver diseases, there is a presence of decreased levels of albumin. More so, albumin is lost in the urine of nephritic syndrome patients. Albumin has a molecular weight of about 65000 D (Nagai et al., 2016) which is similar to that of veterinary species with minor variations (Throop, Kerl, & Cohn, 2004). Albumin in urine is usually measured to identify and monitor patients with kidney damage (Miller et al., 2008; Miller, Seegmiller, Lieske, Narva, & Bachmann, 2017). Similarly, serum albumin is measured for an investigation into liver diseases, protein- energy malnutrition, nephritic syndrome and others. 38 University of Ghana http://ugspace.ug.edu.gh More so, serum albumin’s measurement has more clinical importance as it indicates more than one pathological change. This is because there are a number of diseases that may cause a decrease in serum albumin concentration in the human body. As a result, it is used in different settings for the purposes of diagnoses, disease progression monitoring, and/or as an indicator for other further testings. 2.5.2.10 Bilirubin. The use of serum total bilirubin level in clinical decisions is critical, with some studies applying different dimensions of this analyte to common cardiovascular and kidney diseases studies (Kunutsor, Kieneker, Burgess, Bakker, & Dullaart, 2017; Tsai et al., 2015). The authors discovered log-linear inverse relationship between circulating total bilirubin level and cardiovascular disease risk, which is independent of well-known risk factors (Kunutsor et al., 2017). Elsewhere, it was found that higher serum total bilirubin concentrations within the reference range were related to lower total white blood cell counts, regardless of other classical cardiovascular risk factors (Tsai et al., 2015). 2.5.2.11 Aspartate aminotransferase (AST). In the human body, the organs, liver and kidney, as well as cardiac and skeletal muscles are places where large amounts of aspartate aminotransferase are found. AST was formerly called serum glutamic oxaloacetic transaminase (GOT). Raised levels of AST are often linked with malfunction of the liver (Bishop et al., 2013). In fact, AST elevations habitually dominate in patients with cirrhosis, even so, for liver diseases that characteristically have an increased ALT (Green & Flamm, 2002). Elevated levels of AST may not necessarily be exclusive to malfunction of the liver, because the enzyme may be from other source organs (Sohn et al., 2013). Hence, the level of AST complemented with 39 University of Ghana http://ugspace.ug.edu.gh alanine aminotransferase (ALT) in the serum are useful for the diagnosis of the liver and heart for injuries. 2.5.2.12 Alanine aminotransferase (ALT). It is very common to find greater concentrations of alanine aminotransferase (ALT) in the liver with the kidney, heart, muscle containing small amounts. When the liver is damaged, elevated levels of ALT can be observed, nonetheless, elevated levels of up to 300U/L is considered non-specific (Gowda et al., 2009). It is noteworthy, that well-defined elevated values of ALT over 500 U/L are commonly observed among persons with liver-related problems, such as ischemic liver injury (Gowda et al., 2009). ALT measurement is critical to the diagnosis and monitoring of various types of hepatitis which usually account for the elevated levels of aminotransferase. Lipid reducing and anti-diabetic drugs are known to raise ALT levels (Sorbi et al., 1999) which means close monitoring is necessary before any decision making in the laboratory. 2.5.2.13 Alkaline phosphatase (ALP). The ALP enzyme is derived in small amounts from the placenta, small intestines, kidneys and leukocytes, whereas the liver and bones provide the high amounts. Its apparent importance is revealed in the event of liver damage, as it is the first enzyme that is elevated in such condition. Of course, if it is a liver cell damage disease, it is the aminotransferase which increases significantly. ALP enzyme analysis is very useful for distinguishing between the disease conditions cholestatic and hepatocellular, (Liu et al., 2016) especially when combined with gammaglutamyl transferase. It is worth noting that, bone growth and healing of a broken bone may cause an increase in ALP even under normal conditions (Gowda et al., 2009). 40 University of Ghana http://ugspace.ug.edu.gh 2.5.2.14 Gamma-glutamyl Transferase (GGT). Liver and biliary related diseases are always associated with high levels of serum GGT (Gowda et al., 2009). Increased level of GGT is observed in about one out of every three patients with chronic hepatitis C infection (Giannini et al., 2001). Alcoholism is found to be associated with raised GGT levels. McCullough (2002) asserts that the level of GGT could rise two or three times above the UL of every one out of two non-alcoholic fatty liver diseased people. ALP alongside GGT are good correlates and useful for biliary disease, even though ALP is usually the first to be used (Williams et al., 2016). It is noteworthy also that barbiturates and phenytoin drugs and congestive heart failure could raise the level of GGT (Wang et al., 2016). 2.5.2.15 Urea. It is one of the main tests used to determine kidney health. It is expected that the kidney excretes the urea through urine. Therefore, in conditions where the urea is not effectively excreted by the kidney, there are raised levels of Blood Urea Nitrogen (BUN) which is an indication of abnormality in the kidneys. The normal human adult is expected to have between 6–20 mg/dL level of BUN. This range may vary according to different assays used. Raised BUN levels are associated with deaths in patients with acute and chronic heart failure (Cauthen et al., 2008; Filippatos et al., 2007). Elsewhere, BUN has shown to be a good and independent predictor of acute myocardial infarction (AMI) (Richter et al., 2018). Moreover, serum urea nitrogen level is relatively higher in males than females (Kamath et al., 2001). Also, a urea test is readily available, inexpensive and have standardized reference intervals. 41 University of Ghana http://ugspace.ug.edu.gh 2.5.2.16 Creatinine. It is one of the most affordable and most commonly measured analytes in clinical chemistry. Measurement of serum creatinine is limited because, it does not only vary with Glomerular Filtration Rate (GFR) but also with muscle mass as a result of being a product of muscle catabolism (Dalton, Kilpatrick, Nichols, & Maylor, 2010; Perrone, Madias, & Levey, 1992). It is also noteworthy, that serum creatinine level could be influenced by age, gender, ethnicity, protein intake, and muscle mass (Ceriotti et al., 2008; Cholongitas et al., 2007; Patel et al., 2013). Serum creatinine level is also an important measure of liver function (Kuster, Bargnoux, Pageaux, & Cristol, 2012) even though it is mainly an indicator of kidney function. 2.5.2.17 Glomerular Filtration Rate (GFR). Glomerular filtration rate is the best index of renal function (Soares et al., 2013). This is a key determinant of the stage of chronic kidney disease (CKD) because clinical decisions such as drug dosage, timing of renal replacement therapy, and application of potentially nephrotoxic contrast agents are based on GFR (Schaeffner, 2017). Admittedly, ageing has been an important factor and determinant of GFR, with a usually declining GFR as people age (Rule et al., 2010). Additionally, sex/gender have also been identified to be an important factor that influences GFR with females exhibiting either lower (Rule et al., 2010) or higher values (Poggio et al., 2009) compared with males. Ethnicity/race have also been linked with the level of GFR even though it still remains debatable (Soares et al., 2013). 2.5.2.18 Total Carbon dioxide (TCO2). Measurement of bicarbonate and total carbon dioxide are usually used interchangeably (Mohd, Sthaneshwar, Junaidah, & Yap, 2010) with none as an estimator for the other. 42 University of Ghana http://ugspace.ug.edu.gh Abnormalities in serum total carbon dioxide is usually the first point of recognition of acid-base disorder among patients. The reference interval for total CO2 in the serum is usually 23 to 30 mEq/L (Centor, 1990). Metabolic acidosis or compensation for respiratory alkalosis is often associated with low levels of total CO2. A total CO2 test is a member of the metabolic panel of test that measure electrolytes and blood gases. 2.5.2.19 Anion Gap. The serum anion gap is defined as the sum of serum chloride and bicarbonate concentrations subtracted from the serum sodium concentration [Na+ - (Cl− + HCO3 −)]. This is calculated from electrolytes from the chemical laboratory (Kraut & Madias, 2007). It is recommended that potassium is included in the formula where concentration is abnormally high or low (Kraut & Nagami, 2013). The serum anion gap is used to detect paraproteins, identify errors in measurement of electrolytes and assessment of patients with suspected acid-base disorders (Kraut & Nagami, 2013). Raised values of anion gap may be attributed to laboratory errors, yet most commonly it reflects metabolic acidosis, metabolic alkalosis, hyperphosphatemia, or paraproteinemia (Kraut & Madias, 2007). Metabolic acidosis is related to raised levels of serum anion gap which is usually caused by overproduction of acid or decreased acid excretion (Emmett, 2006). 2.5.2.20 Diabetes test. Diabetes tests are primarily done using two parameters – glucose and haemoglobin A1c (HbA1c). The glucose parameter is easy to use and cheap, as a result, it is usually the first and preferred choice for many clinicians and researchers. In cases of mass screening, it is applied as a first-step either as fasting or random. Moreover, it is used as a measure of risk for the development of diabetic complications, and as a measure of the quality of diabetes 43 University of Ghana http://ugspace.ug.edu.gh care (Herman & Cohen, 2012). The normal serum glucose level (fasting) is < 5.6 mmol/L. An HbA1C level of 6.5% or higher is also suggestive of diabetes. Meanwhile, any value < 5.7% is considered normal. 2.5.2.21 Lipid Studies. Lipid studies are a group of indicators that describes varying levels of lipids in the blood. Low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides are the most commonly reported. Two major serum lipids, cholesterol and triglycerides are transported in the blood by lipoprotein particles. It is usually called “good cholesterol” while LDL is considered “bad cholesterol”. The normal reference range is < 100 mg/dL for LDL. Thus, while high HDL is positive, the high level of LDL is a serious health threat. Even though total cholesterol is an important parameter, using it independently can be misleading because of the influence of HDL and LDL. The normal reference range is < 200 mg/dL. Triglycerides concentration in the blood should be < 150 mg/dL. Generally, an abnormal lipid profile has been found to be a good predictor of coronary heart diseases (Islam, Choudhury, Mainuddin, & Wahiduzzaman, 2014; Musa et al., 2013; Sherpa et al., 2011). Reference values for lipid parameters have shown variations with respect to gender and ethnicity (Furusyo et al., 2013; Jalali, Honomaror, Rekabi, & Latifi, 2013). 2.5.2.22 Immunoglobulin A, M, G (IgA, IgM, IgG). Immunoglobulins fundamentally play the role of fighting viruses, bacteria and toxins. The human body produces different immunoglobulins that fight different antigens. Interestingly, this characteristic can result in a condition called autoimmune disease. Thus, 44 University of Ghana http://ugspace.ug.edu.gh a condition whereby the body mistakenly produces antibodies against itself by treating healthy organs as invaders. Immunoglobulins are generally in the forms of IgA, IgG and IgM. The first antibody to be produced is the IgM immediately after initial antigen encounter. In addition, it assists with the later development of the immune response. IgA is found in mucous membranes, mainly in the respiratory and digestive tracts. IgA deficiency seems to play a part in asthma and allergies. IgG is the most abundant antibody (75%) fighting against infections. It plays role in fighing parasitic infections. Generally, low levels of immunoglobulins indicate a weak immune system which increases the body’s chances of getting infections. In contrast, high serum level IgE could be an allergy or even an overactive immune system. 2.5.2.23 Iron Studies. Serum ferritin, transferrin and iron are the main parameters for iron studies. An elevated level of transferrin means iron-deficiency anaemia (IDA), whereas a lower level means the presence of a liver disease or hemolytic anaemia. Both serum ferritin and transferrin work closely with serum iron. The iron studies all together are important for diagnosing anaemia both IDA and non-IDA, monitoring iron therapy response and aiding in the diagnosis of iron overload. It is also noteworthy that sometimes elevated levels may be observed in the presence of inflammation. 2.5.2.24 Vitamins Studies (B12, D 25-OH, folate). In the case of a direct measure of serum vitamin B12, deficiency results in haematologic, neurologic, psychiatric, and other chronic illnesses (Headstrom, Rulyak, & Lee, 2008; Selhub, Morris, Jacques, & Rosenberg, 2009). The laboratory evaluation of possible deficiency of vitamin B12 is often driven by the presence of macrocytic anaemia (Carmel, 45 University of Ghana http://ugspace.ug.edu.gh 2008). It is also important to note that vitamins, folate, and B12 serve as coenzymes in one-carbon metabolism (Selhub et al., 2009). Vitamin D is critical to bone health, serum calcium and phosphate levels. Thus, it has been found that increased risk of fractures, osteomalacia in adults, cancer, autoimmune disease, infectious disease, and heart disease are associated with Vitamin D (25-OH) deficiency (Basit, 2013). Ethnicity or race is important when measuring total serum Vitamin D because skin pigmentation affects the synthesization process. Moreover, physical activity, body mass index (BMI) and gender also affect vitamin D (25-OH) concentration (Lanteri, Lombardi, Colombini, & Banfi, 2013; Vuistiner et al., 2015). 2.5.2.25 Endocrine Studies. Prolactin has a primary function of initiating breastfeeding and maintaining lactation (Lahiri, Baruah, Ghosh, & Sengupta, 2014). The elevated prolactin concentration among pregnant females is for obvious reasons. Prolactin level is an important determinant of hypothalamic-pituitary disorders. An elevated level of prolactin may be caused by estrogen, thyrotropin-releasing hormone (TRH), renal disease, hypothyroidism, stress, exercise, hypoglycemia and drugs like Chlorpromazine and reserpine (Lahiri et al., 2014). Luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are both important for the reproductive system. LH acts synergistically with FSH to stimulate follicular growth and ovulation. The amount of FSH changes during a woman's menstrual cycle and is maximum just before she ovulates. FSH helps control the production of sperm in males with a stable level among men after puberty. Among female adults the normal range is 4.5 to 21.5 IU/L while it is 1.5 to 12.4 IU/L among adult males. In recent times, FSH is 46 University of Ghana http://ugspace.ug.edu.gh habitually used in assisted reproductive technology (ART). Estrogen is an important hormone in females. Among females who are not pregnant, the estradiol (E2) is the main estrogen. Serum estradiol (E2) contributes to oocytes/follicular maturation and uterus preparation for implantation (Anifandis et al., 2005). It is considered to be in normal range when it is 15 – 350 pg/mL for menstruating females and <10 pg/mL for females who have reached menopause. Progesterone remains an important part of the reproductive system. Interestingly, the concentration of progesterone varies widely across the menstrual cycle through ovulation and pregnancy. LH is again necessary for the effective functioning of the progesterone, testosterone and estrogen. 2.5.2.26 Thyroid Studies. Thyroid is among the largest endocrine gland in human, with two important hormones that interfere with various aspects of metabolism - Tetraiodothyronine or thyroxine (T4) and triiodothyronine (T3). Thyroid hormones dysfunctions due to a variety of disorders related with the thyroid gland are well established, which comprised of many abnormalities in the thyroid gland itself leading into hypothyroidism and hyperthyroidism (Mansourian, 2010).. The normal reference range of free T3 for adults is 260-480 pg/dL and free T4 is 0.7-1.9 ng/dl. There is also evidence of the strong correlation between T3 and T4 and age (Meng et al., 2015; Qiu et al., 2018). Thus, other authors suggest that these data should be interpreted based on age and gender. Elsewhere, thyroid hormones have been associated with kidney structure and function, including acute kidney injury (Basu & Mohapatra, 2012; Joo, Kim, Go, & Song, 2018). 47 University of Ghana http://ugspace.ug.edu.gh 2.5.2.27 Thyroid-Stimulating Hormone (TSH). There are two main forms of thyroid: dysfunction-hypo and hyperthyroidism. Hypothyroidism is defined as “elevated TSH with normal free thyroxine (FT4), whereas hyperthyroidism is decreased TSH and normal FT4” (Biondi, 2012; Cooper & Biondi, 2012). TSH is a less sensitive marker of thyroid insufficiency in older people, therefore, clinicians should be extra careful not to ignore clinical hypothyroidism when TSH concentration is elevated (Laurberg et al., 2011). Moreover, many factors such as ethnicity, age, type of assays, body mass index (BMI), smoking and iodine status can affect the results of thyroid function tests (Brown et al., 2016; Chaker et al., 2016; Sun, Shan, & Teng, 2014; Yan et al., 2011). Atzmon et al. (2009) point out that the upper limit and median of TSH positively correlates with age. Therefore, using one reference interval values for both older and young people is likely to lead to over-diagnosis and potential mistreatment. Interestingly, the TSH reference interval distribution is seen as not normal with median values usually between 1-1.5 mU/L (Biondi, 2013; Fontes et al., 2013; Vadiveloo, Donnan, Murphy, & Leese, 2013). The upper limit, however, is usually about 4.2-4.5mU/L which is assay-dependent (Lewandowski, 2015). 48 University of Ghana http://ugspace.ug.edu.gh 2.6 Empirical Studies Table 2.1:Empirical Studies of Reference Interval Author(s) Country of study Analytes Target Statistical method Findings Al-Mawali et al. (2013) Oman Haematological Gender, age, Non-parametric Significant difference between males and females and ABO blood for RBC, Hb, Hct. group Other CBC parameters showed no significant differences between genders. Mulu et al. (2017) Ethiopia Haematological Gender Non-parametric Hb = 15.69-17.84g/d; Plt Females = 120-379 x 106 cells/mm3 males = 106-352 x106 cells/mm3. WBC = 3.5-11.5 x103/mm3; MCV = 89.5-107.5 fl, MCH = 28-34 pg and MCHC = 30-33.2g/dl, Siraj et al. (2018) Eritrea Haematological Gender Non-parametric significant difference between males and females in the RI for erythrocyte count, Hb, Hct, MCV, MCH, MCHC and differential WBC. All haematological analytes were higher in males than in females except for platelet count (Segolodi et al., 2014) Botswana Haematological and Gender Non-parametric AST = 13.0–42.0; ALT = 7.0–46.0; biochemistry Creatinine = 0.5–1.1; BUN = 5.0–21.0 Bilirubin (total) = 0.2–1.8; Bilirubin (direct)= 0.1–0.4; Amylase = 47.0–176.0; Phosphate (inorganic) = 2.2–4.3; Chloride = 98.0–107.0; CO2 = 19.9–29.1; Potassium = 3.6–5.2; Sodium (indirect) = 135.0– 143.0; Hb = 10.4–17.6; Hct = 32.9–51.7 Significant difference between males and females in the RI for all analytes Mekonnen et al. (2017) Ethiopia Biochemistry Gender Non-parametric Males had high (P<0.05) mean and 2.5th-97.5th Age percentile ranges of ALT, AST, ALP, creatinine and direct bilirubin. The RI of amylase, LDH, total protein and total bilirubin were not significantly different between the two sex groups (P>0.05). (Du et al., 2016) Mongolia haematological Gender Significant sex differences for all parameters Age except for Lym% and Bas%. Bakrim et al. (2018) Morocco Haematological Gender Non-parametric significant sex difference for all the haematological parameters except for Lym# 49 University of Ghana http://ugspace.ug.edu.gh Sundaram et al. (2008) India Haematological and Ethnicity Non-parametric Total bilirubin, ALT, albumin, creatinine, total biochemistry protein, lipid profile, CK, uric acid and lactate and haematological (MCH, MCHC, LYMP levels) parameters presented higher upper limits. (Lawrie et al., 2009) South Africa Haematological Gender Non-parametric sex-specific differences in RBC and platelet Ethnicity parameters. Ethnic-specific differences were found for some RBC parameters in the black female cohort. In addition, black males and females both generally had lower neutrophil and higher lymphocyte counts than a combined Asian/Caucasian/coloured ethnic group. (Sairam et al., 2014) India Haematological and Gender Non-parametric RI for Hb (Males: 12.3-17 g/dL; Females: 9.9- biochemical 14.3 g/dL), platelet count (Males: 1.3-3.8; Females: 1.3-4.2 Lakhs/µL), erythrocyte sedimentation rate (Males: 2-22; Females: 4- 55 mm/h), serum uric acid in males: 3.5- 8.2 mg/dL, gamma glutamyl transferase (Males: 13-61 U/L), fasting blood glucose (Males: 78- 110 mg/dL), total cholesterol (Males: 115- 254 mg/dL), low density lipoprotein (Males: 60- 176 mg/dL) and triglycerides (Males: 55- 267 mg/dL, Females: 52-207 mg/dL) Samaneka et al. (2016) Zimbabwe Haematological and Gender Non-parametric Males had significantly higher RBC, Hb, Hct, biochemical MCH compared to females. Females had higher WBC, PLT, neutrophils, and Lym compared to males. No sex-related differences in eosinophils, monocytes, and absolute basophil count. Males had significantly higher levels of urea, sodium, potassium, calcium, creatinine, amylase, total protein, albumin and liver enzymes levels compared to females. Females had higher cholesterol and lipase compared with males.. Chisale et al. (2015) Malawi haematological Gender No statistically significant differences in Age haematological results among different ethnic, Ethnicity age, and body mass index groups. BMI 50 University of Ghana http://ugspace.ug.edu.gh Kibaya et al. (2008) Kenya Haematological and Region Parametric and Significant sex differences for RBC and WBC. biochemical Gender non-parametric Kenyan subjects had lower median Hb (9.5 g/dL; range 6.7-11.1) and neutrophil counts (1850 cells/microl; range 914-4715) compared to North Americans. Kenyan clinical chemistry reference ranges were comparable to those from the USA, with the exception of the upper limits for bilirubin and blood urea nitrogen, which were 2.3-fold higher and 1.5-fold lower, respectively (Dosoo et al. 2012) Ghana Haematological and Gender Non-parametric Hb (males = 113–164; females = 88–144) WBC biochemical = 3.4–9.2; PLT (males = 88–352; females = 89– 403) ALT (males = 8–54; females = 6–51) creatinine (male= 56–119; females = 53–106). Kone et al. (2017) Mali Haematological Gender Non-parametric WBC = 3237.5-11900; RBC = 3.56-6.17; Hb =12.2-17.38; Plt =145.4-614.4; Lym# = 1200- 3800; Neu# =1040-6220; Mon# = 100-660; eos# = 0-1026; Significant sex differences in RBC, Hb level and MPV. (Asare, Ezekiel, Nkumpoi, Ghana Biochemical Non-parametric Thus, half of the manufacturers RI (MRI) & Amoah, 2012) represented  25 percentile for FF, 2HPP, LDL- C, ALB and ALP. The MRI for Urea was <25th - >97.5th. (Koram et al., 2007) Ghana Haematological and Gender Non-parametric Hb (males = 14.2 g/dl; females = 12.0 g/dl), ALT biochemical (males=26.1 U/L; female = 19.6 U/L) and Creatinine, (males = 108 mmol/L; females = 93mmol/L). 51 University of Ghana http://ugspace.ug.edu.gh Different categories of empirical studies covering reference intervals were analyzed which span various countries and regions. The review covers reference interval studies mainly across Africa which analyzed haematological, biochemical and immunological parameters. Table 2.1 presents various studies that established RIs for local populations. A total of 18 previous studies which were largely African based were reviewed. The studies cover a period between 2008 and 2018. Two of the studies were in Kenya (Kibaya et al. 2008; Odhiambo et al. 2015), two in Ethiopia (Mulu et al. 2017; Mekonnen et al. 2017), two in India (Sairam et al. 2014; Sundaram et al. 2008) and three in Ghana (Dosoo et al. 2012; Asare et al. 2012; Koram et al. 2007). Elsewhere in Morocco (Bakrim et al., 2018), Oman (Al-Mawali et al. 2013), Zimbabwe (Samaneka et al. 2016), Mali (Kone et al. 2017) , Malawi (Chisale et al. 2015), Mongolia (Du et al., 2016), Botswana (Segolodi et al., 2014), Eritrea (Siraj et al. 2018) and South Africa (Lawrie et al., 2009) one study each were reviewed. In all these studies, haematological and biochemical analytes were the focus of analysis, with some studies restricted to establishing RI for only haematological, biochemical or both analytes. The studies that were reviewed included population-based approach where study participants volunteered (e.g. Kone et al. 2017; Kibaya et al. 2008; Dosoo et al. 2012). While other studies were conducted using blood samples from established blood banks where volunteers had donated their blood for other purposes (e.g. Chisale et al. 2015). Elsewhere, Kone et al. (2017) set out to establish haematological RIs for healthy Malian adults. The study established RIs for the following haematological parameters - (WBC) count of 5200 cells/μL [3237.5–11900], RBC count of 4.94 10^6 [3.56–6.17], Hb 52 University of Ghana http://ugspace.ug.edu.gh of 14.2 g/dL [12.2–17.38], platelet count (Plt) of 275 10^3/μL [145.4–614.4], lymphocytes 2050/μL [1200–3800], neutrophils 2200/μL [1040–6220]; monocytes 200/μL [100– 660]; eosinophils 131/μL [0–1026]; CD4 902 cells/μL [444–1669] and CD8 485 cells/μL [0– 1272]. Additionally, it was revealed that significant differences among males and females existed for RBC, Hb and MPV. It was further revealed that Hb levels differed from those reported in East and Southern Africa (Kone et al., 2017). Kibaya et al. (2008) studied haematological and clinical chemistry analytes for adults in Kericho, a rural community in Kenya, it was established among the 1293 HIV seronegative participants that people from Kenya had lower median haemoglobin levels (9.5 g/dL; range 6.7–11.1) and neutrophil counts (1850 cells/ml; range 914–4715) in comparison to North Americans. Yet, clinical chemistry ranges were similar to those from USA, except bilirubin and blood urea nitrogen. Interestingly, significant sex variations were observed. In Malawi, Chisale et al. (2015) studied the haematological distribution of healthy individuals in Blantyre. In comparison with existing manufacturers’ reference intervals in use in Malawi, the study found that study participants (19-35 years) had results below the lower limit of the manufacturers’ RIs. Thus, 35% for haemoglobin, 15.2% for neutrophils, about 88% for basophils, and approximately 24% for eosinophils (Chisale et al. 2015). Platelets and monocytes were also above the upper limits of the manufacturer’s RIs with significant gender differences observed in leucocyte and basophil counts. Also, haematological RI study conducted by Bakrim et al. (2018) among Moroccan adult population using a non-parametric approach found significant variations between males and females for all the haematological parameters measured except lymphocytes counts. Similarly, obvious variations were observed with resuts compared to other RIs reported in Arabic, Caucasian and African populations. 53 University of Ghana http://ugspace.ug.edu.gh The two studies that reported in India by Sairam et al. (2014) and Sundaram et al. (2008) both studied haematological and clinical chemistry parameters. For Sairam et al. (2014), a sample of 10,665 healthy individuals were recruited in multiple centres in India. Using a non-parametric approach the 95% RIs were calculated for haemoglobin (Males: 12.3–17 g/dL; Females: 9.9–14.3 g/dL), platelet count (Males:1.3–3.8; Females: 1.3–4.2 Lakhs/lL), erythrocyte sedimentation rate (Males: 2–22; Females:4–55 mm/h), serum uric acid in males: 3.5–8.2 mg/dL, GGT (Males: 13–61 U/L), fasting blood glucose (Males: 78–110 mg/dL), total cholesterol (Males: 115–254 mg/dL), LDL (Males:60–176 mg/dL) and triglycerides (Males: 55–267 mg/dL, Females: 52–207 mg/dL). The authors concluded that their findings varied significantly from RIs currently used in India. Also, in 2008, Sundaram et al. established RIs for biochemical and haematological parameters using a sample of 213 subjects recruited in Chennai, southern India. Biochemical - total bilirubin, ALT, albumin, creatinine, total protein, lipid profile, creatine phosphokinase, uric acid and lactate) and haematological (MCH, MCHC and lymphocyte levels) parameters presented higher upper limits. In addition, the upper limits of WBC, platelet count, Hct, RBC and Hb level were low in comparison to current RIs. Furthermore, it was found that RIs ethnic- related variations existed for some haematological and biochemical parameters. Mekonnen et al. (2017) sampled 446 participants from the Gojjam zone in Northwest Ethiopia. Clinical chemistry parameters were determined among this population using a non-parametric technique. The results indicated that income influenced ALT, total protein and creatinine, with high-income people having high values than low-income people. In addition, it was found that there were no age-related differences for AST, amylase, total protein, total bilirubin, ALP and direct bilirubin except ALT. Likewise, Lawrie et al. (2009) studied a South African population, in an attempt to establish reference intervals 54 University of Ghana http://ugspace.ug.edu.gh for haematological parameters. Using a cross-sectional design 719 HIV-negative participants were recruited from the Gauteng region. Targeting gender and ethnic comparison, the non-parametric analysis revealed that, there were gender-specific differences in RBC and Platelet parameters. In addition, the RIs generated for this population (black) was different from those in Asian, Caucasian and coloured ethnic group. Therefore, confirming the need for locally derived RIs for use in clinical use and clinical trial studies. Samaneka et al. (2016) found significant gender-specific differences in the RIs of Zimbabwean adult (18-55 years) population for haematological and biochemical parameters. The authors found RBC, Hb, Hct, MCH were significantly higher in males than females. On the other hand, WBC, platelets, absolute neutrophil counts, and absolute lymphocyte counts were higher in females compared to males. Furthermore, it was found that males had significantly higher levels of albumin, total protein, potassium, urea, creatinine, calcium, amylase, sodium and liver enzymes levels relative to females. Yet, females had higher cholesterol and lipase compared with males (Samaneka et al. 2016). Finally, the authors confirmed that WBC, neutrophils, cholesterol and creatinine kinase were values that are significantly different from the current RIs in use. Two of the previous studies that were conducted in Ghana had two distinctive features. Firstly, for Asare et al. (2012), the authors aimed at establishing RIs for haematological and biochemical analytes among Ghanaian healthy adults using the non-parametric technique. Using a sample of 6300 adults aged 25-65 years, the study established that mean values of most of the analytes determined represented the 25th – 75th and not the 95th or 97.5th percentile (Asare et al., 2012). This points to the fact that there were significant 55 University of Ghana http://ugspace.ug.edu.gh differences in the analytes established by this study compared with the existing manufacturers’ RIs that come with the package kits. Similarly, Dosoo et al. (2012) studied the people in Kintampo Municipality in the Brong Ahafo region where a total of 691 volunteers were recruited. The authors studied 16 haematological and 22 biochemical analytes using the non-parametric technique. The study revealed that more than half (53%) of study participants would have been considered unhealthy using manufacturers’ package insert for the haematological parameters. And up to one-fourth (25%) of the population would have the same fate for the biochemical parameters. Prior to these two studies in Ghana, Koram et al. (2007) studied the RIs for both haematology and biochemistry parameters for Ghanaians. Using a sample of 400 volunteers, the authors established the RIs using the non-parametric technique. It was found that the RIs for the Ghanaian population was significantly different from other existing RIs that were in use by clinicians. Thus, it was found that haemoglobulin levels from this study were lower and liver function parameters higher. The authors, therefore, concluded that further studies be done to establish RIs specific to the local population, as they attributed some of the differences to genetic and environmental factors. These findings from the previous Ghanaian studies are important foundations for further studies into RIs establishment for Ghanaian population - an important backdrop that necessitated this current study of establishing RIs for the healthy adult population in Ghana, taking into consideration ethnicity, gender, age and nutrition. In addition, the scope of the current study embraces haematological, biochemical and immunological parameters by adopting both parametric and non-parametric statistical techniques. 56 University of Ghana http://ugspace.ug.edu.gh Table 2.2:Multicenter Studies on Reference Intervals Author(s) Reference country Analytes Target Statistical method Findings (ies) (Ceriotti et al., 2010) Italy Biochemical Gender Non-parametric Regional differences were observed for RI of AST. Inter- China Regions regional differences were not statistically significant for ALT, Turkey but partitioning was required due to significant sex differences. Nordic countries RIs for ALT (females = 8–41 U/; males 9–59 U/L), GGT (females = 6–40 U/L; males = 12–68 U/L). (Ozarda et al., 2014) Turkey Biochemical Gender Parametric No significant differences of reference values among seven Age LAVE regions for the 25 analytes. Significant sex-related and age- Regions related differences were observed for 10 and seven analytes, respectively. (Xia et al., 2016) China Biochemical Gender Parametric significant sex-related and age-related differences were Age LAVE observed in 12 and 12 analytes, respectively Regions (Ozarda et al., 2017) Turkey Haematological Regions Parametric Regional differences in RIs were observed among the 7 regions LAVE of Turkey (Yamamoto et al., 2013) Japan Haematological Gender Parametric No significant differences in between-region among all Biochemical Age LAVE analytes. Significant between-sex and -age variations in 14 and Regions 15 analytes, respectively. (Ichihara, Ceriotti, Tam, et South Korea Biochemical Gender Parametric SDRs for sex and age were significant for 19 and 16 analytes, al., 2013) Japan Hormones Age respectively. Regional difference was significant for 11 Malaysia Inflammatory Regions analytes, including HDL-cholesterol and inflammatory Vietnam markers. Regionality differences in analytes were not observed Indonesia in the data for only Japan China (Karita et al., 2009) Kenya Haematological,bi Region Non-parametric While some significant gender and regional differences were Rwanda ochemical, and Gender observed, creating consensus African study intervals from the Uganda immunological complete data was possible for 18 of the 25 analytes. Zambia Compared to reference intervals from the U.S., lower HCT and HGB levels, particularly among women, lower WBC and neutrophil counts, and lower amylase were observed. Both genders had elevated eosinophil counts, immunoglobulin G, total and direct bilirubin, LDH and creatine phosphokinase, the latter being more pronounced among women. 57 University of Ghana http://ugspace.ug.edu.gh Table 2.2 presents a review of selected multicenter reference interval studies. This particular review points out that several of the multicenter reference interval studies that have been conducted were done in Asia. In fact, out of the seven studies that were reviewed it was found that three studies were explicitly done in Asian countries (Ichihara, Ceriotti, Tam, et al., 2013; Xia et al., 2016; Yamamoto et al., 2013) while one included an Asian country (Ceriotti et al., 2010). Interestingly, two were also conducted in Turkey (Ozarda et al., 2014, 2017) with another multicenter study involving Turkey (Ceriotti et al., 2010). One study was also conducted in Africa (Karita et al., 2009). All the studies that were reviewed were conducted between the year 2009 and 2017. In addition, some of the studies were nationwide multicenter studies (Ozarda et al., 2014, 2017; Xia et al., 2016; Yamamoto et al., 2013) while the rest were multi-country multicenter studies (Ceriotti et al., 2010; Ichihara, Ceriotti, Tam, et al., 2013; Karita et al., 2009). All the studies except two (Ceriotti et al., 2010; Karita et al., 2009) used the parametric statistical approach in the generation of the reference intervals. These were largely complemented by the application of LAVE. Ceriotti et al. (2010), applied the non- parametric method in the RI generation which found statistically significant differences for AST in the four regions, even though the differences are considered to be clinically negligible. On the other hand, ALT, however, did not have inter-regional differences albeit significant gender differences. Yet, for Ichihara et al's. (2013) study among Southeast Asian countries it found inter-regional significant differences in many analytes that covered hormone, biochemical, inflammatory, etc. Even though when the data was limited to only Japan, no significant differences were found between regions within Japan among all the analytes. Noteworthy, however, is the differences in the statistical techniques 58 University of Ghana http://ugspace.ug.edu.gh adopted by the two studies. While Ceriotti et al. (2010) used non-parametric, Ichihara et al. (2013) used the parametric technique with LAVE. The intriguing outcome of these multicenter studies is that every one of them that explored the relationship between reference values and variation between male and female had similar results. Thus, even in studies where no between-regions were found, there were clear sex- variations in RIs. For example, as indicated in Table 2, there were no inter- regional differences in RI of the analytes in Turkey and Japan, yet sex-related variations were found (Ozarda et al., 2014; Yamamoto et al., 2013). Furthermore, it was revealed that age-related variations were common amongst these multicenter studies even in studies which did not record significant inter-regional differences. Moreso, out of the multi-country studies, it is fascinating that the one conducted among African countries provides an important contribution in their results. Thus, Karita et al. (2009) suggest that “while some significant gender and regional differences were observed, creating consensus African study intervals from the complete data was possible for 18 of the 25 analytes”. This provides a foundation for further studies in the quest to have regional-specific RI that are most appropriate clinically and economically. In all the studies, it was found that the authors adopted the CLSI-IFCC guidelines which provide the standard for multicenter reference interval studies. Hence, adequate sample sizes, use of parametric or non-parametric statistical techniques, strict eligibility criteria, specimen handling, etc. were adhered to. 59 University of Ghana http://ugspace.ug.edu.gh 2.7 Conclusion One overarching observation runs through the studies that were reviewed. Almost all the studies conducted used the non-parametric technique in the derivation of the RIs. Similarly, haematological analytes dominated the studies that were conducted particularly in Africa. Even in the studies that targeted biochemical analytes, they were usually done alongside haematological analytes. Moreover, the review showed that ethnicity/race, gender and age were important factors that influenced the RIs for specific parameters. Thus, many of the studies found that significant differences existed for certain haematological and biochemical parameters. This was supported by other findings that revealed that many of the existing RIs used currently by clinicians and for screening for clinical trial studies in local Africa and Asia setting were not appropriate due to significant differences; thus calling for locally derived RIs studies. 60 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE METHODS AND MATERIALS 3.0 Introduction This chapter covers the methods and research procedures used in collecting and analysing data generated during the study. It deals with the study design, profile of the study area, sources of data, sample size, research instruments, sampling techniques, data collection, laboratory analysis and procedures for the data analysis. Also, the details of the inclusion and exclusion criteria, standard operating procedures involved in reference interval derivation, as well as all other protocols that guided this study are covered. 3.1 Study Design The study adopted a cross-sectional design that employed simple random sampling technique to recruit healthy individuals, aged 18 – 95 years (primary target between 18 to 60+ years). Aside the administration survey, the study also involved laboratory analysis of the participants’ blood samples to establish a reference interval for the Ghanaian adults. Healthy individuals, aged 18 to 60+ years, who were willing to volunteer as participants were recruited from residents in Accra and Tamale. The aim of the research was advertised via radio stations in both study sites, posters, visits to institutions, churches, mosques, and other social centres to create public awareness. A simple random sampling technique was used to recruit potential study participants into the study based on the eligibility criteria (inclusion/exclusion). The study participants were issued with a structured questionnaire, which captured their socio-demographic characteristics, lifestyle characteristics, nutrition pattern as well as clinical information. 61 University of Ghana http://ugspace.ug.edu.gh Participants’ blood pressure and anthropometric measurements were taken as well. In addition, fasting blood samples were also taken from the volunteers. 3.2 Study Area/ Study Sites Ghana is a tropical country in Africa with a surface area of 238,533 km2. It is situated at the edge of the Atlantic Ocean in the Western Africa shoreline. Ghana consists of sixteen regions, including many islands now, but during the time of study it was ten regions. The current population is about 29.46 million (Ghana Statistical Service, 2018). Out of the ten regions, two were selected as the study sites thus the Greater Accra region (Southern part) and Northern region (Northern part) of Ghana. These two regions were purposively selected to find out the effect of locality and ethnicity on reference intervals of the various analytes among Ghanaian adults. The study locations selected for the study were Tamale the regional capital of the Northern region and Accra, the capital of the Greater Accra region. The two study sites are of Metropolitan status. Tamale is Ghana’s fourth largest city with an estimated population of 360,579 according to the Ghana Statistical Services (GSS, 2010). The city is located at 600 km north of Accra with a surface area of 750 km2. The specific recruiting sites for the study within the Tamale Metropolis included Tamale Central, Nyohini Mosque, Changli, Tamale Zongo, Tamale Technical University, Ola Cathedral and Tamale Teaching Hospital. Accra is also a Metropolis in the Greater Accra region and it is the most urbanized and densely populated area of the ten regions with an estimated population of over 3 million (GSS, 2010). Accra is the capital city of Ghana located in the coastal savanna ecological 62 University of Ghana http://ugspace.ug.edu.gh zone in the south-eastern part of Ghana. The specific recruiting sites within the Accra metropolis were Police Service Headquarters, Prison Service Headquarters, School of Biochemical and Allied Health Sciences - Korle-Bu, Corporate Institutions (eg banks), and social gathering places (churches and mosques). 3.3 Study Population The study population involved Ghanaian healthy individuals, aged 18 – 95 years (primary target between 18 to 60+ years) who met all the inclusion criteria of the study and resident in the selected study areas. 3.4 Sample Size The sample size was chosen following recommended protocol published by the IFCC, Committee on Reference Intervals and Decision Limits (C-RIDL), for multicenter studies (Ozarda et al., 2013) The guidelines recommend a minimum number of 500 reference individuals as an appropriate sample size in establishing RIs in all multicentre studies. A target sample size from each country is set at 500 or more so that: ● Between-country differences can be clearly detected ● Country-specific RIs are also obtained in a reproducible manner ● When there is an interest to explore within-country regional variations, it is recommended to obtain at least 120 samples from each local area (Ozarda et al., 2013). In Ghana, since this study was part of a multicentre study the sample size was 501, with equal distribution of participant’s age and gender profile. 515 participants were recruited for the study. After the laboratory analysis, 14 participants had high levels of some analytes 63 University of Ghana http://ugspace.ug.edu.gh such as glucose, lipids, iron deficiency anaemia, and thyrotoxicosis so they were excluded from the results, hence 501 participants were recruited in the study. 3.5 Sampling Procedure 3.5.1 Inclusion criteria. The participants should: ● Be 18 – 60+ years of age (with even distribution of gender and age, without much consideration about the evenness of the age distribution above age 60 years). ● Be feeling subjectively well. ● Ideally not be taking any medication. However, subjects who take a small number of medications (vitamins) occasionally can be recruited. ● Be a Ghanaian, who resides in Accra or Tamale. 3.5.2 Exclusion criteria. ● Participants with BMI ≥ 35 (Body mass index) ● Participants with high consumption of ethanol ≥ 70g/day ● Participants with high intake of tobacco >20 cigarettes/day ● Participants taking regular medication for chronic diseases (Diabetes mellitus, hypertension, hyperlipidaemia allergic disorders, depression, etc.) ● Participants who had recently recovered from acute illness, injury, or surgery requiring hospitalisation (≤14 days) ● Known carrier state of HBV, HCV, or HIV ● Pregnancy or within one year after childbirth ● Incomplete data (questionnaire/consent form not completed). 64 University of Ghana http://ugspace.ug.edu.gh 3.5.3 Selection of Study Participants. The study population involved healthy individuals. This study used a multi-stage sampling technique. First, the two study locations were randomly selected after the ten (10) regions of Ghana were categorized into northern and southern zones. In this, the Northern region and Greater Accra regions were selected as the main study areas. Subsequently, the participants were randomly selected using simple random sampling (SRS) technique to recruit eligible participants from Nyohini Mosque, Changli, Tamale Zongo, Tamale Technical University, Ola Cathedral and Tamale Teaching Hospital, Accra Police Service Headquarters, Prison Service Headquarters, School of Biochemical and Allied Health Sciences - Korle-Bu, and Banks. A total of 515 study participants aged 18 – 60+ years with fair gender and age distribution from the Northern and Greater Accra regions in Ghana were recruited. Structured questionnaires were administered to volunteers to capture their demographics, lifestyles, nutrition pattern as well as their clinical information. This was necessary to identify ineligible volunteers that were to be excluded from the study. The selection of eligible participants was solely based on inclusion and exclusion criteria, which were in accordance with the IFCC/C-RIDL protocol (Ozarda et al., 2013). Upon completion of the questionnaire and consent form, the eligible participants in Accra were then scheduled for an appointment at Medlab Ghana, an ISO certified laboratory. Volunteers’ fasting venous blood sample collection was performed at Medlab Ghana in Accra while in Tamale participants’ venous blood samples were drawn on site and transported to Medlab Ghana in Accra within 6hrs each day by flight. Written informed consent was obtained from all volunteers before enrolling them into the study. After the laboratory analysis results of 14 participants that had high levels of glucose, lipids, thyrotoxicosis, and iron deficiency anaemia were excluded from the reference individuals. Hence the total sample size of reference individuals recruited for the study was 501 (Figure 3.1) 65 University of Ghana http://ugspace.ug.edu.gh Total Participants Recruited based on inclusion criteria = 515 Excluded Participants after laboratory analysis High glucose = 3 High lipids = 3 Thyrotoxicosis = 4 Iron deficiency anemia = 4 Total Participants after laboratory analysis = 501 Sample size used for statistical analysis = 501 Figure 3.1:Flow diagram for recruitment of Participants 3.6 Data Collection Techniques and Tools 3.6.1 Source of Data Primary data was obtained through interviews using a structured questionnaire adopted from the IFCC protocol with modifications factoring Ghanaian cultural settings. The questionnaire consisted of participants’ socio-demographic, lifestyles, nutrition pattern, anthropometric measurement as well as clinical information. This information was necessary to aid the study to recruit the eligible participants. The laboratory analysis results of participants’ blood samples on the various biochemical parameters also served as primary data. Secondary data was obtained from reference intervals adopted by some diagnostic laboratories in Ghana, in order to compare Ghanaian reference intervals results of the specified analytes. The secondary data were useful in that their analyses helped to detect the variations of analytes between the current RI in use and the Ghanaian RI derived. 66 University of Ghana http://ugspace.ug.edu.gh 3.6.2 Sociodemographic and lifestyle characteristics The data on sociodemographic features which included age, gender, occupation and ethnicity were obtained from study participants using a structured questionnaire. Similarly, lifestyle characteristics including smoking status, alcohol consumption, engagement in physical activities (exercise) and nutrition pattern were obtained from volunteers as well. Also, data on participants’ current health status were also obtained. 3.7 Variables 3.7.1 Independent Variables. The independent variables that affect the variability of reference intervals of haematological and biochemical analytes among the healthy population within Greater Accra and Northern Region of Ghana include: ● Age ● Sex ● Ethnicity ● Body mass index (BMI) ● Blood Pressure ● Alcohol ● Smoking ● Physical activity (Exercise) ● Statistical Method 3.7.2 Dependent Variables. Reference intervals of haematological and biochemical analytes among the Ghanaian healthy population are the dependent variables. 67 University of Ghana http://ugspace.ug.edu.gh 3.8 Blood pressure and anthropometric measurement 3.8.1 Blood Pressure Measurement. Blood pressure readings of study participants were taken using a new digital automatic sphygmomanometer (Omron BP74N 5, UK) with an adjustable cuff. Participants’ blood pressure measurement were taken on the day of enrolment. The participants were made to sit in an upright position and relaxed. The cuff was wrapped on participants exposed arm 2cm above the elbow with the tube at the centre of the arm facing the front. Under a relaxed condition, the start bottom was press for the cuff to inflate, then slowly deflate. Readings of the systolic and diastolic blood pressure of the participants were recorded after the measurement was complete. 3.8.2 Weight Measurement. The study participants’ body weights were measured using Seca 799 (SecaGermany, Hamburg, Germany) electronic column weight measuring scale to ensure participants’ eligibility. The scale was placed on flat ground to ensure even surface. Participants were made to stand bare-footed upright at the centre of the measuring scale platform with their weight evenly distributed on both feet. Weight readings were recorded in kilograms after measurements. 3.8.3 Height Measurement The height measurement of the study participants was performed using a wall mounted Seca 213 stadiometers (SecaGermany, Hamburg, Germany). Participants were made to stand upright and bare-footed without a hair gear or cup. It was ensured participants’ backs and heels were against the height rule on the wall and their heads at a front view. Readings were taken in the right position to avoid parallax errors and were recorded in centimetres. 68 University of Ghana http://ugspace.ug.edu.gh 3.8.4 Body mass index (BMI) Measurement. The BMI calculation was based on comparing participants’ weight against their height. The equation used in the BMI calculation for study participants is indicated below: BMI = Weight (kg)/ Height (m)2 This was to assess participant’s risk factor for the development of disease, as well as to confirm participants’ eligibility to be enrolled in the study. 3.8.5 Waist Circumference. The waist circumferences of the study participants were measured using a flexible standard tape measure. Participants were made to stand straight and their feet closed together (about 12-15cm) with their weight equally distributed on each other. Measurements were done at the level midway between the lower rib margin and iliac crest with the tape all around the body waist in a horizontal position. Readings were taken and recorded in centimetres after the measurements. 3.8.6 Physical Activity (Exercise level). For physical exercise level calculation, the various exercise type intensity was designated with numerical values. Thus participants, who did not involve in any form of exercise was recorded as none = 0, light intensity exercise like walking = 0.5, jogging = 1.0, Soccer = 2.0). Exercise duration of 30minutes or more for one to seven day(s) =1~7. For example, jogging 30 min 3days per week (1.0  3days), the level is 3; 30-60mins walking 6days/week (0.5 6days), the level is 3. 69 University of Ghana http://ugspace.ug.edu.gh 3.8.7 Alcohol. The quantity of ethanol in grams contained in an alcoholic beverage consumed by participants was calculated by formula Mass = fluid oz  1 ounce  alcohol by volume (ABV) (%) density of ethanol. The alcohol content of alcoholic beverages varies from one beverage to the other; hence different kinds of alcoholic beverages were calculated differently using specific ABV (%). For example, participants who drink wine: 5 fluid oz of wine, with an alcohol content of 12.0% by volume, it is calculated as (5 x 29.5735 x 0.12 x .789 ≈ 14 grams of alcohol); for beer with 12 fluid oz, alcohol % varies. If the alcohol content is 5.0% by volume then the quantity of ethanol is calculated as (12 x 29.5735 x 0.05 x .789 ≈ 14 grams of alcohol). The total quantity of ethanol consumed by participants per week is the quantity of ethanol in grams multiplied by the number of days. 3.9 Laboratory analysis 3.9.1 Sample Collection, Storage, and Measurements. Blood samples taking was done under basal conditions, including overnight fasting 10 to 14 hours, sitting for at least 20 minutes prior to sampling, and avoidance of strenuous muscular exertion for three days or working a night shift before the sampling as recommended in the protocol. This helped to avoid variations since the posture prior to venipuncture could influence laboratory test (Shimizu, Ichihara & Kouguchi, 2017). A fasting blood sample of 24mL was drawn via venipunture into two (2mL) plastic evacuated tubes containing ethylene diamine tetra-acetic acid (EDTA) (BD Vacutainer Blood Collection Tube, South Africa), three (4mL) BD Vacutainer serum-separating tubes containing a clot-activator, one (4mL) BD Vacutainer sodium fluoride tube and one (4mL) lithium heparin tube (Becton-Dickinson Corp, South Africa) between 7:00 am – 10:00 am according to internationally recognised guidelines to ensure the standardisation 70 University of Ghana http://ugspace.ug.edu.gh of the pre-analytical phase. Anthropometric measurements (weight, height, waist circumference) together with the candidate’s blood pressure were documented on the day of sample collection. In the Northern region (Tamale), participants’ venous blood samples were taken on site by a qualified phlebotomist, blood samples were preserved with ice and transported to Medlab within 6 hours each day by flight for three days. 3.9.2 Pre-laboratory Analytical Process. After 15 to 30 minutes of participants’ blood sample collection, the samples were pre- analysed as follows: sorting, spinning (centrifugation) and aliquoting the serum and plasma into the 2ml serum separating tubes (SST), heparin and fluoride sample tubes. All participants’ blood samples received in the laboratory were double-checked to confirm the uniformity of their sample codes with their identification codes. All the participants’ samples were segregated into the various haematology, chemistry, and immunology sections for sample processing. The blood sample for the haematological analytes found in the purple top tube which contains EDTA as an anticoagulant were analysed with Sysmex XN1000 (Sysmex Corp, Kobe, Japan) analyser within 4 hours to 12 hours after sample collection, whiles that of the biochemical and immunoglobulin analytes were centrifuged at 1200g for 10 minutes at room temperature to separate the serum and plasma from the blood. The serum and plasma from each tube were promptly divided into 5 aliquots of 1-2mL each, using well-sealed freezing Eppendorf or sarstedt tubes, and was stored at −80 °C for 6 months until all participants samples were received for the analysis. Before the analysis, the frozen serum and plasma samples were thawed for at least 4 hours and then transferred to the analyser for analysis. tumour markers, thyroid test, hormones and immunological analytes which were not under the accreditation of Medlab Ghana were transported to Synlab, Germany for analysis. 71 University of Ghana http://ugspace.ug.edu.gh 3.9.3 Sample Transporting Procedure. The frozen samples of each participant were placed in 2mL cryovial tubes stored in 5”X5” holding boxes. Each set of participants’ samples were covered with an absorbent material (paper towel) and taped tightly to secure them such that there was enough absorbent material to absorb any liquid or leaked samples. All the wrapped boxes were placed in a Ziploc biohazard bag with each participant barcode and study number labelled on the sample for easy identification. All the samples of each participant in the Ziploc biohazards bag were placed in a Styrofoam box pack with dry ice at -80 oC to preserve the samples and transported to Synlab, Germany for the analysis of tumour markers, thyroid test, hormones and immunological analytes. 3.9.4 Target Analytes and Measurements. The target analytes for the study were haematological analytes, lipids profile, kidney and liver function tests, Immunoglobulins, tumour markers, thyroid, hormones, iron and vitamins. The analytes measured in serum included alkaline phosphatase (ALP), gamma- glutamyl transferase (GGT), aspartate aminotransferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), amylase (AMY), creatine kinase (CK), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL- C), low-density lipoprotein cholesterol (LDL-C), calcium (Ca), inorganic phosphate (IP), magnesium (Mg), phosphate(P), iron (Fe), total protein (TP), albumin (ALB), globulin, total bilirubin (TBIL), direct bilirubin (DBIL), glucose (GLU), urea (UN), creatinine (CRE), uric acid (UA), sodium (Na), potassium (K), chloride (Cl), total CO2 (TCO2), C-reactive port (CRP), C3 complement, C4 complement, anion gap, corrected CA+, eGFR male, eGFR female, free triiodothyronine (FT3), free thyroxine (FT4), thyroid stimulating hormone (TSH), vitamin B12 (VIT B12), folate (FOL), ferritin (FER), vitamin D (VIT D), parathyroid 72 University of Ghana http://ugspace.ug.edu.gh hormone (PTH) and insulin (INS), prolactin, Haemoglobin A1C, prostate specific antigen (PSA), oestrogen, progesterone, luteinizing hormone (LH), follicle stimulating hormone (FSH), testosterone, somatotrope hormone, thyreoperoxidase, thyroglobulin, immunoglobulin A, G and M (IgA, IgG and IgM), carcino-embryonic antigen (CEA), alpha 1 fetoprotein(AFP), cancer antigen125 (Ca-125), Cortisol and Transferrin. 3.9.5 Haematological Analytical Procedure. 3.9.5.1 Reagents and Equipment. ● CELLPACK DCL (DILUENT) – Reagent for measuring the numbers and sizes of RBC and platelets. ● SULFOLYSER - Reagent used for the automated determination of haemoglobin concentration of blood. ● LYSERCELL WNR – Lysing reagent was used in haemolyzing red blood cells ● FLUOROCELL WNR – This reagent was used to stain the nucleated cells in diluted and analysing white blood cell count, the counts and percentages of neutrophils, lymphocytes, monocytes and eosinophils basophil. 3.9.5.2 Determination of Haematological analytes. Haematological (full blood count) analytes included haemoglobin (HGB), hematocrit (HCT), mean corpuscular haemoglobin volume (MCV), mean corpuscular haemoglobin concentration (MCHC), mean platelet volume (MPV), red cell distribution width (RDW), Red blood cells (RBC), White blood cells (WBC), neutrophils (NEUT#), eosinophils (EOS#), lymphocytes (LYMPH#), basophils (BASO#), monocytes (MONO#) and platelet (PLT) count were analysed with the Sysmex XN1000 (Sysmex Corp, Kobe, Japan) automated analyser. The Sysmex XN-1000 is a quantitative automated haematology 73 University of Ghana http://ugspace.ug.edu.gh analyzer for in-vitro diagnostic use for determining 21 haematological parameters. Four milliliters (4mL) of whole blood sample of the study participants drawn into EDTA-3K anticoagulant tubes (Becton-Dickinson Corp, South Africa) with a purple cork. Each blood sample tube was immediately inverted for 3~4 times to ensure uniformity and prevent coagulation. The barcoded sample details of each participant were scanned and recorded on the information processing unit of the sysmex analyser automatically. Participants barcoded samples were placed on the sysmex analyser rack feeder (10 samples at a time) for analysis. All barcoded samples on the rack were then placed into the sample collection point of the analyser with the start bottom pressed, the analyser aspirated 90l of blood sample of each participants from the barcoded sampler tube in the rack feeder for full blood counts analysis. The Sysmex XN-1000 analyzer directly measured the White blood cell (WBC), Red blood cell (RBC), Haemoglobin (Hb), Haematocrit (Ht), Platelets (PLT), Nuetrophil count (NEUT#), Lymphocytes count (LYMPH#), Monocytes count (MONO#), Eosinophil count (EOS#), and Basophil (BASO#). The remaining parameters were calculated Mean corpuscular volume (MCV), Mean corpuscular haemoglobin (MCH), Mean corpuscular haemoglobin concentration (MCHC), Red cell distribution width (RDW), Mean platelet volume (MPV), and differential percentages by the analyzer. The Sysmex XN-1000 (Sysmex Corp, Kobe, Japan) counts the number and sizes of red blood cells (RBC) and platelet (PLT) using electronic resistance detection enhanced by hydrodynamic focusing. Hematocrit (HCT) was measured as the ratio of the total RBC volume to whole blood using cumulative pulse height detection. Haemoglobin (HGB) was converted to SLS-haemoglobin and read photometrically. All laboratory investigations were carried out in accordance with the laboratory’s standard operating procedures (SOPs). Sysmex XN-1000 manufacturer’s manual and reagent/kit 74 University of Ghana http://ugspace.ug.edu.gh package inserts requirements were adhered to in order to ensure good laboratory practices and procedures. 3.9.6 Biochemical Analytical Procedure. All the chemistry analytes were performed using Beckman AU480 coulter analyser (Beckman Coulter, Tokyo, Japan). The 3 SST blood sample tubes of each participant were centrifuged to separate serum from the blood cell. The aspirated serum of each participant was mixed together to ensure homogeneity. The serum samples from each tube were aliquoted into 2mL Eppendorf or sarstedt tubes labelled with each participant unique barcode number. Samples were packed in cryo box and refrigerated at -80C until all 515 samples were collected for analysis. During the analysis of the chemistry analytes, samples were thawed in batches of 20 at room temperature. The barcoded samples for the chemical analytes were programmed using their unique accession codes into the Beckman Coulter AU480 (Beckman Coulter, Tokyo, Japan) analyser’s information processing system unit (IPU). 20 barcoded samples were arranged on the rack feeder and loaded into the analyser for the analysis of the following biochemical analytes alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), aspartate aminotransferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), amylase (AMY), creatine kinase (CK), creatinine (CRE), total protein (TP), albumin (ALB), globulin, total bilirubin (TBIL), direct bilirubin (DBIL), glucose (GLU), urea (UN), uric acid (UA), sodium (Na), potassium (K), chloride (Cl), total CO2 (TCO2), C-reactive port (CRP), anion gap, corrected CA+, eGFR male, eGFR female, ferritin, transferrin. triglycerides (TG), total cholesterol (TC), high-density 75 University of Ghana http://ugspace.ug.edu.gh lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), calcium (Ca), inorganic phosphate (IP), magnesium (Mg), phosphate(P) and iron (Fe). After analysis, the results for each analyte were transmitted into Medlab Ghana’s Schuylab platform (an internet module that allows easy access to participants’ laboratory results on- line from any computer equipped with a web browser). 3.9.7 Gamma-Glutamyl Transferase Analytes (GGT). 3.9.7.1 Required Materials, Reagents and Equipment. GGT Reagent Cartridge, Saline (monoammonium salt and Sodium hydroxide solution), control material and Beckman Coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.7.2 Determination of serum GGT concentration by enzymatic rate method. Gamma-Glutamyl Transferase (GGT) was determined by the enzymatic rate of Gamma- glutamyl-3-carboxy-4-nitroanalide method based on the recommendation of the IFCC (AU480 Beckman Coulter User Manual, 2017). A volume of 50L of serum was aspirated by the analyzer and assayed for GGT concentration. The principles of the method involve GGT catalysing the transfer of the gamma-glutamyl group from the substrate (gamma- glutamyl-3-carboxy-4-nitroanlide) to the glycylglycine acceptor to yield 5-amino-2- nitobenzoate. The rate of absorbance increases at 410/480 nm due to the formation of 5- amino-2-benzoate which is directly proportional to the GGT activity in the sample (AU480 Beckman Coulter User Manual, 2017; Schumann et al., 2002; IFCC, 1986). 3.9.7.3 Reaction. L-γ-Glutamyl-3-carboxy-4-nitroanilide + Glycylglycine –GGT→ 5-Amino-2- nitrobenzoate + L-γ-Glutamyl-glycylglycine 76 University of Ghana http://ugspace.ug.edu.gh L-γ-Glutamyl-3-carboxy-4-nitroanilide + L-γ-Glutamyl3-carboxy-4-nitroanilide –GGT→ 5-Amino-2-nitrobenzoate + L-γ-Glutamyl-γ-glutamyl -3- carboxy-4-nitroanilide 3.9.8 Alkaline phosphatase (ALP). 3.9.8.1 Required Materials, Reagents and Equipment. Alkaline Phosphatase Reagent Cartridge, Saline, pNPP Substrate, AMP buffer, control material, Beckman Coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.8.2 Determination of serum ALP concentration by kinetic rate method. Alkaline phosphatase (ALP) was determined by kinetic rate method based on the recommendation of the German Society for Clinical Chemistry (AU480 Beckman Coulter User Manual, 2017). A volume of 50L of serum was aspirated by the analyzer and assayed for ALP concentration using a 2-Amino-2-Methyl-1-Propanol (AMP) buffer to measure ALP activity in the serum. In the reaction, the ALP catalyzes the hydrolysis of the colourless organic phosphate ester substrate, p-Nitrophenyl phosphate (pNPP), to the yellow coloured product p-Nitrophenol (pNP) and phosphate. This reaction occurs at an alkaline medium at pH of 9.8. The system monitors the rate of change in absorbance at 410/480 nm over a fixed-time interval. This rate of change in absorbance is directly proportional to the ALP activity in the serum (National Center for Health Statistics, 2015). 3.9.8.3 Reactions. P-Nitrophenyl phosphate + H20 P- Nitrophenol +Phosphate Alkaline medium 77 University of Ghana http://ugspace.ug.edu.gh 3.9.9 Aspartate Aminotransferase (AST). 3.9.9.1 Required Materials, Reagents and Equipment. Aspartate aminotransferase Cartridge, saline, pyridoxal phosphate liquid reagent, control material, Beckman coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.9.2 Determination of serum AST concentration by IFCC method. AST procedure utilizes a tris buffer with a pyridoxal-5-phosphate method based on the recommendations of the IFCC. In this method, a volume of 50L of serum was aspirated by the analyzer and assayed for AST concentration. Aspartate aminotransferase (AST) catalyses the transformation of aspartate and 2-oxoglutarate, forming L-glutamate and oxalacetate. The addition of pyridoxal phosphate to the reaction mixture ensures maximum catalytic activity of AST. The oxalacetate is reduced to L-malate by malate dehydrogenase (MDH), while ß-Nicotinamide Adenine Dinucleotide (reduced form) (NADH) is simultaneously converted to ß-Nicotinamide Adenine Dinucleotide (NAD) (AU480 Beckman Coulter User Manual , 2017). The decrease in absorbance due to the consumption of NADH is measured at 340 nm and was proportional to the AST activity in the sample (AU480 Beckman Coulter User Manual, 2017; Schumann et al., 2002b; IFCC, 1986). 3.9.9.3 Reaction. AST L-Aspartate + α-Oxoglutarate L-Glutamate + Oxalacetate MDH Oxalacetate + NADH + H+ L -Malate + NAD+ 78 University of Ghana http://ugspace.ug.edu.gh 3.9.10 Alanine Aminotransferase (ALT). 3.9.10.1 Required Materials, Reagents and Equipment Alanine aminotransferase Cartridge, pyridoxal phosphate liquid reagent, Saline, control material, Beckman coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.10.2 Determination of serum ALT concentration by IFCC method. ALT was determined by the tris buffer with a pyridoxal-5-phosphate method based on the recommendations of the IFCC. A volume of 50L of serum was aspirated by the analyzer and assayed for ALT concentration. ALT transfers the amino group alanine to 2- oxoglutarate to form pyruvate. The addition of pyridoxal phosphate to the reaction mixture ensures maximum catalytic activity of ALT. The pyruvate enters a lactate dehydrogenase (LDH) catalysed reaction with ß-Nicotinamide Adenine Dinucleotide (NADH) to produce lactate and ß-Nicotinamide Adenine Dinucleotide (NAD). The decrease in absorbance due to the consumption of NADH is measured at 340 nm and is directly proportional to the ALT activity in the sample (AU480 Beckman Coulter User Manual, 2017; Schumann et al., 2002a; IFCC, 1986). 3.9.10.3 Reaction L-Alanine+2-Oxoglutarate -ALT→ Pyruvate+L-Glutamate Pyruvate+NADH+H+ -LDH*→Lactate + NAD+ 3.9.11 Lactate Dehydrogenase (LDH). 3.9.11.1 Required Materials, Reagents and Equipment. LDH Reagent Cartridge, Saline, Control Material, Beckman coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 79 University of Ghana http://ugspace.ug.edu.gh 3.9.11.2 Determination of serum LDH concentration by enzymatic rate method. The LDH concentration in the participants’ serum were determined by utilizing an enzymatic rate method. A volume of 50L of serum was aspirated by the analyzer and assayed for LDH concentration. In the reaction, the LDH catalyzes the reversible oxidation of L-Lactate to Pyruvate with the concurrent reduction of ß-Nicotinamide Adenine Dinucleotide (NAD) to ß-Nicotinamide Adenine Dinucleotide (NADH). The system monitors the rate of change in absorbance at 340 nm over a fixed-time interval. The rate of change in absorbance is directly proportional to the activity of LDH in the sample. (AU480 Beckman Coulter User Manual, 2017; Bais & Philcox, 1994). 3.9.12 Amylase (AMY) 3.9.12.1 Required Materials, Reagents and Equipment Amylase Reagent Cartridge, Saline, control material, and Beckman Coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.12.2 Determination of serum Amylase concentration by Beckman Olympus- blocked CNPG3 method. α-Amylase concentration in serum was determined by Beckman Olympus-blocked CNPG3 method. A volume of 50l of serum was aspirated by the analyzer and assayed for Amylase concentration. In the reaction, α-Amylase hydrolyzes the 2-chloro-4- nitrophenyl-α-D-maltotrioside (CNPG3) as a substrate to release 2-chloro-4-nitrophenol (CPNP) and form 2-chloro-4-nitrophenyl-α-D-maltoside (CNPG2), maltotriose, and glucose without any ancillary enzymes. The rate of formation of the 2-chloro-4- nitrophenol can be detected spectrophotometrically at 410 nm to give a direct 80 University of Ghana http://ugspace.ug.edu.gh measurement of α-amylase activity in the sample (AU480 Beckman Coulter User Manual, 2017. 3.9.12.3 Chemical Reaction. α-amylase 10 CNPG3 9 CNP + CNPG2 + 9G3 + G 3.9.13 Creatine Kinase (CK). 3.9.10.13.1 Required Materials, Reagents and Equipment. Creatine Kinase Reagent Cartridge, Saline, Control Material, Beckman coulter AU480 (Beckman Coulter, Tokyo, Japan). 3.9.13.2 Determination of serum CK concentration by Immune-Inhibition (IFCC) method. Creatine Kinase (CK) was determined based on the recommendations of the IFCC with immune-inhibition. A volume of 50L of serum was aspirated by the analyzer and assayed for CK concentration. In the reaction CK reversibly catalyses the transfer of a phosphate group from creatine phosphate to adenosine diphosphate (ADP) to give creatine and adenosine triphosphate (ATP) as products, The ATP formed is used to produce glucose- 6-phosphate and ADP from glucose. This reaction is catalysed by hexokinase (HK) which requires magnesium ions for maximum activity. The glucose-6-phosphate is oxidized by the action of the enzyme glucose-6-phosphate dehydrogenase (G6P-DH) with simultaneous reduction of the coenzyme nicotinamide adenosine dinucleotide (NADP) to give NADPH and 6-phosphogluconate. The rate of increase of absorbance at 340/660 nm due to the NADPH is directly proportional to the activity of CK in the sample. 81 University of Ghana http://ugspace.ug.edu.gh 3.9.13.3 Chemical Reaction. Creatine phosphate + ADP CK (AMP, NAC) Creatine +ATP ATP + Glucose HK ADP + G6P G6P + NADP+ +H2 G6P-DH Gluconate-6P + NADPH + H+ 3.9.14 Creatinine (CRE). 3.9.13.1 Required Materials, Reagents and Equipment Creatinine Reagent Cartridge, System Calibrator (Cat. No.66300), Urine Calibrator Cat.No.ODC0025, Antifoam, control material, Saline and Beckman coulter AU480 (Beckman Coulter, Tokyo, Japan). 3.9.14.2 Determination of serum Creatinine concentration by kinetic method. Creatinine concentration in serum was determined by the kinetic method. A volume of 50L of serum was aspirated by the analyzer and assayed for creatinine concentration. Picric acid in an alkaline medium reacts with creatinine to form yellow–orange coloured compound with alkaline picrate. The rate of change in absorbance at 520/800 nm is proportional to the creatinine concentration in the sample. 3.9.14.3 Chemical Reaction. Creatinine + Alkaline Picrate Orange coloured compound 3.9.15 Principle Procedure for Total Protein (TP). 3.9.15.1 Required Materials, Reagents and Equipment. Beckman Coulter Total Protein Reagent Cartridges (R1 and R2), System Calibrator (Part Number: 66300), and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 82 University of Ghana http://ugspace.ug.edu.gh 3.9.15.2 Determination of serum Total Protein concentration by timed rate biuret method. The timed rate biuret method was used in determining the amount of total protein concentration in participants’ samples. A volume of 50L of serum was aspirated by the analyzer and assayed for total protein. Proteins in the sample combine with cupric ions (Cu2+) in an alkaline solution, which forms copper-protein complexes. The proteins and polypeptides formed contains at least two peptide bonds to produce a violet coloured complex. The absorbance of the complex at 540/660 nm is directly proportional to the concentration of total protein in the sample. 3.9.16. Principle Procedures of Albumin (ALB). 3.9.16.1 Required Materials, Reagents and Equipment. Albumin Reagent Cartridge, System Calibrator (Cat.No.66300), Saline, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.16.2 Determination of serum Albumin concentration by Bromocresol Green (BCG) method. Albumin concentration in serum was determined by Bromocresol Green (BCG) method. A volume of 50L of serum was aspirated by the analyzer and assayed for albumin concentration. Albumin binds with bromocresol green (BCG) at an acidic pH (4.3) to produce a colour change of the indicator from yellowish green to bluish green colour. The absorbance of the albumin-BCG complex is measured bichromatically (600/800nm) and is proportional to the albumin concentration in the sample. 83 University of Ghana http://ugspace.ug.edu.gh 3.9.16.3 Chemical Reaction. Albumin + bromocresol green albumin-bromcresol green-complex 3.9.17 Principle Procedures of Urea. 3.9.17.1 Required Materials, Reagents and Equipment. System Calibrator Cat.No.66300 and Urine Calibrator Cat.No.ODC0025, Urea Reagent Cartridge, Saline, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.17.2 Determination of serum Urea concentration by kinetic method. Urea concentration in serum was determined by the kinetic method. A volume of 50L of serum was aspirated by the analyzer and assayed for urea concentration. Urea is hydrolysed in the presence of water and urease to produce ammonia and carbon dioxide. The ammonia produced in the first reaction combines with 2-oxoglulate and NADH in the presence of glutamate-dehydrogenase (GLDH) to yield glutamate and NAD. The decrease in NADH absorbance per unit time is proportional to the urea concentration. 3.9.17.3 Chemical Reaction. Urease Urea + H2O 2 NH4 + + CO 2- 3 GLDH NH4+ + α-Oxoglutarate + NADH L-glutamate + NAD+ + H2O 3.9.18 Principle Procedures of Total Bilirubin (TBIL). 3.9.18.1 Required Materials, Reagents and Equipment. Total Bilirubin Reagent Cartridge, System Bilirubin calibrator Cat. No 66300, control material, saline, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 84 University of Ghana http://ugspace.ug.edu.gh 3.9.18.2 Determination of serum Total Bilirubin concentration by Dichlorophenyl Diazonium method. Total Bilirubin concentration in participants’ serum samples was determined by Dichlorophenyl Diazonium method. A volume of 50L of serum was aspirated by the analyzer and assayed for total bilirubin. A stabilized diazonium salt, 3-5- dichlorophenyldiazonium tetrafluoroborate (DPD), reacts with conjugated bilirubin and unconjugated bilirubin directly in the presence of surfactant an accelerator to form azobilirubin. The increase in absorbance at 450 nm is proportional to the total bilirubin concentration. 3.9.18.3 Chemical Reaction. Caffeine Bilirubin + DPD Azobilirubin Surfactant 3.9.19 Principle Procedures of Direct Bilirubin (DBIL). 3.10.19.1 Required Materials, Reagents and Equipment. Direct Bilirubin Reagent Cartridge, Systems Calibrator Cat. No 66300, Deionized water, control material and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.19.2 Determination of serum Direct Bilirubin concentration by Dichlorophenyl Diazonium method. Direct Bilirubin concentration in participants’ serum samples was determined by Dichlorophenyl Diazonium method. A volume of 50L of serum was aspirated by the analyzer and assayed for direct bilirubin. In the reaction stabilized diazonium salt, 3,5 85 University of Ghana http://ugspace.ug.edu.gh Dichlorophenyl diazonium tetrafluoroborate (DPD) couples directly with direct (conjugated) bilirubin in an acid medium to form azobilirubin. The absorbance at 570 nm is proportional to the direct bilirubin concentration in the sample. 3.9.19.3 Chemical Reaction. Bilirubin + 3,5-Dichlorophenyl Diazonium (BF)4 Azobilirubin 3.9.20 Principle Procedures of Glucose (GLU). 3.9.20.1 Required Materials, Reagents and Equipment. Beckman Coulter Glucose Reagent Cartridges (R1 and R2), Beckman Coulter System Calibrator (Part Number: 66300), Beckman Coulter Control Serum Levels 1 and 2 (Part Numbers ODC0003 and ODC0004) Biorad Liquid Assayed Multiqual Premium Level 1 Biorad Liquichek Spinal Fluid Control Levels 1 and 2 Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.20.2 Determination of serum Glucose concentration by hexokinase method. Glucose concentration in participants’ serum samples were determined by hexokinase method. A volume of 50L of serum was aspirated by the analyzer and assayed for glucose concentration. In the reaction glucose phosphorylated by hexokinase (HK) in the presence of adenosine triphosphate (ATP) and magnesium ions to form glucose-6-phosphate (G-6- P) and adenosine diphosphate (ADP). G-6-P is then oxidized by glucose-6- phosphate dehydrogenase (G-6-PDH) in the presence of nicotinamide adenine dinucleotide (NAD) producing 6-phosphogluconate and NADH. The formation of NADH causes increase in absorbance at 340 nm which is directly proportional to the concentration of glucose in the sample. 86 University of Ghana http://ugspace.ug.edu.gh 3.9.20.3 Chemical Reaction. Glucose + ATP HK G6P + ADP G6P + NAD + G6PDH 6-Phosphogluconate + NADH + H+ 3.9.21 Principle Procedures of Uric acid (UA). 3.9.21.1 Required Materials, Reagents and Equipment. Uric acid Reagent Cartridge, saline, control material, System Calibrator No. 66300, Urine Calibrator (Cat. No.ODC0025), Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.21.2 Determination of serum Uric acid concentration. Uric acid concentration in the serum was determined by uricase peroxidase without ascorbate oxidase. A volume of 50L of serum was aspirated by the analyzer and assayed for uric acid concentration. Uric acid in the sample is oxidized to allantoin and hydrogen peroxide in the presence of uricase. The Trinder reaction is utilized to measure H2O2. The formed H2O2 reacts with N, N-bis (4-sulfobutly)-3,5-dimethylaniline, disodium salt (MADB) and 4-aminoantipyrine (4-AAP) in the presence of peroxidase to produce a chromophore, which is read biochromatically at 660/800nm. The amount of dye formed is proportionate to the uric acid concentration in the sample. 3.9.21.3 Chemical Reaction Uricase Uric Acid + O2 + H2O Allantoin + CO2 + H2O2 Peroxides 2 H2O2 + 4-AAP + MADB 3 H2O + Blue Dye + OH - 87 University of Ghana http://ugspace.ug.edu.gh 3.9.22 Principle Procedure of Total Cholesterol . 3.9.22.1 Required Materials, Reagents and Equipment. Cholesterol Reagent Cartridge, System Calibrator (No.66300), saline, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.22.2 Determination of serum Total Cholesterol concentration enzymatic timed- point method. Cholesterol Reagent is used to measure total cholesterol concentration by the enzymatic timed-point method using the Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50L of serum was aspirated by the analyzer and assayed for cholesterol concentration. In the reaction, total cholesterol esterase (CE) hydrolyzes cholesterol esters to free cholesterol and fatty acids. Free cholesterol is oxidized to cholestene-3-one and hydrogen peroxide (H2O2) by cholesterol oxidase (CO). Peroxidase (POD) catalyzes the reaction of hydrogen peroxide with 4-aminotipyrine (4- AAP) and phenol to produce a coloured quinoneimine product. The red quinoneimine dye formed was measured spectrophotometrically at 540/600 nm as an increase in absorbance of total cholesterol (TC). The intensity of the coloured complex formed is directly proportional to the concentration of total cholesterol in the sample. 3.10.22.3 Chemical Reaction. Cholesteryl ester Cholesteryl ester hydrolase + H2O Cholesterol + fatty acid Cholesterol oxidase Cholesterol + O2 Cholest-4-en-3-one + H2O2 peroxidase 2H2O2 + 4-aminophenazone + phenol Quinoneimine + 4 H2O 88 University of Ghana http://ugspace.ug.edu.gh 3.9.23 Principle Procedures of High-Density Lipoprotein Cholesterol (HDL- C). 3.9.23.1 Required Materials, Reagents and Equipment. HDL Reagent Cartridge, HDL-Cholesterol Calibrator (ODC0011), saline, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.23.1 Determination of serum HDL-Cholesterol concentration by Enzymatic Immunoinhibition method. HDL-Cholesterol concentration was measured by the enzymatic Immunoinhibition method using the Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50L of serum was aspirated by the analyzer and assayed for HDL-Cholesterol concentration. The HDL-Cholesterol test was a two-reagent homogenous system for the selective measurement of serum HDL-Cholesterol in the presence of other lipoprotein particles. The assay comprised of two distinct phases. In phase one, free cholesterol in non-HDL-lipoproteins was solubilized and consumed by cholesterol oxidase, peroxidase, and DSBmT to generate a colourless end product. In phase two a unique detergent selectively solubilizes HDL- lipoproteins. The HDL cholesterol is released for reaction with cholesterol esterase, cholesterol oxidase and a chromogen system to yield a blue colour complex which is measured bichromatically at 600/700nm. The resulting increase in absorbance is directly proportional to the HDL-C concentration in the sample. 3.10.23.3 Chemical Reaction. Reaction Phase 1 Accelerator + CO LDL, VLDL Chylomicrons Colorless end product DSBmT + Peroxidase 89 University of Ghana http://ugspace.ug.edu.gh Reaction Phase 2 HDL Specific detergent HDL Cholesterol HDL Disrupted CHE and CHO HDL Cholesterol + H2O + O2 Cholest-4-en-3-one + H2O2 Peroxidase H2O2 + DSBmT + 4-AAP Blue color complex 3.9.24 Principle Procedures of Low-Density Lipoprotein Cholesterol (LDL-C). 3.10.24.1 Required Materials, Reagents and Equipment. LDL Reagent Cartridge, LDL-Cholesterol Calibrator (ODC0011), saline, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.24.2 Determination of serum LDL--Cholesterol concentration by Enzymatic selective method. The LDL-Cholesterol test is a two-reagent homogenous system. The assay is comprised of two distinct phases. A volume of 50L of serum was aspirated by the analyzer and assayed for LDL-Cholesterol concentration. In phase one a unique detergent solubilizes cholesterol from non-LDL- lipoprotein particles. This cholesterol is consumed by cholesterol esterase, cholesterol oxidase, peroxidase and 4- aminoantipyrine to generate a colourless end product. In phase two a second detergent in reagent 2 releases cholesterol from the LDL – lipoproteins. This cholesterol reacts with cholesterol esterase, cholesterol oxidase and a chromogen system to yield a blue colour complex which can be measured bichromatically at 540/660nm. The resulting increase in absorbance is directly proportional to the LDL-C concentration in the sample. 90 University of Ghana http://ugspace.ug.edu.gh 3.10.24.3 Chemical Reaction. Reaction Phase 1 CHE and CHO HDL-C, VLDL-C LDL-C Chylomicrons Cholest-4-en-3-one + Fatty acids +H2O2 Peroxidase H2O2 – 4-AAP LDL-C +Colorless end product Reaction Phase 2 CHE and CHO LDL Cholesterol Cholest-4-en-3-one + Fatty acids +H2O2 Peroxidase H2O2 + DSBmT + 4-AAP Blue color complex 3.9.25 Principle Procedure of Triglycerides (TG). 3.9.25.1 Required Materials, Reagents and Equipment. Triglycerides Reagent Cartridge, System Calibrator No. 66300, Saline, Two levels of control material, Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.25.2 Determination of serum Triglycerides concentration by glycerol phosphate oxidase – Peroxidase method. Triglyceride concentration was measured by glycerol phosphate oxidase – Peroxidase (GPO-POD) method using Beckman Coulter AU 480 chemistry auto analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50L of serum was aspirated by the analyzer and assayed for triglycerides concentration. The procedure is based on a series of coupled enzymatic reactions. The triglycerides concentration in the serum was determined after enzymatic hydrolysis with a combination of microbial lipases to yield glycerol and fatty acids. The glycerol is phosphorylated by adenosine triphosphate (ATP) in the presence of glycerol kinase (GK) to produce glyceol-3-phosphate. The glycerol-3-phosphate is oxidized by molecular oxygen in the presence of GPO (glycerol phosphate oxidase) to 91 University of Ghana http://ugspace.ug.edu.gh produce hydrogen peroxide (H2O2) and dihydroxyacetone phosphate. The formed H2O2 reacts with 4-aminophenazone and N, N-bis(4-sulfobutyl)-3,5-dimethylaniline, disodium salt (MADB) in the presence of peroxidase (POD) to produce a chromophore, which is read at 660/800nm.The increase in absorbance at proportional to the triglyceride content of the sample. 3.10.24.3 Chemical Reaction. Lipase Triglycerides + 3 H2O Glycerol + 3 Fatty Acids GK, Mg2+ Glycerol + ATP Glycerol-3-phosphate + ADP GPO Glycerol-3-phosphate + O2 H2O2 + Dihydroxyacetone phosphate Peroxidase 2 H2O2 + MADB + 4AAP Blue Dye + OH- + H2O 3.9.26 Principle Procedure of Electrolytes (Na+ K+, Cl-). 3.9.26.1 Required Materials, Reagents and Equipment. The Beckman Coulter 480 AU chemistry analyzer (Beckman Coulter, Tokyo, Japan) uses the Ion Selective Electrode (ISE) methods for determining sodium (Na+), potassium (K+), and chloride (Cl-). It employs crown ether membrane electrodes for sodium and potassium; and a molecular oriented polyvinylchloride (PVC) membrane for chloride that are specific for each ion of interest in the sample. An electrical potential is developed according to the Nernst Equation for a specific ion. When compared to the Internal Reference Solution, this electrical potential is translated into voltage and then into the ion concentration of the sample. 92 University of Ghana http://ugspace.ug.edu.gh 3.9.27 Principle Procedure of Sodium (Na+). 3.9.27.1 Required Materials, Reagents and Equipment. Ion selective electrode (ISE) electrolyte buffer reagent, ISE electrolyte reference reagent, Saline, control material, Low Serum standard calibrator and High serum standard calibrator, Beckman Coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.27.2 Determination of serum Sodium (Na+) concentration by ion selective electrode method . The determination of sodium is one of most important functions in the clinical laboratory. The method used in determing sodium concentration is ion selective electrode (ISE) module. A volume of 50L of serum was aspirated by the analyzer and assayed for sodium concentration. The ISE Module for sodium employs crown ether membrane electrode that is specific for sodium ion in the sample. An electrical potential is developed according to the Nernst Equation for sodium. When compared to the internal reference solution, this electrical potential is translated into voltage and then into sodium ion concentration of the sample. The Nernst equation for ISE E=E0+(2.030 RT/nF)logC , where E is potential, E0 is a constant characteristic of a particular ISE, R is the gas constant (8.314 J/K.mol), T is the temperature (in K), n is the charge of the ion and F is Faraday constant (96,500 coulombs/mol). 3.9.28 Principle Procedure of Potassium (K+). 3.9.28.1 Required Materials, Reagents and Equipment. Ion selective electrode (ISE) electrolyte buffer reagent, ISE electrolyte reference reagent, ISE MID Standard solution, Saline, control material, Beckman Coulter AU480 analyzer 93 University of Ghana http://ugspace.ug.edu.gh (Beckman Coulter, Tokyo, Japan), ISE low serum standard (Cat No.66317) and ISE high standard Cat (No. 66316). 3.9.28.2 Determination of serum Potassium concentration by ion selective electrode method. The method used in determining potassium concentration involved ion selective electrode. A volume of 50L of serum was aspirated by the analyzer and assayed for potasium concentration. The ISE Module for potassium employs crown ether membrane electrode that is specific for potassium ion in the sample. An electrical potential is developed according to the Nernst Equation for potassium. When compared to the internal reference solution, this electrical potential is translated into voltage and then into potassium ion concentration of the sample. 3.9.29 Principle Procedure of Chloride (Cl-). 3.9.29.1 Required Materials, Reagents and Equipment. Ion selective electrode (ISE) electrolyte buffer reagent, ISE electrolyte reference reagent, ISE low serum standard Cat No:66317, ISE high serum standard Cat No:66316, Urine Calibrator Cat.No.ODC0025, control material, Beckman Coulter AU480 analyzer (Beckman Coulter, Tokyo, Japan). 3.9.29.2 Determination of serum Chloride concentration by ion selective electrode method. The chloride test was determined by ion selective electrode method. A volume of 50L of serum was aspirated by the analyzer and assayed for chloride concentration. The ISE Module for chloride employs a molecular oriented polyvinylchloride (PVC) membrane 94 University of Ghana http://ugspace.ug.edu.gh that is specific for chloride ion in the sample. An electrical potential is developed according to the Nernst Equation for Chloride. When compared to the Internal Reference Solution, this electrical potential is translated into voltage and then into chloride ion concentration of the sample. 3.9.30 Principle Procedure of Phosphorous. 3.10.29.1 Required Materials, Reagents and Equipment. Phosphorus Reagent Cartridge, Saline, control material, System Calibrator Cat.No.66300, and Beckman Coulter 480 AU chemistry analyzer (Beckman Coulter, Tokyo, Japan). 3.9.30.2 Determination of serum Phosphorous concentration. A volume of 50L of serum was aspirated by the analyzer and assayed for phosphorous concentration. inorganic Phosphate reacts with molybdate to form a heteropolyacid complex. The absorbance at 340/380 nm is directly proportional to the inorganic phosphorous concentration in the sample. 3.10.30.3 Chemical Reaction. H+ Phosphate + Molybdate  Heteropolyacid complex 3.9.31 Principle Procedure of Calcium. 3.9.31.1 Required Materials, Reagents and Equipment. Calcium Reagent Cartridge, System Calibrator No.66300, Urine Calibrator CatNo.ODC0025, control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 95 University of Ghana http://ugspace.ug.edu.gh 3.9.31.2 Determination of serum Calcium concentration by Arsenazo method. The calcium test was determined by Arsenazo method using the Beckman Coulter AU480 chemistry analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50L of serum was aspirated by the analyzer and assayed for calcium concentration. Calcium ions (Ca2+) reacts with Arsenazo III (2,2’-[1,8-Dihydroxy-3,6-disulphonaphthylene-2,7-bisazo]- bisbenzenear-sonic acid) to form an intense purple colored complex. In this method the Ca-Arsenazo III complex is measured bichromatically at 660/700 nm. The resulting increase in absorbance of the reaction mixture is directly proportional to the calcium concentration in the sample. 3.10.31.3 Chemical Reaction. Ca2+ + Arsenazo III Ca-Arsenazo III complex (purple) Acidic Medium 3.9.32 Principle Procedure of Magnesium. 3.9.32.1 Required Materials, Reagents and Equipment. Magnesium Reagent Cartridge, Saline, Two levels of control material, System Calibrator Cat. No.66300, Urine Calibrator Cat.No.ODC0025, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.32.2 Determination of serum Magnesium concentration by Xylidyl Blue and Tris- Buffer method. Magnesium concentration was determined by Xylidyl Blue and Tris-Buffer method using Beckman Coulter AU480 chemistry analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50 L of serum was aspirated by the analyzer and assayed for magnesium concentration. The magnesium reagent utilizes a direct method in which magnesium ions form a complex 96 University of Ghana http://ugspace.ug.edu.gh with xylidyl blue in a highly concentrated solution. The colour produced is measured bichromatically at 520//800 nm and is proportional to the magnesium concentration in the sample. Calcium interference is eliminated by glycoletherdiamine-N, N, N’,N’ – tetraacetic acid (GEDTA) (Tietz, 1987). 3.10.32.3 Chemical Reaction. pH 11.4 Mg2+ + Xylidyl blue Purple complex 3.9.33 Principle Procedure of C – Reactive Protein (CRP). 3.9.33.1 Required Materials, Reagents and Equipment. C-Reactive protein Reagent Cartridge, Serum Protein Multi-Calibrator (Cat.No. ODR3021), control material, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.33.2 Determination of serum C – Reactive Protein (CRP) concentration by Immunoturbidimetric method. C-Reactive protein concentration was determined by Immunoturbidimetric method using Beckman Coulter AU480 chemistry analyzer (Beckman Coulter, Tokyo, Japan). A volume of 50L of serum was aspirated by the analyzer and assayed for C-Reactive protein concentration. Serum is mixed with reagent thus R1 buffer at a pH 7.4 and R2 antiserum solution, CRP reacts specifically with anti-human CRP antibodies to yield insoluble reagent. The absorbance of these aggregates is directly proportional to the CRP concentration in the sample. 97 University of Ghana http://ugspace.ug.edu.gh 3.10.33.3 Chemical Reaction. CRP(sample) + anti-CRP(antibodies) CRP(sample)-antibody complex 3.9.34 Principle Procedure of Carbon Dioxide. 3.9.34.1 Required Materials, Reagents and Equipment. Bicarbonate reagent cartridge, Bicarbonate calibrator Cat. No. ODC0019, Three levels of synchron multilevel controls material Cat No.657365, and Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan). 3.9.34.2 Determination of serum Carbon Dioxide concentration by Enzymatic method. Carbon dioxide in serum was determined using the bicarbonate reagent, which utilizes the enzymatic method. A volume of 50L of serum was aspirated by the analyzer and assayed for carbon dioxide concentration. In this process bicarbonate (HCO -3 ) and phosphoenolpyruvate (PEP) are converted to oxaloacetate and phosphate in the reaction catalyzed by phosphoenolpyruvate carboxylase (PEPC). Malate dehydrogenase (MD) catalyzes the reduction of oxaloacetate to malate with the concomitant oxidation of reduced nicotinamide adenine dinucleotide (NADH). The oxidation of NADH results in a decrease in absorbance of the reaction mixture measured bichromatically at 380/410nm proportional to the Bicarbonate content of the sample. 3.10.33.3 Chemical Reaction. PEPC + Mg2+ Phosphoenolpyruvate + HCO - 3 Oxaloacetate +H2PO - 4 MDH Oxaloacetate + NADH analog + H+ Malate +NAD+ analog 98 University of Ghana http://ugspace.ug.edu.gh 3.9.35 Hormones Analytical Procedure. Some hormones and thyroid analytes (PSA, Oestrogen, Progesterone, LH, FSH, Prolactin, TSH, FT3, FT4) with Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.36 Principle Procedure of Estradiol. 3.9.36.1 Required Materials, Reagents and Equipment. Estradiol Reagent Cartridge, Streptavidin-coated microparticles, Anti‐estradiol‐ Ab~biotin, two biotinylated monoclonal, anti‐estradiol antibodies mesterolone, buffer solution, Estradiol derivative labelled with ruthenium complex, and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.36.2 Determination of serum Estradiol concentration by competitive assay method. The Estradiol in participants’ serum was determined using estradiol III reagent which employs a competitive assay principle using two monoclonal antibodies specifically directed against 17β‐ estradiol. Endogenous estradiol released from the sample by mesterolone competes with the added estradiol derivative labelled with a ruthenium complex for the binding sites on the biotinylated antibody. A volume of 25μl of the participants’ serum was aspirated by the Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan) and assayed for estradiol concentration. The serum was incubated with two estradiol (E2)-specific biotinylated antibodies and mesterolone that liberates protein-bound E2. The higher the E2-concentration in the sample, the lower is the amount of unbound biotinylated antibody. Unbound biotinylated antibody was now saturated by adding estradiol conjugated to a ruthenium complex. Simultaneously, 99 University of Ghana http://ugspace.ug.edu.gh streptavidin-coated paramagnetic microparticles were added to bind all biotinylated components in the reaction mixture. The reaction mixture was aspirated into the measuring cell where the microparticles were fixed to the electrode surface by magnetic action. The unbound substances were subsequently removed. Chemiluminescence was then induced by applying a voltage and measured with a photomultiplier. The signal yield was roughly reversely proportional to the total E2 concentration in the sample. Results were determined via a calibration curve which was instrument-specifically generated by 2-point calibration and a master curve provided via the reagent barcode. 3.9.37 Principle Procedure of Progesterone. 3.9.37.1 Required Materials, Reagents and Equipment. Progesterone reagent cartridge, Streptavidin-coated microparticles, Anti-progesterone- Ab~biotin (gray cap), Biotinylated monoclonal anti‐progesterone antibody, phosphate buffer, Progesterone-peptide~Ru(bpy), Progesterone coupled to a synthetic peptide labelled with ruthenium complex, phosphate buffer, and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.37..2 Determination of serum Progesterone concentration by competitive assay method. The determination of the presence of progesterone adopted the competition assay method using Cobas e411 automated analyser by Roche within a duration of 18 minutes. A volume of 20 μl of the participants’ serum assayed for progesterone concentration was incubated with a biotinylated, monoclonal progesterone-specific antibody, ruthenylated 100 University of Ghana http://ugspace.ug.edu.gh progesterone derivative, and danazol (to release progesterone). The antibody binding sites were occupied by either progesterone or the ruthenylated derivative, with the proportion of each depending on the concentration of progesterone in the sample. Streptavidin-coated microparticles were added to the reaction mixture and the immune complexes bound to the solid phase via biotin–streptavidin interactions. The reaction mixture was transferred to a measuring cell and the microparticles were magnetically captured onto the surface of an electrode; unbound sample was washed away before a chemiluminescent reaction is induced by applying a voltage to the electrode. Chemiluminescence was measured by a photomultiplier and the concentration of Progesterone within the sample was calculated using a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode. 3.9.38 Principle Procedure of Luteinizing Hormone (LH). 3.9.38.1 Required Materials, Reagents and Equipment. Luteinizing hormone reagent cartridge, streptavidin-coated microparticles, Anti-LH- Ab~biotin (gray cap), Biotinylated monoclonal anti‐ LH antibody, TRIS buffer, Anti-LH- Ab~Ru(bpy), Monoclonal anti‐ LH antibody labelled with ruthenium complex, and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.38.2 Determination of serum Luteinizing Hormone concentration by sandwich assay method. Luteinizing hormone in serum was determined using the Luteinizing hormone reagent, which utilizes the sandwich assay method using Cobas e411 automated analyser. A volume of 20L of serum was aspirated by the analyzer and assayed for luteinizing hormone concentration. The serum was incubated with both a biotinylated, monoclonal 101 University of Ghana http://ugspace.ug.edu.gh LH-specific antibody and a ruthenylated, monoclonal LH-specific antibody to form a sandwich complex. Streptavidin-coated microparticles were added to the reaction mixture and the complex bound to the solid phase via biotin-streptavidin interactions. The reaction mixture was transferred to a measuring cell and the microparticles were magnetically captured onto the surface of an electrode; unbound sample was washed away before a chemiluminescent reaction is induced by applying a voltage to the electrode. Chemiluminescence was measured by a photomultiplier and the concentration of LH within the sample is calculated using a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode. 3.9.39 Principle Procedure of follicle‐stimulating hormone (FSH). 3.9.39.1 Required Materials, Reagents and Equipment. Follicle‐ stimulating hormone reagent cartridge, Streptavidin-coated microparticles, Anti- FSH-Ab~biotin (gray cap), Biotinylated monoclonal anti‐ FSH antibody, MES buffer, Anti-FSH-Ab~Ru(bpy), Monoclonal anti‐ FSH antibody labelled with ruthenium complex, Procell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.39.2 Determination of serum Follicle‐stimulating hormone concentration by sandwich assay method. Follicle‐ stimulating hormone in serum was determined using the sandwich assay method with Cobas e411 automated analyser. A volume of 40L of serum was aspirated by the analyzer and assayed for Follicle‐ stimulating hormone concentration. The serum and a biotinylated monoclonal FSH‐ specific antibody, and a monoclonal FSH‐ specific antibody labelled with a ruthenium complex form a sandwich complex. After addition of 102 University of Ghana http://ugspace.ug.edu.gh streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture is aspirated into the measuring cell where the microparticles are magnetically captured onto the surface of the electrode. Unbound substances were then removed with ProCell reagent. Application of a voltage to the electrode then induces chemiluminescent emission which is measured by a photomultiplier. Results were determined via a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode. 3.9.40 Principle Procedure of Prolactin. 3.9.40.1 Required Materials, Reagents and Equipment. Prolactin II reagent cartridge, Streptavidin-coated microparticles, Anti-prolactin- Ab~biotin (gray cap), Biotinylated monoclonal anti‐prolactin antibody, phosphate buffer, Anti-prolactin-Ab~Ru(bpy), monoclonal anti‐prolactin antibody labelled with ruthenium complex, ProCell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.40.2 Determination of serum Prolactin hormone concentration by sandwich assay method. Prolactin in serum was determined using the sandwich assay method with Cobas e411 automated analyser within a duration of 18 minutes. A volume of 20 μl of participants’ serum was aspirated by the cobas e411 analyzer and assayed for prolactin concentration. The serum and a biotinylated monoclonal prolactin- specific antibody form a first complex. After addition of a monoclonal prolactin‐ specific antibody labelled with a ruthenium complex and streptavidin-coated microparticles, a sandwich complex was 103 University of Ghana http://ugspace.ug.edu.gh formed and bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture was aspirated into the measuring cell where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances were then removed with ProCell. Application of a voltage to the electrode then induces chemiluminescent emissions which was measured by a photomultiplier. Results were determined via a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode or e‐ barcode. 3.9.41 Principle Procedure of Testosterone. 3.9.41.1 Required Materials, Reagents and Equipment. Testosterone II reagent cartridge, Streptavidin-coated microparticles, Anti-testosterone- Ab~biotin, Biotinylated monoclonal anti-testosterone antibody releasing reagent 2- bromoestradiol; MES buffer, Testosterone-peptide~Ru(bpy), Testosterone derivative, labelled with ruthenium complex, ProCell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.41.2 Determination of serum Testosterone concentration by competition assay method. Testosterone in serum was determined using the competition assay method with cobas e411 automated analyser within a duration of 18 minutes. A volume of 20 μl of participants’ serum was aspirated by the Cobas e411 analyzer and assayed for prolactin concentration. The serum was incubated with a biotinylated monoclonal testosterone- specific antibody. The binding sites of the labelled antibody become occupied by the sample analyte (depending on its concentration). After addition of streptavidin-coated microparticles and a testosterone derivate labelled with a ruthenium complex, the complex 104 University of Ghana http://ugspace.ug.edu.gh becomes bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture was aspirated into the measuring cell where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances were then removed with ProCell (reagent). Application of a voltage to the electrode then induces chemiluminescent emission which was measured by a photomultiplier. Results were determined via a calibration curve which was instrument specifically generated by 2-point calibration and a master curve provided via the reagent barcode or e-barcode. 3.9.42 Principle Procedure of Thyroid‐stimulating hormone (TSH). 3.9.42.1 Required Materials, Reagents and Equipment. TSH reagent cartridge, Streptavidin-coated microparticles, Anti-TSH-Ab~biotin, Biotinylated monoclonal anti-TSH antibody, phosphate buffer, Anti-TSH-Ab~Ru(bpy), Monoclonal anti-TSH, ProCell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.42.2 Determination of serum TSH concentration by sandwich assay method. TSH in serum was determined using the sandwich assay method with Cobas e411 automated analyser within a duration of 18 minutes. A volume of 50 μl of participants’ serum was aspirated by the Cobas e411 analyzer and assayed for TSH concentration. 50 μl of serum, a biotinylated monoclonal TSH‐ specific antibody and a monoclonal TSH‐ specific antibody labelled with a ruthenium complex react to form a sandwich complex. After addition of streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture was aspirated into the measuring cell where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances are then removed with ProCell. Application 105 University of Ghana http://ugspace.ug.edu.gh of a voltage to the electrode then induces chemiluminescent emission which is measured by a photomultiplier. Results were determined via a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode or e‐ barcode. 3.9.43 Principle Procedure of Ferritin. 3.9.43.1 Required Materials, Reagents and Equipment. Ferritin reagent cartridge, Streptavidin-coated microparticles, Anti‐ Ferritin‐ Ab~biotin (gray cap), Biotinylated monoclonal anti‐ ferritin antibody, phosphate buffer, Anti‐ ferritin‐ Ab~Ru(bpy), Monoclonal anti‐ ferritin antibody labelled with ruthenium complex ProCell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.43.2 Determination of serum Ferritin concentration by sandwich assay method. Ferritin in serum was determined using the Sandwich assay method with Cobas e411 automated analyser within a duration of 18 minutes. A volume of 10µl of participant’s serum was aspirated and assayed for ferritin concentration. The serum reacts with a biotinylated monoclonal ferritin‐ specific antibody, and a monoclonal ferritin‐ specific antibody labelled with a ruthenium complex form a sandwich complex. After addition of streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture was aspirated into the measuring cell where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances were then removed with ProCell. Application of a voltage to the electrode then induces chemiluminescent emission which was measured by a photomultiplier. Results were determined via a calibration curve which was instrument- 106 University of Ghana http://ugspace.ug.edu.gh specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode or e‐ barcode. 3.9.44 Principle Procedure of Prostate‐Specific Antigen (PSA). 3.9.44.1 Required Materials, Reagents and Equipment. PSA reagent cartridge, Streptavidin-coated microparticles, Anti-PSA-Ab~biotin (gray cap), Biotinylated monoclonal anti‐ PSA antibody, phosphate buffer, Anti-PSA- Ab~Ru(bpy), Monoclonal anti‐ PSA antibody labelled with ruthenium complex ProCell reagent and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan). 3.9.44.2 Determination of serum PSA concentration by sandwich assay method. PSA in serum was determined using the Sandwich assay method with Cobas e411 automated analyser within a duration of 18 minutes. A volume of 20µl of participant’s serum was aspirated by the analyzer and assayed for PSA concentration. The serum reacts with a biotinylated monoclonal PSA‐ specific antibody, and a monoclonal PSA‐ specific antibody labelled with a ruthenium complex react to form a sandwich complex. After addition of streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin. The reaction mixture was aspirated into the measuring cell where the microparticles were magnetically captured onto the surface of the electrode. Unbound substances were then removed with ProCell. Application of a voltage to the electrode then induces chemiluminescent emission which was measured by a photomultiplier. Results were determined via a calibration curve which was instrument- specifically generated by 2‐ point calibration and a master curve provided via the reagent barcode or e‐ barcode. 107 University of Ghana http://ugspace.ug.edu.gh Other analytes such as Alpha 1-Fetoprotein, C3 Complement, C4 Complement, CA 125, Carcinoembryonic Antigen, Cortisol in serum, Folic acid, Immunoglobulin A in serum, Immunoglobulin G in serum, Immunoglobulin M in serum, Insulin, Thyroperoxidase abs, Parathyroid hormone, Thyroglobulin abs, Transferrin in serum, Vitamin B12, which were not under Medlab Ghana accreditation were shipped to Synlab in Germany for analysis. The samples of each of the participants placed in 2mL cryovial tubes stored in 5”X5” holding boxes. Each set of participants’ samples were covered with an absorbent material (paper towel), taped tightly to secure such that there is enough absorbent material to absorb any liquid or leaked samples. All the wrapped boxes were placed in a Ziploc biohazard bag labelled with each participant barcode and study number for easy identification. All the samples of each participants in the Ziploc biohazards bag were placed in a Styrofoam box pack with dry ice at a temperature of -80 oC to preserved the samples. All the analytes sent to synlab were analysed using Centaur XP Siemens (Siemens, Bayer, Germany), except Vitamin D (25-OH) and Somatotrope Hormone which were analysed using Qtrap Api 4500 ABSCIEX (Bayer, Germany), and Immulite 2000 Xpi Siemens analyser (Siemens, Bayer, Germany) respectively. 3.10 Quality Control Before any data collection, a one-day training session was held with both research assistants and study participants on data collection, data quality, data accountability, and data confidentiality. A pilot study with questionnaire was done with 20 volunteers of the Medlab Ghana staff and all the challenges concerning questionnaire ambiguity and other specific questions were addressed appropriately. A database was developed in Microsoft Access database (Access 2010), which was used to double-check all errors and logistic discrepancies for data validation. 108 University of Ghana http://ugspace.ug.edu.gh 3.10.1 Laboratory Quality Control. All laboratory analyses were carried out in accordance with the laboratory’s standard operating procedures (SOPs) and all equipment calibration were done according to manufacturers’ specification. Manufacturers manuals and reagent/kit package inserts requirements were adhered to in order to ensure good laboratory practices and procedures. 3.10.2 Mini-panel. For the purpose of this study a mini panel of sera was prepared from healthy volunteers consisting of three males and two females. These sera were measured each day to monitor the stability of the assay over the study period, in order to assess between-day variations of test results. This was done for all assays including those that were analysed at Synlab, Germany. This quality control procedure was used to identify any unexpected bias that occurred between testing days. This allows for the re-calibrating of the particular assay with the manufacturer’s calibrator or check for some other abnormality on the analyser that could be causing the shift in repeatability. 3.11 Data Management Data were entered into Microsoft Access database (Access 2010). Standardized queries were used to conduct range and logic checks, and discrepant entries were rectified after a review of collected data. Data management techniques used involved sorting, merging and data reorganization to suit the statistical technique used for analysis. After a double data entry, comparison of two datasets was done to delete duplicate. String variables such as sex, ethnicity, physical execerise etc were coded into numeric where necessary. 109 University of Ghana http://ugspace.ug.edu.gh 3.12 Statistical Analysis The questionnaire information (anthropometric measurement) was entered manually using Microsoft Accessdatabase (Access 2010) and exported to Microsoft Excel 2016. The clinical laboratory results were retrieved from the stored data in the Sysmex XN 1000 (Sysmex Corp, Kobe, Japan) auto-analyzer, Beckman Coulter AU 480 Chemistry Analyzer (Beckman Coulter, Tokyo, Japan) auto- analyzer, and Roche Cobas e411 analyser (Roche Cobas, Tokyo, Japan) software transmitted to schuynet Medlab Ghana. The laboratory results were downloaded from schuynet (an internet module that allows easy access to patient laboratory results on-line from any computer equipped with a web browser) and manually entered into Microsoft Excel. The statistical analysis conducted in this study were in four parts: ● Descriptive analysis of the demographics of study participants with Stata version 13 software (Stata Corp., College Station, Texas, United States) ● Multiple regression analysis (MRA) was performed to identify source of variations of Reference Values by StatFlex version 6.0 statistical software (Artech Inc., Osaka, Japan). ● Partitioning of reference values by sex and age was judged by the computation of standard deviation ratio (SDR) using 3-level nested ANOVA with StatFlex version 6.0 statistical software (Artech Inc., Osaka, Japan). ● Derivation of reference intervals was done by “Reference Interval Master” originally developed by Ichihara (Ichihara, 2010). 3.12.1 Descriptive Analysis. Descriptive analysis to summarize the demographics, lifestyle, and nutritional pattern of the study participants was performed. Stata version 13 software (Stata Corp., College 110 University of Ghana http://ugspace.ug.edu.gh Station, Texas, United States) was used in analysing the simple descriptive statistics for the demographics of study participants. Frequencies, median, and mean were used to present the demographic results. 3.12.2 Data Regression. As part of this study the possible factors influencing the reference values were of interest and multiple regression analysis (MRA) was performed to identify such factors. Analysis of sources of variation is essential to assess the need for RIs partitioning. Multiple regression analysis was adopted in exploring the source of variation of reference values because the source of variation of test results varies, and can be mutually correlated. Multiple regression analysis was performed separately for each sex by setting RVs of each analyte as a dependent variable. Variables such as age, sex, ethnicity (Akan=1 and Non- Akan=1), BMI, blood pressure (SBP and DBP), alcohol consumption, smoking, exercise were the primary independent variables. Furthermore, physical exercise, cigarette smoking, and alcohol consumption were also analysed to ascertain their influence on the results. For alcohol calculation: none = 0; less than 25g Ethanol/day=1; 25~50g=2; 50~100g=3; 100g~=4 whiles physical exercise intensity was also categorized as: none=0; less intense exercise like walking =0.5 ; jogging = 1 and high intensity exercise like running and football = 2. All exercise done for 30 mins or more for 1~7day =1~7. The level of exercise was calculated as the number of times exercised with in the week by the denoted exercise number. For example jogging 30 mins 3 times per week is calculated as (3×1=3), walking 6days/week, the level is (6 × 0.5 = 3). A given explanatory variable was considered to be of practical importance when its standardized partial regression coefficient, which corresponds to the partial correlation coefficient (rp), was  0.20. 111 University of Ghana http://ugspace.ug.edu.gh 3.12.3 Partitioning Criteria. The data was partitioned according to the magnitude of variation of test results and this was determined by specific factor being expressed as standard deviation (SD). With age, gender, ethnicity as variables of interest, the magnitude of between-age SD (SD-age), between-gender SD (SD-sex), between-ethnicity SD (SD-ethnicity) and net-between individual SD (SD-indiv) were calculated using a 3-level nested ANOVA as standard deviation ratio (SDR-reg, SDR-age, and SDR-sex) (Ichihara, 2014; Ozarda et al., 2013). In this study, a SDR  0.40 was used as a guide to consider partitioning reference values by the sources of variation (SV) (Ichihara, 2014; Ichihara & Boyd, 2010). Hence, any of the SDR that was  0.40 was an indication of the presence of significant difference and therefore was subjected to partitioning for their RI. Moreover, in the analysis, age was stratified as (18–29, 30–39, 40–49, 50–59, and 60+ years); ethnicity was stratified as (Akan and Non-Akan); sex was male and females. 3.12.4 Latent abnormal values exclusion (LAVE) method. One important aspect of the data analyses was the detection and removal of outliers because they could easily confound the RI results. In this study, even though the exclusion criteria were strictly adhered to, it was inevitable that some individuals with latent diseases of common occurrence such as latent anaemia of menstrual/nutritional origin, subclinical infections, metabolic syndrome and inflammation were included in non-negligible number. Therefore, prior to derivation of the RIs, the latent abnormal values exclusion (LAVE) method was used to exclude individuals with such latent diseases. The latent abnormal values exclusion (LAVE) method is a secondary exclusion procedure applied to refine the data (Ichihara, 2014; Ichihara & Boyd, 2010). 112 University of Ghana http://ugspace.ug.edu.gh This method is a type of iterative optimization approach for derivation of multiple RIs simultaneously. Initially mutually related analytes for haematology, mainly folate, Vitamin B12, Fe, TF, ferritin, RBC, Hb, Ht, and MCV served as reference test for erythrocytes analytes whiles IgG, IgA, IgM, C3, CRP and ferritin served as reference test for leukocytes. With regards to the chemistry analytes, the reference test used as exclusion criteria were Alb, Glb, Glu, AST, ALT, LDH, GGT, TG, HDL, LDL-C and CRP. RIs are determined for each reference analyte independently, in which no exclusion of values is made in the initial computation of RIs. From the second computation, any individual who has two or more results outside the RIs derived in the previous computation among the reference tests (mutually related analytes) is excluded. The algorithm then used those initial values of RIs to judge the abnormality of each individual’s record by counting the number of abnormal results in tests other than the one for which the RI is being determined. Reference values were excluded only when the concurrent test abnormalities occurred in tests that were significantly correlated with the results of the test for which the RI was being computed. This process was repeated six times at which point the RIs were near stable. According to Ichihara et al. (2017), this method is better than the traditional methods of outlier exclusion. This is because, first, truncation of reference distributions does not occur as it removes individuals whose results are outside RIs of analytes other than the one for which the RI is being derived. Secondly, there is an effect on analytes whose values were rarely outside the RI in healthy individuals (Ichihara, Ozarda, Barth, Klee, Qiu, et al., 2017). 113 University of Ghana http://ugspace.ug.edu.gh 3.12.5 Reference Interval Derivation. In this study, the reference intervals for each analyte was derived using the “Reference Interval Master” originally developed by Ichihara. Both parametric and non-parametric methods were used in deriving the RIs. However, it is worthy to mention that, the non- parametric method was used only for comparison purposes. For the parametric method, the reference values were first transformed to Gaussian distribution by use of the modified Box-Cox power transformation formula (Ichihara, 2014; Ichihara & Boyd, 2010). In determining the mean and SD, the final RI was calculated as the mean ± 1.96SD, which corresponds to the central 95% limits or lower limit (LL) and upper limit (UL) under transformed scale. Then, the limits are reverse transformed to get the LL and UL in the original scale. For the nonparametric method, reference values were first sorted in ascending order, and the 2.5th and 97.5th percentile points were determined as the lower limits and upper limits respectively. The 90% confidence intervals (CIs) for both the lower and upper limits were estimated by use of the bootstrap method through iterative resampling 50 times. For smoothing the RI, average values for LL, median, and UL were adopted as the final RIs. On the other hand, for the nonparametric method, reference values were first sorted in an ascending order, and the 2.5th and 97.5th percentile points were determined. 3.13 Ethical Approval The study protocol was approved by the Ethical Review Committee (CHS-Et/M.8-P 4.14/2016-2017) of the College of Health Sciences, University of Ghana. Written and verbal consent was obtained from all voluntary eligible participants prior to the study. Confidentiality of data was maintained at all times. All abnormal results were discussed with the participants and referred for further tests or to see a medical practitioner where applicable. These results were however not included in the analysis. 114 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR RESULTS 4.0 Introduction This chapter presents the descriptive demographic statistics and the outcome of the study according to the four main objectives set out. Foremost, the reference intervals derived for haematological and clinical chemistry analytes are presented in sections 1 and 2 of this chapter. Subsequently, the sources of variation represented by the standard deviation ratios and multiple regression analysis are presented which takes care of the objective 3. Finally, the results of reference intervals derived with the application of LAVE and without the application of LAVE are presented in the last section. 4.1 Participant Demographics A total of 501 volunteers participated in the study with male to female ratio of 54: 46 (n=270: 231). As shown in Table 4.1 the participants’ mean age was 41.3 ± 13.5 years. In Table 4.2, age was further categorized into age-group as 18−29 years (24.7%), 30−39 years (22.5%), 40−49 years (23.9%), and 50+ years (28.8%). Composition of volunteers were 77.2% from Accra and 22.8% from Tamale. With regard to ethnicity, Akan constituted 40.3% (202) whiles non-Akans were 59.7% (299). The mean BMI of the participants was 25.8 ± 3.9 kg/m2 (range = 18.6−34.9 kg/m2). Regarding participants’ blood pressure, the mean and SD for systolic was 121.2 ± 11.5 mmHg (range = 88-140mmHg) whiles diastolic blood pressure mean and SD were 78.9 ± 7.2 mmHg (60- 90mmHg). About 13% of the participants take alcohol with the vast majority of them only occasionally, and almost all of them (99.4%) were non-smokers. About 13% take alcohol and almost all the participants (99.4%) were non-smokers. Approximately, 37.1% of the participants engage in different level of physical body exercise shown in Table 4.2. 115 University of Ghana http://ugspace.ug.edu.gh Table 4.1:Characteristics of age mean, gender, BMI, SBP and DBP of study participants Variables Observations Mean Standard Minimum Maximum Deviation Age Participants 501 41.3 13.5 18 92 Males 270 40.9 13.5 18 92 Females 231 41.8 13.6 18 90 BMI Participants 501 25.8 3.9 18.6 34.9 Males 270 24.9 3.4 18.6 34.9 Females 231 26.9 4.0 18.6 34.9 SBP Participants 501 121.2 11.5 88 140 Male 270 122.5 10.9 90 140 Females 231 119.6 12.0 88 139 DBP Participants 510 78.9 7.2 60 90 Male 270 78.8 7.0 61 90 Female 231 79.0 7.4 60 89 BMI – Body Mass Index, SBP – Systolic Blood Pressure, DBP – Diastolic Blood Pressure Table 4.2:Distribution of age category with gender, ethnicity and exercise of study participants 18-29 30-39 40-49 50+ Total (n) Percentage (%) Gender Males 76 58 58 78 270 53.9 Female 48 55 62 66 231 46.1 Ethnicity Akan 49 50 54 49 202 40.3 Non-Akan 75 63 66 95 299 59.7 Exercise Yes 40 37 49 60 186 37.1 No 84 76 71 84 315 62.9 116 University of Ghana http://ugspace.ug.edu.gh 4.2 Haematological Analytes Table 4.3 presents the RIs for haematological analytes derived parametrically with LAVE for males and females. It was found that RBC had an RI of 4.05 – 6.43 x109/L (median = 4.99 x109/L) for combined male and female. However, the values for males were higher with 4.57 – 6.50 x109/L than females 4.00 – 5.46 x109/L. Haemoglobin ranged 11.2 – 17.4 g/dL with a median of 13.9 g/dL. Yet, for males, it was 12.8 – 17.2 g/dL with a median of 15.0 g/dL while for females it was 10.7 – 14.3 g/dL with a median of 12.6 g/dL. Similarly, haematocrit was estimated as 35.4 – 52.8% (median = 42.7%). with males recording higher reference values for lower limit (LL) and upper limit (UL) 39.4 - 52.1% compared with reference values of females 34.0 – 44.2%. The reference interval of MCV for males and females respectively were 72.2 – 96.2 fL (median = 85.4 fL) and 71.4 – 98.3 fL (median = 86.6 fL); but the RI for combine male and female was 71.9 – 97.4 fL with a median of 86.0 fL. MCH ranged between 23.2 – 31.8 pg with median = 28.1 pg, here the RIs for both males and females were relatively comparable. Similarly, the estimated RIs of MCHC for both males and females were similar. Meanwhile the RI for both sexes combined was 29.6 – 35.2 g/dL, median = 32.3 g/dL. For RDW, RI was 11.7 – 16.2 %, median = 13.3 % with females recording a relatively higher UL (17.3%) than males (16.0%). The RIs for WBC, MPV and PLT were 3.08 – 7.53 x 109/L (median = 4.75 x 109/L), 9.1 – 12.8 fL (median = 10.7 fL), and 126 – 371 x 109/L (median = 246 x 109/L) respectively. It is worthy of note that the RIs for WBC and MPV were similar between males and females, but for PLT, females recorded significantly higher reference values (157 - 402 x 109/L) than males (115 – 339 x 109/L). The reference intervals for Neutrophils (%), Neutrophils (#), Lymphocytes (%) and Lymphocytes (#) were also estimated as 29.0 – 62%, 1.17 – 3.81 x109/L, 27 – 60.0%, and 1.22 – 3.38 x109/L respectively. Comparable reference values were recorded between males and females. 117 University of Ghana http://ugspace.ug.edu.gh Moreover, the results show that the estimated values for Monocytes (%), Monocytes (#), Eosinophils (%), Eosinophils (#), Basophils (%) and Basophils (#) are 5.0 – 13.7%, 0.22 – 0.73 x109/L, 0.40 – 8.7%, 0.02 – 0.39 x109/L, 0.20 – 1.50%, 0.01 – 0.08 x109/L respectively. The respective medians are: 8.5%, 0.40 x109/L, 2.10%, 0.10 x109/L, 0.60%, and 0.03 x109/L. Interestingly, it was revealed that except for Eosinophils (%) and Eosinophils (#) which recorded significant differences in the upper limits (males = 0.50 – 10.30% vs females = 0.40 – 6.50%) and (males = 0.03 – 0.53 x 109/L, vs 0.02 – 0.29 x 109/L,) between males and females all the other analytes had similar RIs (see Table 4.3). Table 4.3: Reference intervals for haematological analytes derived parametrically TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL RBC 1012/L 419 4.05 4.99 6.43 232 4.57 5.34 6.50 181 4.00 4.54 5.46 Hemoglobin g/dL 425 11.2 13.9 17.4 239 12.8 15.0 17.2 188 10.7 12.6 14.3 Hematocrit % 426 35.4 42.7 52.7 239 39.4 45.8 52.1 191 34.0 39.1 44.2 MCV fL 428 71.9 86.0 97.4 238 72.2 85.4 96.2 185 71.4 86.6 98.3 MCH pg 419 23.2 28.1 31.8 232 23.6 28.1 31.9 181 22.3 27.7 31.7 MCHC g/dL 485 29.0 32.3 35.2 263 29.8 32.7 35.5 222 28.6 31.9 34.8 RDW % 428 11.7 13.3 16.2 235 11.7 13.2 16.0 187 12.0 13.6 17.3 MPV fL 475 9.1 10.7 12.8 225 9.0 10.6 12.6 220 9.3 10.7 13.0 Platelets 109/L 483 126 246 371 262 115 228 339 222 157 268 402 WBC 109/L 483 3.08 4.75 7.53 263 2.86 4.67 7.43 221 3.32 4.81 7.69 Neutrophils % % 427 29 44 62 216 28 43 61 197 29 45 61 Neutrophils # 109/L 413 1.17 2.03 3.81 210 1.17 1.95 3.72 189 1.21 2.10 3.69 Lymphocytes % % 485 27 44 60 264 27 43 60 221 26 43 60 Lymphocytes # 109/L 485 1.22 2.01 3.38 263 1.20 1.97 3.38 222 1.27 2.06 3.46 Monocytes % % 485 5.0 8.5 13.7 263 4.9 8.8 13.8 221 5.1 8.1 13.4 Monocytes # 109/L 484 0.22 0.40 0.73 263 0.22 0.41 0.74 221 0.23 0.39 0.72 Eosinophils % % 481 0.4 2.1 8.7 262 0.5 2.4 10.3 219 0.4 1.8 6.5 Eosinophils # 109/L 410 0.02 0.10 0.39 209 0.03 0.12 0.53 188 0.02 0.09 0.29 Basophils % % 485 0.2 0.6 1.5 264 0.2 0.6 1.5 222 0.2 0.6 1.5 Basophils # 109/L 485 0.01 0.03 0.08 263 0.01 0.03 0.08 222 0.01 0.03 0.08 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 118 University of Ghana http://ugspace.ug.edu.gh 4.3 Biochemical Analytes Table 4.4 presents the reference intervals (RIs) for clinical chemistry analytes that were parametrically derived with the application of LAVE. The table is sub-titled and grouped into functional tests that these analytes measure. For calcium (Ca) and corrected calcium (cCa+) their respective RIs were 2.21 - 2.57 mmol/L, median = 2.39 mmol/L and 2.20 - 2.47 mmol/L, median = 2.31 mmol/L. The results of cCa+ for males and females are similar (see table 2) however for Ca males recorded relatively higher LL and higher UL (2.29 - 2.62 mmol/L, median = 2.41 mmol/L) compared with females (2.2 - 2.54 mmol/L, median = 2.35 mmol/L). The RI for magnesium (Mg) was 0.75 - 1.04 mmol/L with results of both sexes relatively similar (males = 0.75 - 1.03 mmol/L, median = 0.88 mmol/L; females = 0.76 - 1.04 mmol/L, median = 0.88 mmol/L). On the other hand, Phosphate (P) had relatively both higher LL and UL among females (0.98 - 1.93 mmol/L, median = 1.25 mmol/L) than males (0.84 - 1.59 mmol/L, median = 1.16 mmol/L). For the combined RI of phosphate is 0.90 - 1.73 mmol/L with a median of 1.20 mmol/L. Creatine Kinase (CK) recorded an estimated RI of 63 – 414 U/L with a median of 174 U/L. Obviously, it was found that males had a significantly wider range (93 – 502 U/L, median = 207 U/L) than females (52 – 276 U/L, median = 137 U/L); this reflected in both higher LL and higher UL. The RI of LDH was also found to be 124 – 278 IU/L with a median of 194 IU/L. Females had higher reference values (134 – 296 IU/L) compared with males (124 – 278 IU/L). Amylase (AMY) on the other hand recorded an RI of 46 – 174 U/L and median of 94 U/L; here males recorded relatively wider RI particularly in the UL (47 – 177 U/L, median = 101 U/L) compared with females (43 – 158 U/L, median = 86 U/L). In addition, it was found that Uric Acid (UA) had RI of 168 – 470 umol/L and 119 University of Ghana http://ugspace.ug.edu.gh median of 295 umol/L. Significant sex variation was observed with males recording considerably higher references values (231 – 487 umol/L, median = 335 umol/L) than females (149 – 377 umol/L, median = 248 umol/L). In C-Reactive Protein (CRP) it was found that RI is 0.22 – 7.98 mg/L with a median of 1.10 mg/L; contrastingly, females presented a higher LL and higher UL (0.47 – 9.85 mg/L, median = 1.32 mg/L) that was twice more than LL and UL of males (0.18 - 4.94 mg/L, median = 0.86 mg/L). C3 complement and C4 complement recorded RIs of 84 – 168 mg/dL, median = 120 mg/dL and 15 – 50 mg/dL, median = 30 mg/dL respectively. For C3, females presented higher LL and higher UL (93 – 172 mg/dL) than males (82 – 164 mg/dL). The same was observed in C4 where females presented with RI of 17 – 53 mg/dL as against 14 – 46 mg/dL in males. Table 4.4:Reference intervals derived parametrically for clinical chemistry analytes TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL Calcium mmol/L 485 2.21 2.39 2.57 261 2.26 2.41 2.60 222 2.20 2.35 2.54 Corrected Ca+ mmol/L 483 2.20 2.31 2.47 261 2.21 2.31 2.47 221 2.19 2.31 2.46 Magnesium mmol/L 486 0.75 0.88 1.04 264 0.75 0.88 1.03 222 0.76 0.88 1.04 Phosphate mmol/L 480 0.90 1.20 1.73 261 0.84 1.16 1.59 219 0.98 1.25 1.93 CK U/L 484 63 174 414 263 93 207 502 219 52 137 276 LDH IU/L 485 129 197 286 263 124 194 278 222 134 199 296 AMY U/L 483 46 94 171 263 47 101 177 220 43 86 158 Uric Acid umol/L 486 168 295 470 264 231 335 487 222 149 248 377 C-Reactive Prot. mg/dL 417 0.22 1.10 7.98 233 0.18 0.87 4.94 188 0.47 1.32 9.85 C3 Complement mg/dL 485 84 121 168 264 82 113 164 222 93 128 172 C4 Complement mg/dL 484 15 30 50 262 14 28 46 222 17 33 53 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 120 University of Ghana http://ugspace.ug.edu.gh 4.3.1 Kidney Function Tests. Table 4.5 below presents the reference intervals (RIs) for kidney function test. Urea recorded relatively similar RIs for males and females; 2.35 - 5.74 mmol/L, median = 3.61 mmol/L and 2.01 - 5.47 mmol/L, median = 3.2 mmol/L. The combined RI is 2.17 - 5.63 mmol/L and median of 3.43 mmol/L. Creatinine recorded an estimated RI of 43 – 107 umol/L and median of 69 umol/L with males recording significantly wider interval (58 – 109 umol/L, median = 80 umol/L) than females (41 – 82 umol/L, median = 59 umol/L). The RI of eGFR was 79 – 188 mL/min, median = 125, with females recording wider interval (83 – 205 mL/min, median = 129) compared with males (79 – 179 mL/min, median = 120). The analytes sodium (Na), potassium (K) and chlorine (Cl) had RIs of 136 – 143 mmol/L, 3.5 - 5.0 mmol/L, 99 – 108 mmol/L respectively. It is noteworthy that, for these analytes, the RIs between males and females were comparable (see Table 4.5). It was also found that total carbon dioxide (TCO2) had an estimated RI of 18.0 - 27.2 mmol/L with a median of 22.6 mmol/L; with males recording 18.9 – 27.9 mmol/L and females recording 17.4 – 26.1 mmol/L. The RI for Anion GAP (ANGAP) was 8.5 – 21.2 mmol/L with comparable results between males and females. Table 4.5:Reference intervals derived parametrically for Kidney function Tests TEST UNITS MALE+FEMALE MALE FEMALE N LL Me UL n LL Me UL n LL Me UL Urea mmol/L 485 2.17 3.43 5.63 263 2.35 3.61 5.74 222 2.01 3.22 5.47 Creatinine umol/L 486 43 70 107 263 58 80 109 221 41 59 82 eGFR mL/min 485 79 124 188 263 79 120 179 222 83 129 205 Sodium mmol/L 483 136 139 143 262 136 139 143 221 135 139 143 Potassium mmol/L 483 3.5 4.2 5.0 263 3.4 4.2 5.0 221 3.5 4.2 4.9 Chloride mmol/L 485 99 103 108 263 98 102 107 221 100 104 108 Total CO2 mmol/L 486 18.0 22.8 27.2 264 18.9 23.2 27.9 222 17.4 22.1 26.1 ANGAP mmol/L 485 8.5 12.9 21.2 264 8.7 13.1 21.9 222 8.2 12.5 20.2 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 121 University of Ghana http://ugspace.ug.edu.gh 4.3.2 Liver Function Tests. Table 4.6 below presents the reference intervals (RIs) for clinical chemistry analytes, specifically liver function tests; the results indicate that total protein (TP) was 65.8 – 83.7g/L, median = 74; with similar RIs recorded between the sexes. The estimated RI for Albumin (Alb) was 37.7 – 48.3g/L; however, males had a relatively shorter interval with a higher LL (39.2g/L) compared with 36.9g/L for females; with a difference in the upper limits too (males = 48.9g/L vs females = 46.2g/L). Furthermore, Globulin recorded an RI of 23.6 - 39.6g/L with a median of 30.9g/L. The main difference between RIs of the sexes was that the females had a relatively higher LL and UL (24.3 - 40.9g/L) as against that of males (23.1 - 38.8g/L). Total bilirubin (Tbil) had an RI of 5.9 - 32.7umol/L, median = 11.3 umol/L; with males recording a wider interval (6.9 - 35.8umol/L, median = 14.0 umol/L) compared with females (5.3 - 23.9umol/L, median = 9.5 umol/L). Similarly, Direct bilirubin (Dbil) recorded some difference between males and females (1.4 - 6.5 umol/L, median = 2.7 umol/L) compared with (0.9 – 4.8 umol/L, median = 1.9 umol/L). However, the combined RI is 1.0 – 5.8 umol/L with median of 2.3 umol/L. AST and ALT RI results are 16 – 37 IU/L and 8 – 42 IU/L respectively; with significant differences between males and females. Thus, for AST, males recorded RI of 17 - 39 IU/L compared with 15 - 27 IU/L for females. Similarly, among males, the RI for ALT recorded an UL which is almost twice (50 IU/L) as high as the UL of females (26 IU/L). In contrast, ALP which had an estimated RI of 39 – 111 IU/L also recorded comparable results between males and females (40 – 111 IU/L and 38 – 112 IU/L). GGT had an RI of 14 – 67 IU/L for the combined sex but significantly contrasting results for males and females. The UL of the males’ RI (87 IU/L) was almost twice the UL of the females’ RI (49 IU/L) even though their LLs were closer (17 IU/L vs 12 IU/L). 122 University of Ghana http://ugspace.ug.edu.gh Table 4.6:Reference intervals derived parametrically for Liver function Tests TEST UNITS MALE+FEMALE MALE FEMALE N LL Me UL n LL Me UL n LL Me UL Total Protein g/L 484 65.3 73.9 83.7 264 65.8 74.6 83.8 221 65.1 73.9 83.5 Albumin g/L 486 37.7 43.0 48.3 264 39.2 44.0 48.9 222 36.9 41.5 46.2 Globulin g/L 485 23.6 30.9 39.6 264 23.1 30.5 38.8 222 24.3 31.4 40.9 Total Bilirubin umol/L 481 5.9 11.7 32.7 262 6.9 14.0 35.8 220 5.3 9.5 23.9 Direct Bilirubin umol/L 480 1.0 2.3 5.8 261 1.4 2.7 6.5 220 0.9 1.9 4.8 AST IU/L 367 16 22 37 207 17 24 39 176 15 20 27 ALT IU/L 368 8 17 42 207 9 21 50 178 7 14 26 ALP IU/L 483 39 67 111 263 40 67 111 220 38 66 112 GGT IU/L 370 14 28 67 211 17 32 87 176 12 23 49 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.3 Diabetes and Lipids Analytes The analytes considered for diabetes screen are glucose and HbA1c%. As shown in Table 4.7 results indicate RIs of 3.94-5.95 mmol/L and 4.24 - 6.27 mmol/L for glucose and HbA1c% respectively. Among these two analytes, it was found that females recorded relatively higher reference values in both LLs and ULs compared to males (Table 4.7). Total cholesterol HDL-cholesterol, and triglycerides recorded RIs of 3.54 – 7.51 mmol/L, 0.86 – 2.01 mmol/L and 0.44 – 1.91 mmol/L respectively. With regards to HDL_cholesterol, females recorded higher reference values (0.94 - 2.03 mmol/L) as against males (0.83 – 1.89 mmol/L). Contrariwise, males had higher reference values (0.43 – 1.95 mmol/L) than females (0.41 – 1.64 mmol/L) for triglycerides. In addition, LDL-cholesterol had RI of 2.01 - 5.62 ratio with a median of 3.41 ratio. The RIs of CHOL_HDL and HDL_LDL were 2.54 - 6.55 ratio and 0.20 - 0.76 ratio 123 University of Ghana http://ugspace.ug.edu.gh respectively. The CHOL_HDL among males had a wider range (2.51 - 6.75 ratio) compared with RI for females (2.58 – 5.97 ratio). On the other hand, HDL_LDL had comparable LL between males and females, yet the UL for males was higher (0.78 ratio) than females (0.73 ratio). Table 4.7:Reference intervals for Diabetes and Lipids Analytes TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL N LL Med UL Glucose, fasting mmol/L 481 3.94 4.86 5.95 261 3.86 4.95 5.98 219 4.03 4.76 5.82 HbA1c% % 483 4.24 5.37 6.27 262 4.19 5.39 6.25 221 4.37 5.33 6.34 Total Cholesterol mmol/L 357 3.54 5.15 7.51 204 3.39 5.16 7.41 174 3.56 5.05 7.52 HDL-C mmol/L 358 0.86 1.28 2.01 204 0.83 1.23 1.89 174 0.94 1.36 2.03 Triglycerides mmol/L 366 0.44 0.80 1.91 207 0.43 0.87 1.95 179 0.41 0.75 1.64 LDL-C Ratio 357 2.01 3.41 5.62 204 1.86 3.47 5.54 173 1.94 3.33 5.48 CHOL_HDL mg/L 358 2.54 3.96 6.55 204 2.51 4.15 6.75 173 2.58 3.70 5.97 HDL_LDL mg/dL 358 0.20 0.38 0.76 204 0.20 0.35 0.78 173 0.21 0.40 0.73 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.4 Immunoglobulins. In this study, three immunoglobulins were considered, IgG, IgA, and IgM. The RIs of these analytes were 1165 – 2164 mg/dL, 114 – 417 mg/dL, and 41 – 237 mg/dL respectively. Even though the LLs of IgG were comparable between males and females, UL was significantly higher in females (2178 mg/dL) than males (2157 mg/dL), as shown in Table 4.8. In both IgA and IgM females presented wider interval (108 – 419 mg/dL) and (51 – 262 mg/dL) respectively compared to males (119 – 411 mg/dL) and (38 – 195 mg/dL). 124 University of Ghana http://ugspace.ug.edu.gh Table 4.8:Reference intervals for Immunoglobulins TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL N LL Me UL n LL Me UL IgG mg/dL 483 1165 1593 2164 262 1163 1594 2157 222 1166 1581 2178 IgA mg/dL 485 114 223 417 221 119 222 411 221 108 225 419 IgM mg/dL 482 41 102 237 262 38 88 195 220 51 120 262 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.5 Iron and Vitamins Analytes. Table 4.9 below presents the reference intervals (RIs) for Iron, ferritin, and transferrin as the three main analytes derived in this study. Here, it was found that iron has an RI of 7.2 – 26.8 umol/L with a median of 14.5 umol/L. Males presented higher LL and higher UL (8.6 – 28.5 umol/L, median = 16.1 umol/L) compared with that of females (5.5 - 22.4 umol/L, median = 12.4 umol/L). More interestingly, ferritin had an RI of 23 – 489 ng/mL with a median of 129 ng/mL but there was a sharp contrast between males and females. Thus, while females had RI of 15 – 300 ng/mL with a median of 76 ng/mL, males recorded RI of 39 – 502 ng/mL with a median of 168 ng/mL. More so, transferrin had an estimated RI of 196 – 314 mg/dL, with males and females presenting comparable LLs but a considerably higher UL (345 mg/dL) in females than males (308 mg/dL). Furthermore, Vitamin B12 and Folate recorded RIs of 305 – 1283 pg/mL and 2.0 - 15.6 ng/mL respectively. For Vit B12, females presented higher LL and higher UL (332 – 1330 pg/mL) compared with that of males (289 – 1199 pg/mL). Folate on the other hand, recorded almost the same LL (2.0 ng/mL vs 2.1 ng/mL) between the sexes but females had significantly higher UL (19.3ng/mL) than males (14.1 ng/mL) as indicated in table 4.9. An RI of 14.4 – 38 ug/L was recorded for Vitamin D with comparable results recorded between males and females (Table 4.9). 125 University of Ghana http://ugspace.ug.edu.gh Table 4.9:Reference intervals for Iron and Vitamins TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL Iron umol/L 422 7.2 14.5 26.8 234 8.6 16.1 28.5 187 5.5 12.4 22.4 Ferritin ng/mL 429 23 129 489 238 39 168 502 182 15 79 300 Transferrin mg/dL 425 196 246 314 236 195 245 308 186 199 248 345 Vitamin B12 pg/mL 485 305 649 1283 262 289 611 1199 222 332 691 1330 Folate ng/mL 478 2.0 6.2 15.6 263 2.1 5.9 14.1 218 2.0 6.7 19.3 Vitamin D ug/L 483 14.4 25 38 262 14.7 26 39 221 14.2 25 37 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.6 Endocrine Parameters. Table 4.10 below presents reference intervals for hormones. Female specific hormones prolactin, estradiol, LH, FSH, progesterone recorded RIs of 4.3-41.1 ng/mL, 32 -1262 pmol/L, 1.3 - 97 mIU/mL, 1.1 - 167 mIU/mL and 0.0 – 79 nmol/L respectively. Testosterone which is a male-specific hormone had an RI of 10 – 41.9 nmol/L. Meanwhile, cortisol recorded RI of 37 – 174 ng/mL with males presenting considerably higher reference values 47 – 183 ng/mL than females 31 – 141 ng/mL. The analytes parathyroid hormone, somatotrope hormone and insulin recorded RIs of 17 – 94 pg/mL, 0.04 – 3.65 ng/mL and 2.0 – 18.2 uU/mL respectively. For parathyroid and somatotrope hormones, the LLs were comparable between males and females, yet the ULs were considerably higher in females than males. 126 University of Ghana http://ugspace.ug.edu.gh Table 4.10:Reference intervals for Endocrine (Hormones) TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL Prolactin ng/mL 216 4.3 11.2 41.1 Estradiol pmol/L 219 32 241 1262 LH mIU/mL 217 1.3 11 97 FSH mIU/mL 219 1.1 11.8 167 Progesterone nmol/L 219 0.0 1.7 79 Testosterone nmol/L 264 10 24.5 41.9 Cortisol ng/mL 481 37 86 174 263 47 101 183 220 31 71 141 PTH pg/mL 482 17 41 94 262 18 38 84 221 16 46 101 STH ng/mL 486 0.04 0.13 3.65 264 0.04 0.09 2.47 222 0.05 0.19 4.51 Insulin uU/mL 479 2.0 6.5 18.2 260 2.0 5.6 18.4 219 2.1 7.6 17.8 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.7 Thyroid Analytes. Thyroid stimulating hormone (TSH) reference values ranged from 0.55 uIU/mL to 3.66 uIU/mL with female recording relatively higher UL (4.0 uIU/mL) than males (3.39 uIU/mL) as shown in Table 4.11. The reverse is true for the LLs where females recorded a lower value (0.46 uIU/mL) than males (0.57 uIU/mL). Free T3 and Free T4 were 3.81 – 6.19 pmol/L and 12.6 – 20.7 pmol/L respectively. Again, RIs for thyroperoxidase recorded relatively higher in the upper limits among females (28 - 114 U/mL) than males (29 - 57.0 U/mL), with similar values for their lower limits. This similar trend was observed in thyroglobulin with females upper limit of RI being three times higher than that of the males upper limit, with relatively similar lower limits, (15.1 - 91 U/mL) and (15.7 - 910 U/mL) 127 University of Ghana http://ugspace.ug.edu.gh Table 4.11: Reference intervals for Thyroid Analytes TEST UNITS MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL TSH uIU/mL 477 0.55 1.46 3.66 262 0.57 1.46 3.39 218 0.46 1.43 4.00 Free T3 pmol/L 485 3.81 4.86 6.19 264 4.10 5.08 6.28 220 3.70 4.59 5.81 Free T4 pmol/L 484 12.6 16.5 20.7 264 12.5 16.6 21.0 221 12.8 16.3 20.4 Thyreoperoxidase U/mL 482 29 34 58 263 29 33 57 220 28 35 114 Thyreoglobulin U/mL 483 15.2 19.8 42 263 15.7 18.9 33 221 15.1 21.4 91 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 4.3.8 Tumour Markers. Table 4.12 presents the reference intervals for tumour markers. Prostate Specific Antigen (PSA) total which is a male-specific was found to range between 0.24 and 2.82 ng/mL with a median of 0.79 ng/mL. Moreover, Alpha 1-fetoprotein had RI of 1.18 – 17.9 ng/mL with males presenting higher UL (19.1 ng/mL) than females (14.9 ng/mL). Similarly, CA125 presented an UL that is almost three times (42 U/mL) the UL of males (15 U/mL) but with comparable LLs (Table 4.12). Again, carcinoembryonic antigen (CEA) recorded an RI that ranged from 0.43 ng/mL to 3.2 ng/mL with males presenting considerably higher UL (4.0 ng/mL) compared to the UL of females (2.8 ng/mL). Table 4.12: Reference intervals for Tumour markers Test derived parametrically. TEST UNIT MALE+FEMALE MALE FEMALE n LL Me UL n LL Me UL n LL Me UL PSA Total ng/mL 257 0.24 0.79 2.82 Alpha 1-Fetoprotein ng/mL 486 1.18 2.6 17.9 264 1.17 3.0 19.1 222 1.21 2.2 14.9 CA125 U/mL 483 2.03 5.8 27 263 1.97 4.7 15 221 2.25 7.9 42 CEA ng/mL 486 0.43 0.80 3.2 264 0.41 0.84 4.0 215 0.46 0.74 2.8 Variation in sample size is as a result of the application of LAVE which removes study participants with latent diseases. 128 University of Ghana http://ugspace.ug.edu.gh 4.4 Relationship between Age, Ethnicity, BMI, SBP, DBP, Alcohol, Hours of Standing and Haematological analytes. Multiple regression analysis was performed on each analyte to explore the effect of BMI, ethnicity, exercise level, blood pressure, alcohol and hours of standing as sources of variations other than gender and age. In this study, rp ≥ 2.0 is considered to be significant. 4.4.1 Age, BMI and Haematology Analytes. Amongst the explanatory variables that were considered in the multiple regression analysis for Haematological analytes, it was found that age is the single most important determinant in RBC, Hb, and Ht. In RBC, the results showed that increasing age had a corresponding decrease in the level of RBC in males (rp = -0.33) p < 0.01. Similarly, it was observed that in males increasing age caused a decrease in the level of Hb and Ht analyte rp = -0.33 and rp = -0.27 respectively with p < 0.01. 4.4.2 Relationship between Ethnicity and Haematology Analytes. Regarding haematological analytes, there was no association between BMI and these analytes except for eosinophil absolute. Eosinophil significantly increase with age in females with rp = -0.21 (p < 0.01). For ethnicity, basophil absolute concentrations were significantly lower in Akans compared to non-Akans for (females rp = -0.22). Also, the level of MCHC was higher in Akan males compared to that of non-Akan males (rp=0.25). 129 University of Ghana http://ugspace.ug.edu.gh Table 4.13: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Haematology Analytes MALE FEMALE Objectiv n R Age Akan BMI SBP DBP EtOH/ Exer HrStand n R Age Akan BMI SBP DBP EtOH Exer HrStan e W Lvl /W Lvl d WBC 269 0.22 -0.11 0.13 0.09 -0.02 0.00 -0.07 -0.11 -0.04 229 0.21 -0.08 -0.11 0.11 -0.09 -0.03 0.07 0.01 -0.04 P-value 0.09 0.04 0.14 0.85 0.97 0.28 0.08 0.46 0.29 0.10 0.13 0.37 0.74 0.28 0.87 0.55 MPV 260 0.19 -0.18 0.00 -0.03 0.01 -0.01 -0.05 0.02 -0.02 227 0.20 0.06 -0.15 0.05 -0.09 -0.02 0.08 0.01 -0.05 P-value 0.01 0.96 0.69 0.90 0.87 0.41 0.81 0.76 0.46 0.03 0.46 0.38 0.80 0.23 0.84 0.42 PLT 269 0.19 -0.14 -0.01 0.01 0.08 0.03 -0.07 0.13 -0.04 229 0.36 -0.37 0.16 0.05 0.21 -0.06 0.04 -0.05 -0.03 P-value 0.03 0.86 0.86 0.34 0.77 0.28 0.03 0.53 0.03 0.00 0.01 0.44 0.03 0.49 0.52 0.48 0.67 Neu % 269 0.13 0.01 0.06 -0.03 0.05 -0.03 -0.04 0.08 -0.07 229 0.17 -0.01 0.10 -0.08 -0.02 -0.07 0.06 -0.03 -0.06 P-value 0.90 0.31 0.67 0.55 0.69 0.56 0.24 0.24 0.92 0.15 0.28 0.87 0.46 0.36 0.67 0.40 Neu # 268 0.18 -0.07 0.11 0.06 0.08 -0.05 -0.06 -0.03 -0.08 227 0.19 -0.03 -0.03 0.04 -0.08 -0.06 0.11 0.01 -0.08 P-value 0.27 0.08 0.34 0.38 0.58 0.32 0.59 0.21 0.70 0.67 0.62 0.44 0.52 0.12 0.91 0.26 Lym % 269 0.14 -0.03 -0.04 0.08 -0.08 0.03 0.03 -0.09 0.02 229 0.17 0.02 -0.07 0.06 0.03 0.05 -0.07 0.06 0.09 P-value 0.63 0.55 0.24 0.38 0.74 0.67 0.17 0.80 0.17 0.79 0.29 0.43 0.80 0.59 0.31 0.40 Lym # 269 0.25 -0.11 0.07 0.15 -0.05 0.00 -0.03 -0.16 -0.03 229 0.21 -0.07 -0.14 0.15 -0.03 0.00 0.03 0.05 0.03 P-value 0.09 0.23 0.02 0.56 0.96 0.64 0.01 0.61 0.21 0.39 0.03 0.03 0.79 0.99 0.71 0.48 Mon % 268 0.20 0.09 -0.09 -0.09 0.01 0.06 0.09 -0.04 0.07 229 0.26 0.00 -0.04 -0.15 0.07 -0.04 -0.01 -0.17 -0.07 P-value 0.17 0.15 0.17 0.87 0.51 0.16 0.58 0.23 0.95 0.53 0.04 0.45 0.62 0.86 0.01 0.30 Mon # 267 0.13 -0.03 0.03 0.02 0.02 0.01 -0.01 -0.12 0.03 229 0.25 -0.06 -0.13 -0.04 0.00 -0.07 0.04 -0.13 -0.08 P-value 0.61 0.63 0.79 0.84 0.93 0.89 0.06 0.66 0.47 0.05 0.59 0.97 0.47 0.55 0.05 0.26 Eos % 267 0.16 0.01 -0.03 -0.02 0.01 -0.04 0.01 0.08 0.12 228 0.21 -0.03 -0.04 0.19 -0.04 0.07 0.04 -0.06 -0.03 P-value 0.85 0.61 0.70 0.94 0.68 0.91 0.23 0.05 0.74 0.57 0.01 0.66 0.45 0.58 0.40 0.68 Eos # 268 0.15 0.02 0.01 0.00 0.01 -0.08 0.00 0.06 0.10 229 0.24 -0.02 -0.09 0.21 -0.08 0.07 0.06 -0.04 -0.05 130 University of Ghana http://ugspace.ug.edu.gh P-value 0.79 0.82 0.99 0.95 0.34 0.95 0.36 0.09 0.81 0.18 0.00 0.45 0.47 0.41 0.54 0.50 Bas % 265 0.19 0.06 -0.15 -0.05 0.09 -0.04 0.03 0.04 0.03 229 0.23 -0.01 -0.16 0.04 0.09 0.06 0.05 -0.01 -0.06 P-value 0.39 0.02 0.43 0.33 0.61 0.63 0.49 0.69 0.85 0.02 0.61 0.36 0.52 0.47 0.88 0.35 Bas # 266 0.13 0.02 -0.08 -0.02 0.10 -0.09 0.01 -0.03 0.01 229 0.27 -0.04 -0.22 0.07 0.05 0.04 0.08 0.04 -0.08 P-value 0.78 0.20 0.74 0.28 0.30 0.88 0.68 0.88 0.59 0.00 0.30 0.61 0.69 0.23 0.60 0.24 RBC 269 0.39 -0.33 -0.08 -0.11 0.03 0.03 0.11 0.06 -0.10 229 0.25 -0.08 -0.13 0.10 -0.05 0.18 0.00 -0.01 -0.11 P-value 0.00 0.17 0.08 0.73 0.69 0.07 0.35 0.07 0.31 0.05 0.16 0.61 0.05 1.00 0.83 0.12 Hb 269 0.23 -0.20 0.02 0.06 0.14 -0.05 0.08 -0.08 -0.01 227 0.25 0.07 -0.06 0.15 -0.17 0.14 0.08 0.01 0.10 P-value 0.00 0.75 0.36 0.12 0.54 0.19 0.20 0.86 0.36 0.39 0.03 0.09 0.12 0.26 0.88 0.15 Ht 269 0.31 -0.27 -0.13 -0.03 0.17 -0.05 0.07 -0.05 -0.07 229 0.22 0.06 -0.15 0.12 -0.05 0.04 0.08 0.01 0.01 P-value 0.00 0.04 0.65 0.04 0.58 0.29 0.38 0.27 0.46 0.02 0.09 0.64 0.68 0.23 0.91 0.85 MCV 268 0.29 0.17 -0.03 0.09 0.13 -0.07 -0.08 -0.08 0.05 227 0.22 0.13 -0.01 0.01 0.05 -0.16 0.07 0.02 0.11 P-value 0.01 0.66 0.14 0.13 0.39 0.20 0.18 0.42 0.10 0.93 0.84 0.60 0.10 0.30 0.72 0.11 MCH 268 0.32 0.19 0.10 0.16 0.09 -0.06 -0.06 -0.08 0.09 226 0.27 0.14 0.09 0.06 -0.09 -0.05 0.06 0.03 0.19 P-value 0.00 0.09 0.01 0.31 0.50 0.37 0.17 0.15 0.07 0.21 0.38 0.37 0.62 0.36 0.61 0.01 MCHC 269 0.31 0.10 0.25 0.16 -0.05 -0.01 0.03 -0.06 0.09 229 0.28 0.04 0.15 0.11 -0.11 0.11 0.02 0.01 0.15 P-value 0.11 0.00 0.01 0.57 0.88 0.62 0.36 0.15 0.64 0.03 0.13 0.28 0.21 0.73 0.86 0.02 RDW 268 0.23 0.05 -0.14 -0.09 -0.08 0.14 0.01 0.02 -0.15 227 0.26 0.02 -0.07 -0.09 0.10 0.04 0.01 0.01 -0.19 P-value 0.44 0.03 0.16 0.36 0.11 0.93 0.80 0.02 0.77 0.29 0.18 0.31 0.64 0.83 0.86 0.00 131 University of Ghana http://ugspace.ug.edu.gh 4.5 Demographic and Physical Characteristics and Clinical Chemistry 4.5.1 Relationship between Age and Clinical Chemistry The clinical chemistry analytes had interesting findings with regards to the effect of age as an explanatory variable on the various parameters. Foremost, it was revealed that amongst males, increasing age caused a moderate decrease (rp = -0.38) (p-value <0.01) in Ca level. CRP was significantly affected by age amongst males with rp = 0.26 (p-value <0.01). Also, amongst females it was observed that UA and Mg increase with respect to age (rp = 0.31) (p-value <0.01) and (rp = 0.21) (p <0.01) respectively. 4.5.2 Ethnicity and Clinical Chemistry Interestingly, among all the clinical chemistry analytes only phosphate, ANGAP and folate were found to be affected by ethnicity. Specifically, Akans had lower phosphate and ANGAP levels than other ethnic groups in both sexes with males (rp = -0.20 and rp = - 0.22) and females, (rp = -0.24 and rp = -0.28) respectively with (p <0.01). Contrastingly, folate concentrations were significantly higher in Akans compared to other ethnic groups for both sexes (males rp = 0.21 and female rp= 0.24) with a statistically significant value (p <0.01). 4.5.3 Relationship between BMI and Clinical Chemistry Several analytes were significantly affected by BMI as one of the explanatory variables used in the multiple regression analysis. In females, it was revealed that BMI is an important determinant of AMY and CRP, with an increasing BMI leading to a decrease in AMY level (rp = -0.30) p <0.01, whiles CRP increases with increasing BMI (rp = 0.37). Likewise, glucose increase significantly with increasing BMI (rp = 0.25) in males, whiles HbA1c% increase with increasing BMI in both males and females. Moreover, BMI was 132 University of Ghana http://ugspace.ug.edu.gh found to moderately affect UA more in males (rp = 0.39) than in females (rp = 0.27). Regarding C3 and C4, BMI was found to be a key determinant in both sexes; in males, C3 was significantly affected by BMI rp = 0.51 than in females rp = 0.35 (p <0.01). Yet, for C4, the effect of BMI rp = 0.21 was the same for both males and females (p-value <0.01). 4.5.4 Blood Pressure (SBP and DBP) and Clinical Chemistry Considering the effect of Blood pressure on clinical chemistry analytes, it was observed that among females SBP was a source of influence on phosphate with a unit increase in SBP leading to 0.20 increase in phosphate level (p <0.05). Similarly, LDH levels increases with a unit increase in SBP rp = 0.20 (p-value <0.05) in females. The results further revealed that among females, SBP (rp = 0.25) and DBP (rp = -0.26) affected CRP moderately at statistical significance of p <0.05. 4.5.5 Exercise level, Standing Hours and Clinical Chemistry Surprisingly, among all the clinical chemistry analytes, only phosphate and Albumin had significant association with exercise level and standing hours. Hours of standing negatively affected phosphate, thus phosphate decreased with an increased change in the hours of standing (rp = -0.21) p <0.01 in females. Similarly, it was also revealed Albumin was significantly affected by the level of exercise in females; here albumin increases with increasing exercise level rp = 0.20 (p-value <0.01). 133 University of Ghana http://ugspace.ug.edu.gh Table 4.14: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Clinical Chemistry Analytes MALE FEMALE Objective n R Age Akan BMI SBP DBP EtO Exer HrSt n R Age Akan BMI SBP DBP EtOH/ Exer HrSta Variable H/W Lvl and W Lvl nd Ca 268 0.39 -0.38 -0.11 -0.08 0.09 0.01 -0.06 0.08 -0.05 229 0.17 0.07 -0.04 -0.04 0.05 -0.01 0.05 0.13 -0.04 p-value 0.00 0.05 0.17 0.29 0.93 0.33 0.15 0.43 0.40 0.59 0.58 0.62 0.90 0.44 0.07 0.58 cCa+ 268 0.20 -0.05 -0.14 -0.02 0.00 0.02 -0.13 -0.02 -0.01 228 0.22 0.12 0.01 0.11 0.05 -0.03 0.08 -0.01 -0.05 p-value 0.47 0.03 0.70 0.97 0.81 0.04 0.81 0.85 0.12 0.89 0.12 0.62 0.74 0.27 0.85 0.46 Mg 269 0.20 0.13 -0.12 -0.08 0.00 0.00 0.04 -0.02 -0.08 229 0.35 0.21 -0.17 -0.09 0.10 -0.01 0.00 0.03 -0.11 p-value 0.05 0.05 0.19 1.00 0.96 0.51 0.76 0.21 0.00 0.01 0.20 0.28 0.88 0.95 0.64 0.10 P 267 0.29 0.13 -0.20 -0.08 -0.05 -0.04 -0.12 -0.01 -0.03 228 0.40 0.00 -0.24 -0.05 0.20 -0.08 -0.03 -0.05 -0.21 p-value 0.06 0.00 0.20 0.58 0.66 0.05 0.85 0.67 0.95 0.00 0.42 0.03 0.34 0.59 0.47 0.00 CK 269 0.23 -0.13 0.00 0.14 0.18 -0.15 0.08 0.04 -0.05 227 0.27 0.01 0.07 0.18 0.19 -0.15 0.00 0.10 0.02 p-value 0.06 0.96 0.03 0.04 0.08 0.22 0.53 0.37 0.92 0.30 0.01 0.06 0.10 0.94 0.15 0.75 AMY 268 0.20 0.03 -0.04 -0.18 0.04 -0.10 0.01 0.02 -0.04 227 0.35 -.03 0.11 -0.30 0.00 -0.06 -0.02 -0.02 -0.04 p-value 0.65 0.55 0.01 0.65 0.27 0.88 0.77 0.52 0.73 0.08 0.00 1.00 0.52 0.75 0.74 0.54 LDH 268 0.32 -0.11 -0.16 0.12 0.15 -0.05 -0.10 -0.13 -0.05 229 0.30 0.00 -0.16 0.11 0.20 -0.12 0.12 0.00 -0.08 p-value 0.09 0.01 0.05 0.07 0.58 0.09 0.04 0.42 0.97 0.02 0.11 0.04 0.18 0.08 0.94 0.22 CRP 268 0.31 0.21 0.03 0.17 -0.06 -0.03 0.01 0.09 -0.07 227 0.44 0.03 0.06 0.37 0.25 -0.26 0.02 -0.04 -0.08 p-value 0.00 0.58 0.01 0.47 0.71 0.91 0.14 0.27 0.67 0.32 0.00 0.01 0.00 0.74 0.55 0.22 UA 266 0.41 0.03 0.00 0.39 -0.07 0.09 -0.04 0.08 -0.01 229 0.44 0.31 0.06 0.27 -0.09 -0.05 0.05 -0.07 -0.03 p-value 0.64 0.98 0.00 0.37 0.28 0.49 0.16 0.90 0.00 0.32 0.00 0.34 0.54 0.47 0.30 0.60 C3 269 0.53 0.00 0.00 0.51 0.03 -0.09 -0.09 0.12 -0.09 228 0.37 0.05 0.00 0.35 -0.04 0.03 0.01 -0.03 0.02 p-value 0.98 0.97 0.00 0.67 0.23 0.09 0.03 0.10 0.47 0.94 0.00 0.65 0.70 0.90 0.68 0.71 134 University of Ghana http://ugspace.ug.edu.gh C4 267 0.31 0.15 0.09 0.21 0.03 -0.02 0.07 -0.07 0.05 228 0.32 0.15 0.10 0.22 0.02 -0.07 -0.02 0.08 0.01 p-value 0.02 0.14 0.00 0.71 0.80 0.24 0.27 0.40 0.05 0.14 0.00 0.87 0.44 0.75 0.23 0.84 Glu 265 0.33 -0.02 -0.01 0.25 0.03 0.14 -0.08 0.04 -0.04 229 0.34 0.14 0.10 0.11 0.11 0.08 -0.09 0.03 0.03 p-value 0.81 0.93 0.00 0.68 0.09 0.21 0.51 0.50 0.07 0.14 0.12 0.27 0.37 0.18 0.69 0.62 HbA1c% 269 0.34 0.12 -0.11 0.25 -0.02 0.05 -0.06 0.07 -0.04 229 0.34 0.08 -0.04 0.22 0.18 -0.08 -0.05 0.01 0.06 p-value 0.07 0.08 0.00 0.79 0.52 0.32 0.23 0.46 0.31 0.56 0.00 0.06 0.38 0.42 0.87 0.35 135 University of Ghana http://ugspace.ug.edu.gh 4.5.6 Age, BMI and Kidney Function Tests Age proved to be a source of variation on some of the kidney function analytes. As shown in Table 4.15, it was observed that age significantly affected eGFR in both males (rp = - 0.43) and females (rp = -0.45) with a statistically significant value (p <0.01). For urea (rp = 0.40), creatinine (rp = 0.24), Na (rp = 0.51), K (rp = 0.25), Cl (rp = -0.22) the effect of age was only observed in females, with all statistically significant at p-value <0.01. On the other hand, ANGAP was influenced by age (rp = -0.26) in only males with a p <0.01. TCO2 increased with age in both males and females, with relatively higher rate in males (rp = 0.33) than females (rp = 0.29) at p <0.01. Interestingly, BMI had no significant effects on all the kidney function analytes, which includes the electrolytes (Na, K and Cl), TCO2, ANGAP, Urea, Cre and eGFR. 4.5.7 Relationship between Blood Pressure (SBP and DBP) and Kidney Function Blood pressure specifically, SBP and DBP had an influence on most of the electrolytes and TCO2. Among the electrolytes, K (rp = -0.23) and Cl (rp = -0.22) both decreased with SBP in females, but no strong association of such analytes in males. Nonetheless, in males TCO2 decreased with one unit change in SBP with (rp = -0.25) whiles ANGAP increased with one unit change in SBP with (rp = 0.28). Again, Cl depicts a positive association with DBP, thus Cl level increased with increase unit change in DBP (rp = 0.20) in males. 136 University of Ghana http://ugspace.ug.edu.gh Table 4.15: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Kidney Function Test Analytes MALES FEMALES Objective n R Age Akan BMI SBP DBP EtO Exer HrSt n R Age Akan BMI SBP DBP EtO Exer HrSt Variable H/W Lvl and H/W Lvl and Na 268 0.34 0.11 -0.17 0.04 0.14 -0.05 -0.16 -0.05 0.00 228 0.65 0.51 -0.09 -0.02 0.15 0.03 0.13 -0.13 -0.09 p-value 0.09 0.00 0.50 0.10 0.55 0.01 0.38 0.94 0.00 0.11 0.65 0.05 0.73 0.02 0.01 0.10 K 268 0.25 0.14 0.08 -0.12 -0.19 0.12 0.12 0.04 0.04 228 0.28 0.25 0.08 -0.03 -0.23 0.16 -0.08 0.10 -0.05 p-value 0.03 0.18 0.06 0.03 0.16 0.06 0.51 0.55 0.00 0.23 0.65 0.02 0.09 0.25 0.16 0.43 Cl 269 0.23 0.12 0.07 0.10 -0.01 0.05 0.01 0.05 0.10 229 0.30 0.22 0.15 -0.05 -0.22 0.20 0.07 -0.14 0.06 p-value 0.08 0.27 0.13 0.95 0.52 0.85 0.43 0.09 0.00 0.03 0.45 0.02 0.03 0.32 0.04 0.38 TCO2 269 0.37 0.33 0.10 -0.02 -0.25 0.12 0.15 0.01 0.10 229 0.46 0.29 0.14 0.12 0.10 -0.08 -0.12 0.16 -0.09 p-value 0.00 0.09 0.69 0.00 0.13 0.01 0.84 0.08 0.00 0.03 0.07 0.29 0.34 0.05 0.01 0.14 ANGAP 269 0.46 -0.26 -0.22 -0.04 0.28 -0.15 -0.20 -0.10 -0.14 229 0.38 -0.01 -0.28 -0.06 0.17 -0.05 0.12 -0.11 -0.04 p-value 0.00 0.00 0.46 0.00 0.06 0.00 0.07 0.01 0.90 0.00 0.35 0.07 0.54 0.06 0.08 0.51 Urea 269 0.24 0.09 -0.07 0.08 -0.18 0.03 0.05 -0.15 -0.04 229 0.40 0.40 -0.02 0.04 -0.08 0.00 0.04 0.14 -0.04 p-value 0.20 0.23 0.22 0.04 0.70 0.41 0.02 0.55 0.00 0.72 0.53 0.38 0.96 0.58 0.03 0.52 Cre 269 0.23 0.14 -0.04 0.04 0.02 -0.12 0.08 -0.08 -0.10 229 0.35 0.24 0.05 0.06 0.13 -0.11 0.10 -0.09 -0.05 p-value 0.03 0.56 0.55 0.83 0.15 0.23 0.20 0.09 0.00 0.46 0.35 0.17 0.22 0.11 0.17 0.42 eGFR 269 0.46 -0.43 0.04 -0.07 -0.01 0.10 -0.07 0.08 0.10 227 0.55 -0.45 -0.04 -0.09 -0.14 0.08 -0.09 0.06 0.07 p-value 0.00 0.46 0.26 0.89 0.18 0.21 0.19 0.08 0.00 0.47 0.14 0.12 0.34 0.11 0.27 0.25 137 University of Ghana http://ugspace.ug.edu.gh 4.5.8 Age and Liver Function Parameters As shown in Table 4.16 it was found that age significantly affected total protein in both males and females yet the magnitude is bigger in males (rp = -0.39) than in females (rp = - 0.25) (p-value <0.01). Albumin was significantly affected by age in males; here albumin decreases with increasing age rp = -0.45 (p <0.01). In contrast, globulin decreased in females with age, rp = -0.26 (p <0.05) faster than in males. 4.5.9 BMI and Liver Function Considering liver function analytes, it was observed that BMI was an important source of variation for only ALT and GGT analytes. BMI had positive effect on ALT and GGT analytes in both sexes; BMI increases with increasing concentration levels of ALT for males and females rp = 0.33 (p <0.01). Similarly, GGT increases with increasing BMI, but relatively stronger in males rp = 0.31 than in females rp = 0.23 for GGT (p <0.01). 138 University of Ghana http://ugspace.ug.edu.gh Table 4.16: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Liver Function Test Analytes MALES FEMALES Objective n R Age Akan BMI SBP DBP EtO Exer HrSta n R Age Akan BMI SBP DBP EtO ExeL HrSt Variable H/W Lvl nd H/W vl and TP 269 0.42 -0.39 0.08 -0.03 0.17 -0.03 0.00 0.12 -0.07 229 0.33 -0.25 -0.03 -0.15 0.11 -0.06 -0.02 0.13 -0.03 p-value 0.00 0.18 0.56 0.03 0.68 0.97 0.05 0.26 0.00 0.69 0.03 0.27 0.48 0.79 0.06 0.67 Alb 269 0.47 -0.45 0.00 -0.06 0.10 -0.01 0.05 0.12 -0.06 229 0.27 -0.05 -0.07 -0.17 0.02 0.02 -0.01 0.20 0.01 p-value 0.00 0.99 0.32 0.19 0.94 0.41 0.04 0.31 0.54 0.27 0.01 0.87 0.85 0.86 0.00 0.94 Glb 269 0.21 -0.17 0.09 0.00 0.14 -0.04 -0.03 0.06 -0.04 229 0.27 -0.26 0.01 -0.08 0.12 -0.08 -0.01 0.03 -0.04 p-value 0.01 0.14 0.96 0.11 0.68 0.66 0.33 0.51 0.00 0.85 0.28 0.25 0.36 0.84 0.65 0.60 AST 268 0.24 -0.06 -0.03 0.16 0.04 0.06 0.13 -0.06 -0.02 229 0.16 0.07 0.01 0.06 -0.04 0.01 -0.05 -0.09 -0.03 p-value 0.39 0.68 0.01 0.67 0.51 0.05 0.32 0.71 0.35 0.91 0.39 0.71 0.92 0.42 0.21 0.62 ALT 267 0.41 -0.17 0.07 0.33 -0.09 0.15 0.14 -0.03 -0.06 228 0.34 -0.10 0.03 0.32 0.03 -0.13 -0.09 0.11 0.04 p-value 0.01 0.25 0.00 0.24 0.05 0.02 0.58 0.29 0.20 0.64 0.00 0.79 0.15 0.17 0.08 0.52 ALP 268 0.14 -0.01 -0.07 0.03 0.00 -0.10 0.02 -0.03 -0.03 228 0.31 0.19 -0.09 0.08 0.11 -0.03 -0.05 -0.02 -0.01 p-value 0.87 0.25 0.68 0.96 0.26 0.71 0.68 0.64 0.02 0.19 0.22 0.24 0.70 0.47 0.76 0.88 GGT 266 0.38 -0.06 -0.10 0.31 -0.02 0.16 0.09 -0.04 0.06 229 0.29 0.02 0.02 0.23 0.15 -0.15 0.05 -0.01 0.10 p-value 0.32 0.10 0.00 0.83 0.05 0.16 0.46 0.28 0.76 0.72 0.00 0.13 0.10 0.42 0.89 0.14 TBil 268 0.20 -0.09 0.10 0.02 0.04 0.00 -0.13 0.07 0.06 229 0.26 0.09 -0.17 -0.01 0.04 0.01 0.14 0.06 -0.03 p-value 0.21 0.09 0.76 0.68 0.97 0.04 0.30 0.30 0.24 0.01 0.94 0.65 0.94 0.04 0.42 0.69 DBil 267 0.21 -0.08 0.05 -0.08 0.03 -0.03 -0.07 0.05 0.11 229 0.25 0.04 -0.16 -0.05 0.10 -0.04 0.14 0.06 0.01 p-value 0.23 0.46 0.22 0.72 0.76 0.30 0.40 0.07 0.64 0.02 0.47 0.32 0.68 0.04 0.40 0.85 139 University of Ghana http://ugspace.ug.edu.gh 4.5.10 Age and Lipids Analytes. Interestingly, it was observed that the effect of age on lipids occurred only amongst females. TG, TC and LDL-Cholesterol significantly decrease with age in females with rp = -0.20 and rp = -0.32 and rp = -0.21 (p <0.01) respectively. 4.5.11 BMI and Lipids Analytes. Unsurprisingly, BMI was found to positively affect TC in males, rp = 0.28 (p <0.01). Similarly, TG was affected by BMI almost at the same magnitude albeit in both males, rp = 0.28 and females rp = 0.26 (p <0.01). Amongst males, BMI remained an important source of variation for HDL-C and LDL-C with rp = -0.24 (p-value <0.01) and rp = 0.31 (p <0.01) respectively. Interestingly, amongst females, while BMI affected LDL-C rp = 0.21 (p <0.01), age also positively affected it rp = 0.21 (p-value <0.05). Again, BMI was positively associated with CHOL/HDL (rp = 0.41) (p <0.01) but negatively associated with HDL/LDL (rp = -0.38) (p <0.01). Consequently, it is obvious that amongst all other explanatory factors considered in this MRA, BMI remains one of the main determinants of lipids level in Ghanaians. 4.5.12 Age and Immunoglobulins. Concerning immunoglobulins, it was found that age is the only factor that significantly affects them. Thus, it was found that IgA in males increases with increasing age (rp = 0.22) p <0.01. On the other hand, IgM was negatively influenced by age in females, with IgM decreasing with increasing age (rp = -0.23) p <0.01. 140 University of Ghana http://ugspace.ug.edu.gh Table 4.17: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Hours of Standing on Lipids Profile Analytes MALES FEMALES Objective n R Age Akan BMI SBP DBP EtOH Exer HrSt n R Age Akan BMI SBP DBP EtOH Exer HrSta Variable /W Lvl and /W Lvl nd TC 269 0.33 0.05 0.09 0.28 0.12 -0.08 -0.03 0.03 -0.02 229 0.36 0.26 0.11 0.18 -0.09 0.07 -0.01 0.07 0.01 p-value 0.43 0.12 0.00 0.16 0.32 0.59 0.59 0.75 0.00 0.09 0.01 0.35 0.44 0.90 0.26 0.86 TG 269 0.43 0.17 0.03 0.28 -0.04 0.05 0.01 0.05 -0.20 229 0.46 0.32 -0.08 0.26 -0.13 0.15 0.00 -0.05 0.04 p-value 0.01 0.66 0.00 0.58 0.57 0.81 0.44 0.00 0.00 0.22 0.00 0.16 0.07 0.95 0.42 0.53 HDL-C 269 0.31 0.05 0.10 -0.24 0.06 -0.13 0.01 0.02 0.10 227 0.19 0.15 0.05 -0.03 -0.08 -0.08 0.00 0.04 -0.03 p-value 0.47 0.11 0.00 0.47 0.11 0.91 0.70 0.10 0.06 0.48 0.63 0.44 0.38 0.96 0.54 0.66 LDL-C 269 0.34 0.02 0.07 0.31 0.12 -0.06 -0.03 0.02 -0.01 228 0.34 0.21 0.09 0.20 -0.11 0.10 0.02 0.07 -0.02 p-value 0.77 0.25 0.00 0.16 0.47 0.68 0.75 0.85 0.00 0.15 0.00 0.27 0.27 0.79 0.28 0.81 CHOL/HDL 269 0.46 0.00 -0.02 0.41 0.07 0.05 -0.04 0.03 -0.09 229 0.34 0.10 0.04 0.22 0.01 0.13 -0.01 0.04 0.04 p-value 0.98 0.69 0.00 0.35 0.55 0.50 0.66 0.11 0.19 0.57 0.00 0.93 0.15 0.89 0.59 0.53 HDL/LDL 269 0.42 -0.01 0.00 -0.38 -0.09 -0.01 0.01 0.00 0.05 229 0.31 -0.08 -0.05 -0.18 0.00 -0.15 -0.01 -0.05 -0.02 p-value 0.91 0.97 0.00 0.26 0.93 0.81 0.97 0.36 0.28 0.41 0.01 1.00 0.10 0.87 0.48 0.75 141 University of Ghana http://ugspace.ug.edu.gh Table 4.18: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, Exercise Level and Hours of Standing on Immunoglobulin Analytes MALES FEMALES Objective n R Age Akan BMI SBP DBP EtOH/ Exer HrSta n R Age Aka BMI SBP DBP EtOH Exer HrSta Variable W Lvl nd n /W Lvl nd IgG 268 0.19 -0.08 0.05 -0.08 0.11 -0.07 -0.12 0.01 -0.02 229 0.27 -0.16 0.09 -0.18 0.10 0.08 0.00 0.00 -0.09 p-value 0.22 0.43 0.20 0.20 0.43 0.07 0.91 0.78 0.04 0.17 0.01 0.33 0.37 0.99 0.97 0.17 IgA 269 0.24 0.22 0.01 -0.01 -0.05 0.08 0.07 -0.01 -0.02 229 0.25 0.19 0.17 -0.09 -0.03 0.06 -0.04 -0.01 -0.01 p-value 0.00 0.90 0.88 0.58 0.32 0.26 0.84 0.78 0.01 0.01 0.20 0.77 0.54 0.55 0.83 0.92 IgM 267 0.18 -0.04 -0.08 -0.11 -0.06 0.05 -0.06 0.08 -0.01 227 0.25 -0.23 0.01 -0.02 0.03 -0.04 0.04 0.05 -0.10 p-value 0.54 0.21 0.08 0.51 0.58 0.32 0.23 0.83 0.00 0.93 0.82 0.79 0.66 0.54 0.51 0.16 142 University of Ghana http://ugspace.ug.edu.gh 4.5.13 Age,Thyroid and Endocrine Parameters. As shown in Table 4.19 several analytes were significantly affected by the explanatory variables used in this MRA, particularly with age. Firstly, FT3 was significantly affected by age; with FT3 level declining with ageing (rp = -0.30) p < 0.01 with this happening only among males. Similarly, in males, cortisol significantly decreased with age (rp = -0.25) p < 0.01. GH was also affected by age but in females (rp = -0.25) p < 0.01, Whiles with PTH, age remained a key determinant in both males and females with about 0.50 unit increase in PTH level with every unit increase in age (p < 0.01). Moreover, it is worth of note that, in the female-specific hormones -estradiol, progesterone, LH, FSH, and PRL age was the single most important determinant. Thus, respectively estradiol, progesterone, PRL levels decreased with increasing age rp = -0.56, rp = -0.40 and rp = -0.44 (all statistically significant at p <0.01). Contrastingly, LH and FSH levels increased with increasing age rp = 0.51 and rp = 0.71 respectively (all statistically significant at p <0.01). 4.5.14 BMI, Thyroid and Endocrine Parameters Regarding Thyroid and Endocrine analytes, BMI was identified as a key determinant in only males expect for insulin analyte. The results revealed that insulin was significantly affected by BMI in males (rp = 0.46) as well as in females (rp = 0.32) p-value < 0.01. However, the effect of BMI also significantly increases with increased levels of GH and PTH in only males with rp = 0.27 and rp = 0.21(p < 0.01) respectively. 143 University of Ghana http://ugspace.ug.edu.gh Table 4.19: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Thyroid and Endocrine Analytes MALES FEMALES Objective n R Age Akan BMI SBP DBP EtOH ExerL n R Age Akan BMI SBP DBP EtOH/ Exer Variable /W vl W Lvl TSH 268 0.15 -0.06 0.06 0.05 0.01 -0.05 0.09 -0.05 226 0.18 -0.10 0.08 0.02 0.01 0.04 0.10 -0.10 p-value 0.37 0.31 0.41 0.93 0.58 0.16 0.42 0.00 0.02 0.47 0.56 0.36 0.35 0.38 FT4 268 0.23 -0.10 0.07 -0.16 -0.02 -0.02 0.01 -0.05 230 0.23 -0.12 -0.11 0.06 0.09 -0.02 0.14 -0.08 p-value 0.14 0.28 0.01 0.78 0.83 0.83 0.39 0.14 0.21 0.43 0.07 0.16 0.77 0.04 FT3 268 0.30 -0.30 0.03 -0.02 0.15 -0.04 -0.02 -0.05 230 0.19 -0.05 -0.06 -0.03 0.10 -0.01 0.05 -0.15 p-value 0.00 0.57 0.75 0.07 0.64 0.73 0.39 0.49 0.35 0.65 0.30 0.91 0.41 0.03 GH 269 0.28 0.00 0.05 -0.27 0.13 -0.07 -0.02 0.02 231 0.41 -0.25 -0.15 -0.18 0.12 -0.13 0.00 -0.15 p-value 0.96 0.44 0.00 0.12 0.37 0.73 0.79 0.00 0.02 0.01 0.20 0.12 0.94 0.02 PTH 268 0.59 0.50 0.06 0.21 0.00 0.01 0.08 0.05 231 0.54 0.46 0.11 0.18 -0.09 0.08 -0.12 0.00 p-value 0.00 0.25 0.00 0.98 0.85 0.13 0.31 0.00 0.06 0.00 0.30 0.31 0.03 0.99 Cortisol 269 0.34 -0.25 0.08 -0.14 0.24 -0.05 0.07 -0.03 231 0.183 0.11 -0.01 -0.09 0.04 0.05 0.09 -0.06 p-value 0.34 0.00 0.21 0.02 0.00 0.51 0.27 0.67 0.15 0.93 0.19 0.68 0.57 0.16 0.39 Testostero 269 0.47 -0.12 -0.12 -0.42 0.15 -0.09 -0.02 -0.03 ne p-value 0.05 0.04 0.00 0.06 0.23 0.72 0.59 Insulin 268 0.59 -0.09 0.01 0.46 -0.07 0.06 0.04 0.05 230 0.437 -0.19 0.15 0.32 -0.04 0.18 -0.06 0.04 p-value 0.45 0.15 0.88 0.00 0.35 0.47 0.54 0.41 0.01 0.02 0.00 0.67 0.04 0.32 0.48 Estradiol 228 0.60 -0.56 0.05 0.01 -0.06 -0.04 0.02 -0.01 p-value 0.00 0.40 0.92 0.50 0.62 0.75 0.88 Progester 228 0.55 -0.40 0.07 -0.06 -0.15 -0.09 -0.05 -0.04 one 144 University of Ghana http://ugspace.ug.edu.gh p-value 0.00 0.24 0.33 0.07 0.25 0.35 0.50 LH 228 0.59 0.51 0.02 -0.06 0.13 -0.02 -0.15 0.03 p-value 0.00 0.69 0.28 0.11 0.81 0.01 0.57 FSH 228 0.75 0.71 -0.01 -0.07 0.07 0.06 -0.06 0.00 p-value 0.00 0.82 0.13 0.28 0.34 0.19 0.98 PRL 228 0.49 -0.44 0.10 -0.09 0.04 -0.04 0.08 -0.07 p-value 0.00 0.12 0.14 0.69 0.66 0.20 0.23 145 University of Ghana http://ugspace.ug.edu.gh 4.5.15 Age, BMI and Iron and Vitamin Metabolism The MRA results for iron and vitamin analytes shown in table 4.10 depicts positive association with age among females; thus, ferritin and folate concentration level increases with increasing age among female (rp= 0.44) and (rp= 0.24) respectively. Again, age proved to be a source of variation in transferrin in both males (rp = -0.22) and females (rp = -0.36) with a statistically significant value (p-value <0.01). Vitamin D was significantly affected by age; with Vitamin D level increasing with ageing (rp = 0.20) p < 0.01. Moreover, the association between BMI and iron analytes existed only in ferritin. Ferritin was significantly affected by BMI in males; here ferritin increases with increasing BMI rp = 0.29 (p <0.01). 4.5.16 Age and Tumour Markers. Amongst the explanatory variables that were considered in this MRA for tumour markers, it was found that age is the single most important determinant in AFP and PSA. In AFP, the results showed that increasing age had a corresponding increase in the level of AFP in females (rp = 0.22) p < 0.01. Similarly, in PSA which is a male-specific analyte, it was discovered that increasing age caused a rise in the level of PSA (rp = 0.44) p < 0.01. 146 University of Ghana http://ugspace.ug.edu.gh Table 4.20: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Iron and Vitamins MALES FEMALES Objective n R Age Akan BMI SBP DBP EtO Exer HrSta n R Age Akan BMI SBP DBP EtOH/ Exer HrSta Variable H/W Lvl nd W Lvl nd Fe 269 0.15 -0.07 -0.02 -0.01 0.12 -0.13 -0.01 -0.07 -0.03 229 0.28 0.09 -0.10 0.14 -0.01 0.02 0.07 0.08 0.17 P-value 0.32 0.72 0.84 0.16 0.14 0.86 0.26 0.60 0.25 0.14 0.05 0.94 0.85 0.31 0.23 0.01 Ferr 269 0.36 0.16 0.03 0.29 -0.12 0.02 0.06 -0.07 -0.02 229 0.46 0.46 -0.02 0.07 -0.06 0.03 0.06 0.00 0.08 P-value 0.01 0.57 0.00 0.16 0.76 0.34 0.24 0.71 0.00 0.75 0.25 0.54 0.70 0.30 0.96 0.20 TF 268 0.29 -0.22 0.08 0.07 0.01 -0.02 -0.03 0.14 0.06 229 0.37 -0.36 -0.05 -0.07 0.11 -0.09 -0.03 0.05 -0.03 P-value 0.00 0.19 0.26 0.94 0.83 0.62 0.02 0.32 0.00 0.46 0.28 0.27 0.28 0.67 0.49 0.61 VitB12 269 0.27 -0.03 0.15 -0.07 -0.18 0.26 0.05 -0.01 0.07 229 0.25 -0.11 0.14 0.17 0.03 -0.13 -0.03 0.03 0.04 P-value 0.68 0.02 0.30 0.04 0.00 0.44 0.83 0.26 0.15 0.03 0.02 0.73 0.17 0.65 0.64 0.54 Folate 268 0.37 0.00 0.20 0.18 0.10 -0.01 0.09 0.16 0.04 227 0.34 0.22 0.23 0.00 -0.15 0.07 -0.02 0.07 0.11 P-value 0.96 0.00 0.00 0.24 0.89 0.13 0.01 0.46 0.00 0.00 0.96 0.12 0.43 0.78 0.32 0.10 VitD 268 0.21 0.20 0.04 -0.04 0.04 -0.02 0.00 0.03 0.02 231 0.19 0.13 0.07 -0.05 -0.07 0.10 -0.07 0.10 0.08 P-value 0.00 0.52 0.50 0.64 0.81 0.99 0.69 0.48 0.10 0.31 0.44 0.47 0.29 0.27 0.14 0.42 147 University of Ghana http://ugspace.ug.edu.gh Table 4.21: Effect of Age, Ethnicity, BMI, SBP, DBP, Alcohol, and Exercise level on Tumour Markers MALES FEMALES Objective n R Age Akan BMI SBP DBP EtOH/ Exer n R Age Akan BMI SBP DBP EtOH/W ExerLv Variable W Lvl l AFP 269 0.14 0.14 0.02 0.11 -0.01 0.02 0.02 -0.08 231 0.29 0.22 0.15 -0.05 0.06 -0.08 -0.06 0.06 p-value 0.77 0.09 0.93 0.83 0.82 0.19 0.77 0.00 0.02 0.47 0.56 0.36 0.35 0.38 CEA 269 0.14 0.09 0.05 -0.01 0.05 0.00 -0.04 -0.06 231 0.28 0.11 -0.08 -0.05 0.18 -0.13 0.02 -0.14 p-value 0.19 0.39 0.86 0.58 0.98 0.48 0.38 0.14 0.21 0.43 0.07 0.16 0.77 0.04 CA125 269 0.17 -0.02 -0.09 0.12 -0.04 0.05 0.05 0.00 231 0.42 -0.43 0.03 -0.03 0.14 -0.13 0.06 -0.08 p-value 0.81 0.15 0.05 0.65 0.58 0.40 0.98 0.00 0.65 0.66 0.12 0.13 0.37 0.20 PSA 267 0.45 0.44 0.01 -0.04 0.11 -0.15 0.02 0.11 p-value 0.00 0.93 0.47 0.19 0.06 0.72 0.07 148 University of Ghana http://ugspace.ug.edu.gh 4.6 Age, Sex and Haematology This section presents the results of sources of variation with respect to sex, age and ethnicity specifically exploring the possibility of partitioning. 4.6.1 Age, Sex and Haematology Analytes. Table 4.22 presents the standard deviation ratios by sex, age and ethnicity. By use of three- level nested ANOVA, between-sex differences with SDRsex ≥ 0.4 were observed for four (4) haematological analytes (PLT, RBC, Hb, and Ht). The SDRs represent the magnitude of sex and age-related changes in the haematology analytes. PLT, RBC, Hb, and Ht were found to have SDR-sex values which were higher than the partitioning criteria (SDR ≥ 0.4). Age-related changes with SDR ≥ 0.4 was noted only for RBC in males. Table 4.22: Standard Deviation Ratios for Sex, Age and Ethnicity in Haematological Analytes Analyte n SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF WBC 500 0.09 0.10 0.06 0.13 0.14 0.19 0.00 MPV 489 0.04 0.06 0.12 0.00 0.14 0.16 0.13 PLT 500 0.46 0.18 0.00 0.26 0.00 0.00 0.13 Neu % 500 0.08 0.09 0.09 0.09 0.00 0.00 0.00 Neu 497 0.13 0.04 0.00 0.10 0.09 0.18 0.00 Lym % 500 0.00 0.13 0.12 0.14 0.00 0.00 0.00 Lym 500 0.09 0.13 0.15 0.00 0.00 0.00 0.14 Mon % 499 0.19 0.00 0.00 0.00 0.19 0.22 0.15 Mon 498 0.00 0.03 0.13 0.00 0.16 0.00 0.27 Eos % 497 0.23 0.00 0.00 0.00 0.10 0.11 0.09 Eos 499 0.15 0.00 0.10 0.00 0.07 0.07 0.08 Bas % 496 0.00 0.00 0.02 0.00 0.15 0.15 0.15 Bas 497 0.00 0.00 0.21 0.00 0.19 0.05 0.27 RBC 500 1.17 0.32 0.41 0.00 0.00 0.00 0.00 Hb 498 1.48 0.19 0.22 0.14 0.00 0.00 0.00 Ht 500 1.37 0.14 0.21 0.00 0.16 0.18 0.15 MCV 497 0.00 0.20 0.28 0.06 0.00 0.00 0.00 MCH 496 0.08 0.16 0.24 0.00 0.16 0.17 0.13 MCHC 500 0.38 0.00 0.00 0.05 0.30 0.36 0.23 RDW 497 0.28 0.00 0.10 0.00 0.13 0.14 0.11 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 149 University of Ghana http://ugspace.ug.edu.gh 4.7 Clinical Chemistry, Kidney and Liver function In clinical chemistry analytes, between-sex differences were observed in Ca, CK, UA, and C3 with SDRsex ≥ 0.4. The magnitude of Ca, CK, UA and C3 were 0.41. 0.59, 0.97, 0.49 respectively. Furthermore, age-related differences were observed in Ca (SDRage = 0.41) with specific age partitioning required in females (SDRage F = 0.45) (Table 4.23). Table 4.23: Standard Deviation Ratios for Sex, Age and Ethnicity in Clinical Chemistry Analytes Analyte n SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF Ca 499 0.41 0.41 0.36 0.45 0.00 0.00 0.00 cCa+ 498 0.00 0.14 0.00 0.26 0.15 0.16 0.14 Mg 500 0.00 0.12 0.07 0.16 0.22 0.09 0.31 P 497 0.33 0.00 0.00 0.00 0.34 0.24 0.44 CK 498 0.60 0.17 0.07 0.24 0.13 0.16 0.09 LDH 499 0.00 0.00 0.00 0.00 0.26 0.26 0.27 AMY 497 0.27 0.18 0.00 0.30 0.00 0.00 0.00 UA 497 0.97 0.28 0.16 0.39 0.00 0.00 0.06 CRP 497 0.25 0.21 0.21 0.20 0.11 0.16 0.00 C3 499 0.49 0.13 0.15 0.09 0.00 0.00 0.00 C4 497 0.37 0.24 0.25 0.22 0.02 0.00 0.10 Glu 496 0.09 0.21 0.18 0.24 0.09 0.13 0.00 HbA1c 500 0.00 0.24 0.20 0.29 0.19 0.20 0.18 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 4.7.1 Age, Sex and Liver Function As shown in Table 4.24, there was between-sex difference (SDRsex = 0.62) and age- related changes (SDRage = 0.43) found in albumin. The age-related changes were noted only in males (SDRageM = 0.48). Furthermore, Tbil and Dbil were also noted for between- sex variations with both analytes recording SDRsex = 0.59 and SDRsex = 0.60 respectively. Similar observation of between-sex differences was found in AST (SDRsex = 0.62), ALT (SDRsex = 0.68) and GGT (SDRsex = 0.54). 150 University of Ghana http://ugspace.ug.edu.gh Table 4.24: Standard Deviation Ratios for Sex, Age and Ethnicity in Liver function analytes Analytes Analyte n SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF TP 500 0.00 0.37 0.37 0.38 0.00 0.06 0.00 Alb 500 0.62 0.43 0.48 0.36 0.01 0.00 0.06 Glb 500 0.17 0.22 0.05 0.30 0.00 0.14 0.00 TBil 499 0.59 0.00 0.00 0.00 0.14 0.09 0.19 DBil 498 0.60 0.00 0.00 0.00 0.00 0.00 0.12 AST 499 0.62 0.05 0.13 0.00 0.14 0.00 0.24 ALT 497 0.68 0.18 0.23 0.05 0.14 0.19 0.00 ALP 498 0.00 0.25 0.00 0.37 0.19 0.00 0.31 GGT 496 0.54 0.08 0.05 0.10 0.12 0.14 0.07 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 4.7.2 Age, Sex and Kidney Function Amongst kidney function tests, it was found that between-sex variation was significantly high in creatinine (SDRsex = 1.16). Yet for eGFR age-related changes (SDRage = 0.60) was noted with both male (SDRageM = 0.48) and female (SDRageF = 0.68) showing significant age-related changes. Moreover, the age-related changes that was found in Na (SDRage = 0.53) was only noted in females (SDRageF = 0.81). Chloride also showed significant between-sex difference (SDRsex = 0.57). Interestingly, between-ethnicity difference in ANGAP was observed in only females (SDRethnF = 0.44) (Table 4.25). Table 4.25: Standard Deviation Ratios for Sex, Age and Ethnicity in Kidney Function Analytes Analyte n SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF Urea 500 0.23 0.24 0.00 0.35 0.11 0.07 0.13 Cre 500 1.16 0.21 0.00 0.33 0.03 0.15 0.00 eGFR 498 0.00 0.60 0.48 0.70 0.00 0.12 0.00 Na 498 0.00 0.53 0.10 0.81 0.20 0.22 0.16 K 498 0.00 0.10 0.00 0.23 0.17 0.17 0.17 Cl 500 0.57 0.14 0.21 0.00 0.18 0.01 0.26 TCO2 500 0.32 0.33 0.31 0.35 0.11 0.00 0.18 ANGAP 500 0.06 0.00 0.15 0.00 0.38 0.32 0.44 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 151 University of Ghana http://ugspace.ug.edu.gh 4.7.3 Age, Sex, Lipids, Immunoglobulins, Iron and Vitamins Among the lipid analytes, triglyceride is the only analyte that shows significant age- related changes in females (SDRageF = 0.43). There was apparently no between sex- variations among the lipid analytes. Regarding immunoglobulins no age-related difference were observed among the analytes, but there is between-sex difference observed in IgM with SDRsex = 0.49. In iron analytes between-sex difference was found in Fe (SDRsex = 0.7), yet for ferritin both between-sex difference (SDRsex = 0.73) and age-related changes (SDRage = 0.44) were observed. It is noteworthy that the age-related change in ferritin is found only in females (SDRageF = 0.54). Folate had ethnicity-related difference (SDRethn= 0.42) and this difference was only revealed in females (SDRethnF = 0.49). 152 University of Ghana http://ugspace.ug.edu.gh Table 4.26: Standard Deviation Ratios for Sex, Age and Ethnicity in Lipids, Immunoglobulins, Iron and Vitamin Analytes LIPIDS PROFILE Analyte n SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF TC 500 0.00 0.24 0.12 0.33 0.15 0.10 0.19 TG 500 0.15 0.35 0.28 0.43 0.13 0.18 0.03 HDL-C 498 0.31 0.00 0.00 0.00 0.15 0.18 0.10 LDL-C 499 0.00 0.21 0.13 0.28 0.04 0.00 0.11 NonHDL 499 0.00 0.25 0.18 0.32 0.05 0.00 0.10 CHOL/HDL 500 0.21 0.19 0.18 0.21 0.00 0.00 0.00 HDL/LDL 500 0.13 0.17 0.17 0.17 0.00 0.00 0.00 IMMUNOGLOBULINS Analyte N SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF IgG 499 0.00 0.08 0.04 0.09 0.00 0.00 0.09 IgA 500 0.00 0.19 0.26 0.06 0.12 0.00 0.21 IgM 496 0.49 0.16 0.00 0.24 0.00 0.06 0.00 IRON AND VITAMINS ANALYTE Analyte N SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF Fe 500 0.53 0.00 0.00 0.00 0.00 0.00 0.00 Ferr 500 0.73 0.44 0.33 0.54 0.00 0.01 0.00 TF 499 0.15 0.34 0.26 0.39 0.00 0.00 0.00 VitaminB12 500 0.23 0.00 0.00 0.00 0.20 0.18 0.23 Folate 497 0.16 0.00 0.00 0.00 0.42 0.33 0.49 Vitamin D 483 0.00 0.21 0.27 0.05 0.00 0.07 0.00 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 4.7.4 Age, Sex, Tumour Markers and Endocrinology Among tumour markers, it was found that CA125 showed between-sex difference (SDRsex = 0.54) and age-related variations in females (SDRageF =0.43). Moreover, PSA which is a male-specific hormone also showed significant age-related variations (SDRageM = 0.47). Also, in the female specific hormones/analytes there was significant age-related changes. These included estradiol (SDRageF = 0.98), progesterone (SDRageF = 0.51), LH (SDRageF = 0.85), FSH (SDRageF = 1.50) and PRL (SDRageF = 0.58). Also, in other 153 University of Ghana http://ugspace.ug.edu.gh analytes such as FT3 and cortisol between-sex differences were observed with SDRsex = 0.59 and SDRsex = 0.60 respectively. The age-related variations (SDRage = 0.62) that were observed in PTH reflected in both males (SDRageM = 0.69) and females (SDRageF = 0.56). Table 4.27: Standard Deviation Ratios for Sex, Age and Ethnicity in Tumour markers, Endocrine and Thyroid Analytes TUMOUR MARKERS Analyte N SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF AFP 486 0.23 0.25 0.29 0.17 0.06 0.00 0.12 CEA 486 0.11 0.29 0.31 0.25 0.11 0.00 0.23 CA125 483 0.54 0.33 0.00 0.43 0.00 0.02 0.00 PSA 261 0.47 0.00 ENDOCRINE STUDIES AND THYROID Analyte N SDRsex SDRage SDRageM SDRageF SDRakan SDRakanM SDRakanF Estradiol 0.26 0.98 0.04 0.00 Progester 219 0.51 0.00 one LH 217 0.85 0.00 FSH 219 1.49 0.00 PRL 216 0.58 0.21 GH 486 0.23 0.18 0.16 0.18 0.10 0.00 0.21 TSH 477 0.00 0.00 0.00 0.00 0.18 0.10 0.23 FT4 484 0.00 0.09 0.13 0.00 0.00 0.00 0.19 FT3 485 0.59 0.08 0.18 0.00 0.25 0.26 0.24 TPOAb 168 0.12 0.00 0.01 0.00 0.00 0.05 0.00 TgAb 193 0.19 0.07 0.00 0.10 0.00 0.00 0.00 PTH 482 0.00 0.62 0.69 0.56 0.00 0.00 0.00 Cortisol 481 0.60 0.19 0.24 0.09 0.00 0.00 0.00 Insulin 479 0.28 0.00 0.05 0.00 0.10 0.00 0.19 SDR ≥ 0.4 are in bold: an indication that there is a need for partitioning according sex, age, or ethnicity 4.8 Graphical Presentation of SDRage and SDRsex in Selected Analytes. Figure 4.1 shows the between-sex differences and age-related changes in the various analytes. The blue-coloured bars represent males RIs while the red-coloured bars represent females RIs. Moreover, the age categories are arranged in hierarchical order and defined on the y-axis in the figures below. 154 University of Ghana http://ugspace.ug.edu.gh 4.8.1 Clinical Chemistry. As depicted in Figure 4.1, Ca showed significant between-sex difference (SDRsex = 0.41) as well as age-related variation for females (SDRageF = 0.45). Even though, the trend in males shows some level of decline in Ca with ageing, the effect is not strong enough (SDRageM = 0.36) to warrant partitioning. All the remaining analytes – TCO2, ANGAP, cCa+, Mg and P did not show any between-sex and age-related changes as their respective bars appear to be in alignment with each other. 4.8.1.2 Kidney Function As shown in Figure 4.2, creatinine and chloride presented between-sex differences with males recording higher RI values than females in creatinine whereas in chloride males record lower RI values than females. Notably, eGFR showed obvious age-related changes; thus, in both males (SDRageM = 0.48) and females (SDRageF = 0.68) eGFR decreased with ageing. Similarly, Na showed age-related variation but only in females (SDRageF = 0.81). It is also worth noting that neither between-sex differences nor age-related changes were observed in urea and K. 4.8.1.3 Liver Function Figure 4.3 shows between-sex differences in AST, ALT and GGT with corresponding SDRsex values of 0.62, 0.68 and 0.54 respectively. However, LDH and ALP did not show any between-sex and age-related difference. Consequently, their combined male and female RI can be adopted for use. Also, in Figure 4.4, albumin which has an SDRsex = 0.62 shows the need to partition RIs between males and females. Furthermore, it is revealed that amongst males, albumin levels 155 University of Ghana http://ugspace.ug.edu.gh decrease with ageing and this is confirmed with the age-related changes (SDRageM = 0.48). It is also observed in TBil, Dbil, and UA that the males have higher reference values than females with all these having SDRsex > 0.4 and confirming the need to partition RIs between males and females. Thus, the use of sex-specific RIs will be more appropriate than using the combined male and female RI for these analytes. 4.8.1.4 Lipids Profile. From Figure 4.5, it is observed that all the analytes except TG do not show any between- sex differences and age-related changes. Thus, all the SDRsex and SDRageM and SDRageF are < 0.4. But for the TG which showed age-related changes, it is only significant in females (SDRageF = 0.43). Thus, it can be observed that TG level increases with ageing. 4.8.1.5 Immunoglobulins. As shown in Figure 4.6, IgM shows significant between-sex differences (SDRsex = 0.49) an indication that shows that males and females must have separate reference intervals. In this, it is further observed that females have reference values that are higher than males. On the other hand, IgG and IgA did not show any sex and age-related differences. 156 University of Ghana http://ugspace.ug.edu.gh Figure 4.1: Graphical representation of SDRage and SDRsex in Selected Clinical Chemistry Analytes 157 University of Ghana http://ugspace.ug.edu.gh Figure 4.2: Graphical representation of SDRage and SDRsex in Selected Kidney Function Tests 158 University of Ghana http://ugspace.ug.edu.gh Figure 4.3: Graphical representation of SDRage and SDRsex in Selected Kidney Function Tests 159 University of Ghana http://ugspace.ug.edu.gh Figure 4.4: Graphical representation of SDRage and SDRsex in Selected Liver Function Tests 160 University of Ghana http://ugspace.ug.edu.gh Figure 4.5: Graphical representation of SDRage and SDRsex in Selected lipids and Diabetes Tests 161 University of Ghana http://ugspace.ug.edu.gh Figure 4.6: Graphical representation of SDRage and SDRsex in Selected Immunoglobulins. 162 University of Ghana http://ugspace.ug.edu.gh 4.9 Derivation and Comparison between LAVE (+) and LAVE (-) methods The use of LAVE to remove latent participants in RI derivation studies is gathering momentum. In this study the researcher examined the differences between RI derived with the application of LAVE (+) and without application of LAVE (-) in both parametric (P) and non-parametric (NP) methods. In the figures presented in this section, the blue-coloured bars represent RIs of males while the red-coloured bars represent RIs of females. Haemoglobin and haematocrit are two haematological analytes that showed evident variations between LAVE (+) and LAVE (-) particularly in the LLs. In females for instance, the LAVE (-) P = 9.6 g/dL for Hb and LAVE (-) P = 32.2% for Ht while LAVE (+)P = 10.7 g/dL for Hb and LAVE(+)P = 34% for Ht as shown in Figure 4.7. Figure 4.7: Comparison between LAVE (+) and LAVE (-) Method in Hb and Ht 163 University of Ghana http://ugspace.ug.edu.gh Figure 4.8: Comparison between LAVE (+) and LAVE (-) Method in MCV, MCH, RBC and Folate In Figure 4.8, MCV and MCH also show clear variations between the LAVE (+) and LAVE (- ). In the RIs derived, it is seen that all the RIs that are LAVE (+) related have narrower intervals compared with LAVE (-) related RIs. It is further revealed that the LLs of the RIs contributed to the differences observed (see Figure 4.8). In Folate however, the variation was more evident 164 University of Ghana http://ugspace.ug.edu.gh in the NP method among females where LAVE (-) NP = 32.7 ng/mL but LAVE (+) NP = 28.2 ng/mL Figure 4.9: Comparison between LAVE (+) and LAVE (-) Method in TF, RDW, Fe and Ferr. As shown in Figure 4.9, TF, RDW, Fe and Ferr all showed variations in the RIs that were derived with the application of LAVE and RIs without application of LAVE. Thus, in the TF, it was observed that among males, both the P and NP methods without the application of LAVE recorded UL values (373.5 ng/mL; 368.7 ng/mL) that were higher than when LAVE is applied (345.1 ng/mL; 353.2 ng/mL). Similar observation was made in RDW, but this time in both 165 University of Ghana http://ugspace.ug.edu.gh males and females. For instance, while the LLs were comparable for all the methods, the ULs of LAVE (+) P= 17.3% was lower than LAVE (-) P = 18.6% in females. Moreover, Fe showed variation in females, particularly in the LLs where the value for LAVE (-) P = 3.3 umol/L was significantly lower than LAVE (+) P= 5.5 umol/L. One of the most revealing findings was the RIs of Ferr among females where LAVE (-) P = 4.5 ng/mL was considerably lower than LAVE (+) P =15 ng/mL. Figure 4.10: Comparison between LAVE (+) and LAVE (-) Method in Neutrophil# and CRP. As shown in Figure 4.10 neutrophil absolute RIs varied between LAVE (+) and LAVE (-). Thus, among females LAVE (+) P = 1.21 – 3.69 x109/L while LAVE (-) P = 1.13 – 4.05 x109/L. Similar, trend is observed in males and NP (see Figure 14). A more interesting finding was observed in CRP where among males the UL of LAVE (-) NP = 11.41mg/dL but LAVE (+) NP = 13.24 mg/dL. In contrast, regarding the P method, it can be observed that the UL of LAVE (- ) P = 6.51mg/dL is higher than the UL of LAVE (+) P = 4.94mg/dL. 166 University of Ghana http://ugspace.ug.edu.gh Figure 4.11: Comparison between LAVE (+) and LAVE (-) Method in HDL-C and LDL-C Figure 4.11 presents the variations between LAVE (+) and LAVE (-) in both P and NP methods for CHOL_HDL and LDL-C. CHOL_HDL showed variations between the LAVE (+) and LAVE (-) for both P and NP methods. Thus, ULs of all the LAVE (+) values were lower than the ULs of the LAVE (-) values. Similarly, for LDL-C their UL values showed significant differences in both males and females. Figure 4.12: Comparison between LAVE (+) and LAVE (-) Method in TG and TC 167 University of Ghana http://ugspace.ug.edu.gh Figure 4.12 presents the variations between LAVE (+) and LAVE (-) in both P and NP methods for TG and TC. In these analytes the LLs values were comparable between LAVE (+) and LAVE (-), yet their UL values showed significant differences in both males and females. Figure 4.13: Comparison between LAVE (+) and LAVE (-) Method in ALP, GGT, AST, and ALT As shown in Figure 4.13, there were noticeable differences between the RIs of GGT, AST, and ALT. In each of the analytes there are narrower RIs for all the ones that LAVE (+) was applied compared with the ones without the application of LAVE (-). These narrower RIs observed in the LAVE (+) can be detected in the ULs of these analytes. For example, in GGT among males, 168 University of Ghana http://ugspace.ug.edu.gh the UL of LAVE (-) NP = 159 IU/L which reduces drastically after the application of LAVE (+) NP = 98 IU/L. In addition, for AST in males, the UL for LAVE (-) P = 45.7 IU/L was significantly higher than LAVE (+) P = 38.7 IU/L. Similar observation is made in ALT, however, for ALP the variations are minimal (Figure 4.13). Figure 4.14: Comparison between LAVE (+) and LAVE (-) Method in Cre, UA, Dbil, and Tbil. Figure 4.14 shows that, in creatinine, variations are observed in two main folds. First, among males, the UL of LAVE (-) NP =110 umol/L while LAVE (+) NP = 112 umol/L, with similar 169 University of Ghana http://ugspace.ug.edu.gh trend for the P method. On the other hand, among females the ULs LAVE (-) NP = 87.69 umol/L and LAVE (-) P = 85.98 umol/L were higher than LAVE (+) NP = 83.07 umol/L and LAVE (+) P = 82.22 umol/L respectively. Again, in UA it was found that males RI derived without LAVE non-parametrically was narrower (231 - 470 umol/L) compared with when LAVE was applied (225 – 477 umol/L). Furthermore, in Dbil and Tbil it was found that variations existed between LAVE (+) and LAVE (-) particularly at their ULs (see Figure 4.14). 170 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE DISCUSSION 5.0 Introduction This chapter presents a discussion of the key findings of the study. The chapter is organized into four main subsections that reflect the specific objectives of the study. Thus, firstly, the key findings regarding haematological analytes are discussed. Next, the discussion centres on the chemistry analytes which are grouped specifically into general chemistry analytes, liver function tests, kidney function tests, diabetes screen, lipids profile, immunoglobulins, thyroid studies, vitamin studies, iron studies, endocrine studies and tumour markers. The third section is a combination of the multiple regression analysis and standard deviations ratio results which primarily discuss the effect of factors such as sex, age, BMI, SBP, DBP, ethnicity, ethanol, and hours of standing. The final section discusses the results of the effect of the application of LAVE and non-application of LAVE on the RI results. 5.1 Reference Intervals for Haematological Analytes Haematological parameters are important clinical indicators for examining patients. Particularly, full blood count (FBC) is commonly requested as part of laboratory tests in diagnosis and patient management. Haematological tests are used to screen for varieties of pathological conditions especially for blood disorders such as anaemia and leukaemia, immunological abnormalities and infections. Apart from the clinical and patient management purposes that haematological parameters serve, they are critical in screening volunteers for clinical trials. This study established RIs for haematological parameters that comprise erythrocytes, leucocytes and platelets. 171 University of Ghana http://ugspace.ug.edu.gh The results from the study showed noteworthy variations between males and females in several of the haematological parameters that were studied. These sex-related differences that were observed in Ghanaians adults corroborate previous studies in other countries (Al- Mawali et al., 2018; Bakrim et al. 2018; Kibaya et al., 2008; Mulu et al., 2017; Samaneka et al., 2016; Segolodi et al., 2014; Siraj et al., 2018). More specifically, it was found that RBC, Hb, Ht, eosinophil (%) and (#) were considerably higher in males compared to females. This finding is supported by previous studies in Africa (Bakrim et al., 2018; Dosoo et al., 2012; Mine et al., 2012; Samaneka et al., 2016) and Asia (Al-Mawali et al., 2018). For example, Bakrim et al. (2018) studied adult Moroccan population and found RIs of RBC (4.37-5.96×1012/L) for males versus (3.86 - 5.20 ×1012/L) for females, Hb (13-17.1 g/dL) for males versus (11-14.8 g/dL) for females, Ht (38.3 - 50%) for males versus (33.5 - 43.9%) for females. The interesting observation here is that the differences in hormonal composition between the sexes are identified as the possible reason that accounts for these variations. For instance, the lower levels of erythrocytes in females may be attributed to the effect of erythropoiesis, menstrual loss, as well as parity (Dosoo et al., 2012; Kibaya et al., 2008; Miri-Dashe et al., 2014). It is also noteworthy that between males and females, platelet count was significantly higher among females than males; a finding that is consistent with what is reported in Ethiopia (Mulu et al., 2017), Morocco (Bakrim et al., 2018), Oman (Al-Mawali et al., 2018), Zimbabwe (Samaneka et al., 2016) and Nigeria (Miri-Dashe et al., 2014). The higher platelet count in females than males may be due to the effect of endogenous female sex hormones, mainly the estrogen and progesterone on the activation of platelet 172 University of Ghana http://ugspace.ug.edu.gh formation. Megakaryocytes use 3-hydroxysteroid dehydrogenase to synthesize 17b- estradiol (E2), which regulates platelet formation (Leng & Bray, 2005; Moro et al., 2005). These between-sex differences observed are critical for the establishment and use of RIs. Thus, it is important further analysis is done to ascertain the need to derive sex-specific RIs for these analytes that showed vast differences between males and females (addressed with SDRs in this study). In this study, it was also found that the Hb results of 12.8-17.2g/dL among Ghanaian males and 10.7-14.3 g/dL among females were comparable with those found among Indian males 12.3 – 17g/dL and Indian females 9.9 – 14.3g/dL (Sairam et al., 2014). Yet, compared to Kenyans and Ethiopians, both Ghanaian males and females had significantly lower values of Hb than their Kenyan and Ethiopian counterparts (Omuse et al., 2018; Mulu et al., 2017). The higher levels of Hb in Kenyans and Ethiopians may be due to the effect of higher altitude (Scheinfeldt et al., 2012;Wang & Semenza, 1996; Windsor & Rodway, 2007). In contrast, however, the Hb among Ghanaians was significantly higher in both males and females than what Miri-Dashe et al. (2014) found among adult Nigerians. More fascinating in these findings is the sharp contrast between the interval among Ghanaians males (12.8-17.2g/dL) and the interval among Nigerians (14.0 -14.4 g/dL). The width for the Ghanaian population (4.4 g/dL) is more than ten times the width for the Nigerian population (0.4g/dL). Nigeria is a West African country that has many nutritional, lifestyle and other environmental factors similar to Ghana, so such a huge variation is not expected. Moreover, the interval is extremely narrower which warrants a critical relook into this study. The methods and approach used to derive the said RIs must be re-evaluated. 173 University of Ghana http://ugspace.ug.edu.gh This study found that between males and females the RIs for MCV, MCH, MCHC, neutrophil (% and #), MPV, monocytes (% and #), WBC and basophil (% and #) were similar(see table 4.3). This finding corroborates other studies among African populations in some of the analytes which show comparable results between males and females; for e.g. MCV, MCH, MCHC (Mulu et al., 2018; Omuse et al., 2018); monocytes, neutrophil, basophil, and WBC (Bakrim et al., 2018). In contrast, previous studies in Kenya and Zimbabwe found variations between males and females for these analytes that recorded the similar results among Ghanaians (Kibaya et al., 2008; Samaneka et al., 2018). Moreover, it was found that the upper limit for WBC (7.53 x109/L) in this study contrasts with that found in previous studies in other countries where they are all higher (Ittermann et al., 2010; Osei‐Bimpong et al.,, 2012; Ambayya et al., 2014, Kratz et al., 2004). In Malaysia for instance, the WBC was found to be 4.1–11.4 x109/L with the upper limit significantly higher than what this study found. Further observation was made regarding the fact that lymphocytes (% and #), eosinophil (% and #) and monocytes (% and #) values were higher compared with both the US population (Kratz et al., 2004) and an existing manufacturer’s guide (Sysmex XN-1000) which is in use in Ghana (appendix III). This finding is consistent with previous studies among African populations that revealed constantly higher reference values in these parameters than that of the manufacturer’s insert (Sysmex XN-1000) (Bakrim et al., 2018; Kibaya et al., 2008; Samaneka et al., 2016; Omuse et al., 2018). This consistency of higher reference values among African populations than their western counterparts may be attributed to genetic/racial and/or environmental factors. 174 University of Ghana http://ugspace.ug.edu.gh Interestingly, other studies among Ghanaian populations that established haematological analytes report RI values that showed significant variations with what this study found. For instance, a study in Ghana by Dosoo and colleagues reported Hb levels among males and females as 11.3–16.4 g/dL and 8.8–14.4 g/dL respectively (Dosoo et al., 2012). In other studies in Ghana, Hb was 11.7-16.5 g/dL and 9.1 – 14.0 g/dL (Koram et al., 2007) and 10.7-18.8 g/dL and 8.2-16.2 g/dL (Addai-Mensah et al., 2019) for males and females respectively. Similarly, variations were observed for platelet count, eosinophil (%), Ht and RBC between the current study and those reported above among healthy adult Ghanaian populations (Addai-Mensah et al., 2019; Dosoo et al., 2012; Koram et al., 2007). In all, the observation was that these analytes presented significantly wider intervals compared with what this study derived. Generally, RI derivation studies use eligibility criteria that recruit apparently healthy individuals. Even though the application of apriori inclusion criteria may exclude several unhealthy volunteers, it is still possible some participants with latent infection may be included in the derivation of the RIs. The differences observed within these Ghanaian population may be attributed to the application of the LAVE method in our study. LAVE is a robust statistical method used as a secondary exclusion technique that removes participants with latent conditions such as anaemia and/or ongoing inflammation. Hence, it reduces the effect of latent diseases on RI derivation. It is noteworthy, that the role of LAVE in improving RIs has been confirmed in several studies (Ichihara et al., 2017a; Ichihara et al., 2014; Ozarda et al., 2014; Yamamoto et al., 2013; Ichihara et al., 2013). In the previous studies conducted in Ghana, there was no application of LAVE as a secondary exclusion technique to eliminate individuals that may have been included but carrying latent conditions such as anaemia and/or ongoing inflammation. Consequently, that will 175 University of Ghana http://ugspace.ug.edu.gh lead to RIs with wider ranges that are influenced by these latent conditions. A situation that is observed in the results of these previous studies conducted among Ghanaians. 5.2 Reference Intervals for Clinical Chemistry Analytes This study aimed at establishing reference intervals for biochemical laboratory indices among healthy Ghanaian adults aged 18 – 65+ years using internationally accepted CLSI guidelines (IFCC, 2008). This was necessary for the establishment of a potential standard that can be used for interpretation of laboratory test results which will be relevant for participants screening for clinical trials and patient care and management generally. Findings from the study showed significant variations that are sex and age-related, as well as variations from other studies from different populations in both African and western countries. Consistent with previous studies, this study revealed that males have significantly higher reference values than females for UA and AMY (Abebe et al., 2018; Kibaya et al., 2008; Mekonnen et al., 2016; Saathoff et al., 2008). Thus, the higher UA values in males than in females have alluded to the fact that females have higher clearance rate (thus the rate at which waste substances are cleared from the blood by the kidney) than males (Jannie et al., 1979). More so, in a multicenter study in China, Xia et al. (2016) found RIs of UA that are significantly higher across the six regions included in their study; with male RIs exceeding 500 umol/L for the UL and that of females exceeding 400 umol/L. Indeed, these findings contradict what is reported in this study. Nonetheless, compared with the Japanese (Yamakado et al., 2015), the RIs were relatively higher among Ghanaians. 176 University of Ghana http://ugspace.ug.edu.gh Conversely, females recorded higher RIs for LDH than in males (see table 4.2). This finding corroborates previous reports of reference values in LDH among adult Kenyans and Ethiopians (Kibaya et al., 2008; Mekonnen et al., 2017). The results however contradict what is reported among adult Tanzanians (Saathoff et al., 2008). Regarding CRP, a more interesting revelation is about the contrasting reference values recorded between males and females in comparison with the existing manufacturer’s insert (Beckman Coulter guide) currently in use in Ghana. Thus, while the RI of Ghanaian females (0.47-9.85 mg/dL) is comparable with that of the manufacturer (0-10 mg/dL), among males the UL of the manufacturer’s insert (10 mg/dL) is approximately twice the UL of Ghanaian males (4.94 mg/dL). It was also noted that the RI for LDH among Ghanaians is significantly higher than what is currently used in Ghana (Beckman Coulter insert) (see Appendix III). Interestingly, the same observation is made for the other studies in Africa (Kenya and Ethiopia) mentioned above, which could be an important pointer to the role of race or environment for these observed differences. Obviously, CK had significantly higher RI in males than females a trend that has been commonly reported elsewhere (Dosoo et al., 2012; Eller et al., 2008; Neal, Ferdinand, Yčas, & Miller, 2009; Samaneka et al., 2016; Saathoff et al., 2008). It was also noted that this study recorded values of CK that differ markedly from studies conducted in other regions such as North America (Neal et al., 2009), Europe (Brewster, Coronel, Sluiter, Clark, & van Montfrans, 2012), and Asia (Xia et al., 2016). The obvious variation that is observed across regions have been attributed to the possible cause of racial and genetic factors (Brewster et al., 2012; Gledhill, Van der Merwe, Greyling, & Van Niekerk, 1988; Neal et al., 2009; Wong et al., 1983). More remarkably, the results of this study present higher RI of CK that is more than the manufacturer’s insert (Beckman Coulter) currently 177 University of Ghana http://ugspace.ug.edu.gh used by some laboratories in Ghana. Undeniably, such disparity impacts clinical decisions and affects case management and resource use in the regions that adopt “foreign-derived” reference intervals for use among local populations. 5.2.1 Reference Intervals for Kidney Function Analytes. Higher reference ranges for creatinine in males compared to females have been reported in studies done in Japan, Turkey, and Ethiopia (Abebe et al. 2018; Yamakado et al., 2015; Ozarda et al., 2014). These findings support what is found among Ghanaians in this study. For creatinine being higher in males than females, it is believed to be an expected phenomenon internationally, because of the greater skeletal, muscle and bone mass in males (Taylor, Fink, & Pappas, 1989). The RI for urea was found to be the same between males and females. Apparently, this contrast with what is reported by previous studies (Abebe et al., 2018; Ichihara et al., 2013; Eller et al, 2008) but agrees with (Kibaya et al., 2008). Furthermore, it was noted that in contrast with what is reported among Turkish (Ozarda et al., 2014), Americans (Kratz et al, 2004), and Asians (Borai et al, 2016; Ichihara et al., 2013), Ghanaian RIs for urea were lower. However, they were comparable or relatively higher than what is reported among other African populations (Saathoof et al., 2008; Dosoo et al., 2012; Kibaya et al., 2008). However, in contrast with other studies that found significant differences between males and females regarding the reference intervals of sodium and potassium (Eller et al., 2008; Saathoff, Schneider, Kleinfeldt, Geis, Haule, et al., 2008; Samaneka et al., 2016) this study found no sex-related differences regarding these analytes. Moreover, it was observed that RI results for sodium is consistent with what is reported among Chinese, Saudi Arabians, 178 University of Ghana http://ugspace.ug.edu.gh Americans and Tanzanians (Xia et al., 2016; Borai et al., 2016; Kratz et al., 2004; Saathoff et al., 2008). On the other hand, it contrasts with what is reported in Kenya with the RI (141.8 – 152.1 mmol/L) considerably higher (Kibaya et al., 2008) than that of Ghanaians (136 -143 mmol/L). Chloride is comparable with what is reported in Zimbabwe (Samaneka et al., 2018) Kenya (Kibaya et al., 2008) Botswana (Segolodi et al., 2014) and Uganda (Eller et al., 2008). Elsewhere in Turkey (Ozarda et al., 2014) and Saudi Arabia (Borai et al., 2016) these are RIs for Cl are in consonance with what is reported among African populations. Thus, there seems to be a trend where Cl levels are relatively stable across the world. Confirmation of this trend may be illustrated by the multicenter study that established RI for selected South- East Asian countries (Malaysia, Indonesia, Hong Kong, Taiwan), Japanese and Chinese (Ichihara et al., 2013). In this, the RI for Cl was found to be 101-108 mmol/L; a range that is similar to the 99 – 108 mmol/L recorded among Ghanaians. Congruent with the aforementioned assertion, the RI for the manufacturer’s guide (Beckman Coulter) is also in the same region of 101-109 mmol/L (Appendix III). eGFR results in this study sharply contrast with what is reported among Japanese. Here the RI for adult Japanese is markedly narrower (Yamakado et al., 2015) than what is found among Ghanaians. Particularly, in the UL, while the value for Japanese is 113 mL/min that of Ghanaians is 188 mL/min. One more interesting revelation is the contrast between the manufacturer’s insert (Beckman Coulter) which only gives an LL but not a UL (>60 mL/min). Consequently, physicians interpret this to mean any observed value that is greater than 60 mL/min is normal. In fact, this contradicts what this study found because 179 University of Ghana http://ugspace.ug.edu.gh it has a UL and other studies also confirm eGFR has a UL (Yamakado et al., 2015; Delanaye et al., 2012; Pottel et al., 2017). 5.2.2 Reference Intervals for Liver Function Analytes. Both the RIs of Tbil and Dbil in males were noticeably higher compared with their female counterparts. This finding is supported by previous studies in Botswana, Tanzania, Kenya, Ethiopia and Zimbabwe (Samaneka et al., 2016; Saathoff et al.2008; Segolodi et al., 2014; Abebe et al., 2018; Kibaya et al., 2008). Moreover, compared with the current study both the RIs of Tbil and Dbil among Ghanaians is considerably lower in males and females compared with RIs of Tbil and Dbil reported in Kenyans, Botswanans, and Ethiopians. On the contrary, RIs of Tbil and Dbil are lower than what is reported by Kratz (2004) among Americans. Other liver function test such as AST, ALT, GGT were also higher in males than in females while no sex-variation was observed for ALP. Again, previous studies have found such variations in African populations (Samaneka et al., 2016; Saathoff et al., 2008; Segolodi et al., 2014; Abebe et al., 2018; Kibaya et al., 2008), Asians (Borai et al., 2016; Yamakado et al., 2015; Xia et al., 2016;), Europeans (Ozarda et al., 2014). Moreover, in terms of the RI results for AST, ALT and GGT, they contrast with what is reported in Turkey (Ozarda et al., 2014) as the Ghanaian reference values are significantly higher. Similarly, compared with Japanese (Yamakado et al., 2015) and Saudi Arabians (Borai et al., 2016) the RIs are higher among Ghanaians for these analytes mentioned above. Finally, the RI for TP found in this study did not show any sex-related variation. This finding corroborates previous findings (Saathoff et al., 2008; Abebe et al., 2018; 180 University of Ghana http://ugspace.ug.edu.gh Mekonnen et al., 2016; Yamakado et al., 2015; Borai et al., 2016). Notably, the TP reference values are relatively comparable with some previous studies among Turkish (Ozarda et al., 2014) but contrast with both RIs of Ethiopians (Abebe et al., 2018) and Japanese (Yamakado et al., 2015). Thus, the RI for Ghanaians is significantly lower than that of Ethiopians but higher than that of Japanese. Albumin an important parameter was found to present relatively narrower RI than what is found among Asians (Ichihara et al., 2013; Borai et al., 2016), Ugandans (Eller et al., 2008), Turkish (Ozarda et al., 2014), and Tanzanians (Saathoff et al, 2008). Yet, for Ethiopians and Americans, the difference was markedly wider. Thus, the approximate interval for Ethiopians was more than twice (37 – 62 g/L) (Abebe et al., 2018) while that of Americans is approximately twice (35 - 55 g/L) (Kratz et al., 2004) that of Ghanaians (37.7 – 48.3 g/L). These observed differences are not readily explicable in the literature as the possible cause of such variation has not been identified yet. The most critical observation here among the analytes for liver function tests is that all the RI results vary considerably from that of the Beckman Coulter’s insert. It is worth noting that, while variations exist in all the analytes, the Dbil, AST, ALT and GGT did not have any specified LLs in the manufacturer’s insert (see appendix III), the Ghanaians had LLs. The ramifications here are that there is a high possibility of many Ghanaians being misdiagnosed using these “foreign-derived” RIs. Consequently, resources will be wasted from both the patients’ perspective and society’s perspective. 181 University of Ghana http://ugspace.ug.edu.gh 5.2.3 Reference Intervals for Diabetes Screen Analytes. Glucose and HbA1c are the two main parameters used in screening for diabetes. In this study, no sex-variations were observed in the RIs for Glu and HbA1c analytes. Regarding Glu, previous studies among Kenyans support this finding. Meanwhile, amongst Chinese (Xia et a., 2016), Japanese (Yamakado et al., 2015), and Turkish (Borai et al., 2016) there exist variations in the RIs of Glu in males and females. Moreover, it was observed that the RIs across these studies are not distinctively farther from each other except for Borai et al. (2016) which reports RI values that are closer to 7.0 mmol/L. Interestingly, Glu remains one of the few analytes that have RI values that are almost identical with the Beckman Coulter standard which is currently used in Ghana. The use of HbA1c remains relatively new as traditionally, clinical screening for diabetes has relied on glucose. However, the use of HbA1c has grown in prominence over time and it is used as a surrogate or complementary test for diabetes. Currently, several medical associations including Australian and American Diabetes Association (ADA) recommend a cut-off of ≥6.5% for testing diabetes. Admittedly, they agree that individuals with an HbA1c of 39–5.7–6.4% are at ‘increased risk’ for diabetes as well as cardiovascular disease (American Diabetes Association, 2010). In this study, it was found that RI for HbA1c which is 4.24 – 6.27%; has an UL that is relatively lower than what is recommended by the ADA. On the other hand, when compared with what is reported among Japanese the range is relatively narrower (4.7 – 6.07%) (Yamakado et al., 2015). It is also noteworthy that the findings obtained in this study are comparable with the Beckman Coulter standard (appendix III). 182 University of Ghana http://ugspace.ug.edu.gh 5.2.4 Reference Intervals for Lipids Profile. Furthermore, the RI of TC among the Ghanaian population for both males and females are consistently higher than that of Kenyans (Kibaya et al., 2008), Nigerians (Miri-Dashe et al., 2014), Zimbabweans (Samaneka et al., 2016), Turkish (Ozarda et al., 2014), Chinese (Xia et al., 2016), Saudi Arabians (Borai et al., 2016) and the USA (Kratz et al., 2004). Interestingly, while the RI of TC have LLs for both males and females Ghanaians, among USA population any RI value below 5.17mmol/L is considered to be normal (Kratz et al., 2004), these two RIs present different implications for clinical decision making and patient management. Moreover, fairly the same RIs for males and females were recorded in this study. This finding corroborates previous studies that reported comparable RIs of TC in females and males (Abebe et al., 2018; Samaneka et al., 2016; Ozarda et al., 2014). HDL-C was found to be relatively higher in females than in males. This finding is confirmed in other studies in China, Japan and Turkey (Xia et al., 2016; Yamakado et al., 2015; Ozarda et al., 2014). The RI of HDL-C in a multicenter study among Japanese conducted by Yamakado et al. (2015) found values that were pointedly higher in both males and females compared to what this study found among Ghanaians. Probably confirming this difference of higher RI of HDL-C in Asians than Africans is the multicenter study that included South-East Asian countries (Malaysia, Indonesia, Hong Kong, Taiwan), Japanese and Chinese (Ichihara et al., 2013). The study reports an average RI of 1.07 - 2.55 mmol/L albeit regional variations of RIs between these countries. On the other hand, compared to previous reports among Turkish, the UL of this analyte is appreciably higher among Ghanaians (Ozarda et al, 2014). 183 University of Ghana http://ugspace.ug.edu.gh This study found that while the LLs of TG were identical for males and females, the UL was markedly raised in males (1.95 mmol/L) than in females (1.64 mmol/L). This is confirmed in a multicenter study among Asians (Ichihara et al., 2013), in addition to other previous studies in China (Xia et al., 2016), Turkey (Ozarda et al., 2014), India (Sairam et al., 2014) and Saudi Arabia (Borai et al., 2016). Of key note here is the considerably higher RI recorded among these Asian populations particularly the UL values. For instance, among Chinese and Saudi Arabian males, the UL values of TG are almost twice (3.58 mmol/L) the UL of the RI for Ghanaian males. Moreover, previous studies among African populations have equally recorded RI values that present appreciably higher values in the UL (Abebe et al., 2018; Kibaya et al., 2008; Eller et al., 2008) than found among Ghanaians. These variations may be due to the lifestyle and ethnicity differences between these study populations (Xia et al., 2016). In addition, it was observed that all the analytes under consideration for the lipids have reference values that vary appreciably from the Beckman Coulter standard (appendix III). Thus, all the RIs for lipid analytes were higher in Ghanaians except for TG that was relatively similar to the Beckman Coulter standard. More interesting is the lack of defined LLs for analytes such as TC, TG, LDL-C, and CHOL_HDL in the Beckman Coulter standard. In contrast, this study recorded defined LLs for these analytes among Ghanaians. This lack of defined LLs and ULs contradicts all previous studies in Asian and African populations mentioned above. At this point, probably the main concern with such a situation may be the ramifications of using such RIs for clinical decision making among local populations. It is obvious that this will lead to inefficiency and waste of medical resources as well as possible further complications of patients’ medical conditions. 184 University of Ghana http://ugspace.ug.edu.gh 5.2.5 Reference Intervals for Immunoglobulins. Immunoglobulins G is one of the analytes that was considered in the Asian Project. In that study even though there were between-region variations in terms of the RIs derived, the average RI for IgG among these Asians was 880 – 1800 mg/dL in males and 950 – 1910 mg/dL in females (Ichihara et al., 2013). This result actually contrasts with what this study found among Ghanaians in two main folds. First, there is variation in the RIs recorded between the two populations, the RI among the Ghanaian population is considerably higher in terms of both the LL and UL in both sexes. Secondly, while the study among Asians show conspicuously higher reference values in females than males, there is no sex- related variation among Ghanaians. It is worth noting that in the above study, the regional- specific RI for South-East Asia (1080 – 2150 mg/dL) was identical to what was found among Ghanaians. In a separate study among American population, Kratz et al. (2004) found the RI of IgG to be 614 - 1295mg/dL. This finding is actually incongruous with the results in this study. The RI for Ghanaians is almost twice (1165 - 2164 mg/dL) that of the American population. In the Asian project (Ichihara et al., 2013), the results showed that IgA had RI that was comparable to what this study recorded among Ghanaians with the ULs of both RIs almost the same. Yet regarding South-East region, again, their RI for IgA was considerably higher especially in the UL. Furthermore, sex-related variation existed for this analyte with females presenting with relatively higher RIs than their male counterparts. This is consistent with the results found among Ghanaians. Meanwhile when the Ghanaian results are compared with previous reports among Americans, the LL is about twice that of Americans while the UL is more than 100 units higher (Kratz et al., 2004). 185 University of Ghana http://ugspace.ug.edu.gh Again, the reference values of IgM were found to be markedly higher in females compared to males. Also, this finding is confirmed by the Asian project where females recorded conspicuously higher RI particularly in the UL than males (Ichihara et al., 2013). More so, in this study IgM recorded RI of 41 – 237 mg/dL for combined male and female. This finding is consistent with previous study among Asians as the results are almost identical (Ichihara et al., 2013) but it contradicts what is reported among Americans where their RI is significantly higher especially the UL (53 - 334 mg/dL) (Kratz et al., 2004). A study conducted over four decades ago among adult Ghanaians that derived RIs for IgG, IgA, and IgM found the following RIs 470 - 2700 mg/dL, 40 - 390 mg/dL and 40 – 310 mg/dL respectively (Riches et al., 1979). In comparison with the current study, the IgG results are extremely narrower contrary to what is reported in the previous study. Moreover, while the UL of IgG is relatively comparable to what is reported in the previous study, the current study records extremely higher LL value than previously reported (Riches et al., 1979). The UL of the RI for IgM in this study is appreciably lower than what is reported previously. Even though, generally, the population used in both studies represent Ghanaians, the earlier study by Riches and colleagues (1979) used male volunteers who are serving officers in the Ghana Armed Forces. It is believed that these group of people have a particular lifestyle that may be different from civilians and this may explain the observed differences in these two studies. In general, the IgG and IgA were significantly raised in this study and the Asian project compared with the Beckman Coulter standard. Thus, this pattern of higher RI values for IgG and IgA in African and Asian populations than Europe may be attributed to a combination of environmental and genetic factors. Moreover, it is reported that the closer 186 University of Ghana http://ugspace.ug.edu.gh an area is to the equator, the higher the serum concentrations of positive inflammatory markers (e.g. IgG); thus, it is assumed that exposure to infectious agents is higher in regions closer to the equator (Ichihara et al., 2013). Hence it is believed that the geographical location of Ghana and these Asian countries may account for the raised levels of IgG in these studies. 5.2.6 Reference Intervals for Iron Studies. This study found sex-related variation in the RI of Fe, with males presenting higher reference values in both the LL and ULs compared to females. This finding agrees with previous studies elsewhere (Bakan et al., 2016; Xia et al., 2016; Borai et al., 2016). Moreover, it was noted that the RI of Fe in Ghanaians are relatively lower than what is reported elsewhere (Bakan et al., 2016; Kratz et al., 2004; Xia et al., 2016; Borai et al., 2016). This observation may be due to nutritional factors such as low intake of iron based food among Ghanaian. Furthermore, in this study RI of ferritin was considerably higher in males (39 – 502 ng/mL) than in females (15 – 300 ng/mL). The raised concentration in males than females is confirmed in other studies (Bosch et al., 2001; Bakan et al., 2016). The reason for the low ferritin concentration among females, mostly those in the reproductive age group than males has been attributed to bleeding and blood loss during menstruation. This study corroborates Kato et al. (2000) studies which found that ferritin concentration in postmenopausal women were twice that of premenopausal women. In accordance with previous studies, the current study found folate levels to be higher in women than men (Ford et al., 1999; Ling et al., 2003). Also, iron levels in females were lower compared to that of males. Considering the UL of the RI found among Ghanaians, it is noted that this 187 University of Ghana http://ugspace.ug.edu.gh value is extremely high counting over 100 units than what is reported in Turkey (Bakan et al., 2016). Transferrin is found to be lower in males than in females particularly in the UL of the RI. Thus, this finding is consistent with what is reported among Asians (Ichihara et al., 2013). Moreover, the reference values found in this study are comparable with the values reported among Asians (Ichihara et al., 2013). Nonetheless, the finding in this study contrasts with reference values reported among Americans. For instance, while the RI for Ghanaians is 196 – 314 mg/dL, that of Americans is notably high, standing at 230 – 390 mg/dL (Kratz et al., 2004). Comparison of RI results for Fe, Ferr, and Tf with Beckman Coulter standard reveals interesting findings. Foremost, the Fe and Tf levels in Ghanaians are considerably lower while than that of the Beckman Coulter standard (appendix III). In contrast, the Ferr level among Ghanaians is pointedly higher with the UL of the females RI as high as twice that of the Beckman Coulter standard (appendix III). Again, the basic understanding here is that the population used to derive the RIs for Beckman Coulter are fundamentally distinct from that of the local population in Ghana. 5.2.7 Reference Intervals for Vitamins Analytes. In this study, it was found that males have reference values of vit. B12 that are lower than females. This finding contradicts what is reported in previous studies among Scandinavians. Because among Norwegians males had higher RI than females (Schwettmann & Berbu, 2015) and among the Swedish, no sex-related difference was found in vit B12 (Wahlin et al., 2002). Additionally, it was found that the reference values 188 University of Ghana http://ugspace.ug.edu.gh for vit B12 were extremely higher in Ghanaians (305 – 1283 pg/mL) than in Norwegians (180.3 - 806.4 pg/mL) and Spanish (213.8 and 763.3 pg/mL) (Schwettmann & Berbu, 2015; Calderon et al., 2018). In contrast, a study in Uganda found reference values of Vit B12 that recorded significantly lower LL that is twice smaller than among Ghanaians. Yet, the UL was relatively closer to that of Ghanaians (117-1158pg/ml) (Galukande et al., 2011). It is also interesting to note that compared to the Beckman Coulter standard, the Ghanaian values remained significantly high (appendix III). Thus, congruent with the Ugandan results it is indicative that the trend may continue to be higher among Africans than Europeans. Hence, the possible reason of nutrition and environmental factors may be responsible for this difference. Folate results indicated that females have higher concentration levels than males. This evidence is shown in the conspicuous variation in the ULs of the two groups (males = 14.1 vs females = 19.3 ng/mL). This finding corroborates previous studies in Turkey (Bakan et al., 2016) but contradicts what is reported in Spain, where no sex-related variation was observed in the reference values for folate (Calderon et al., 2018). Furthermore, the observation that reference value among Norwegians was 2.3 -12.8 ng/mL (Schwettmann & Berbu, 2015) contrasts sharply with the results among Ghanaians as the Norwegians values remain low. However, the results of Ghanaians compare favorably with the RIs of Americans and Spanish (Kratz et al., 2004; Calderon et al., 2018). In the Ugandan study, however, both the LL and UL (4.17 – 20ng/mL) are relatively raised compared to Ghanaians (Galukande et al., 2011). 189 University of Ghana http://ugspace.ug.edu.gh 5.2.8 Reference Intervals for Endocrine Analytes. In this study, prolactin was found in only females. Interestingly, it contrasts with other studies elsewhere. For instance, an RI of 2.73 – 15.1 ng/mL concentration was found among males and 3.88 – 40.1 ng/mL among females who were in the follicle stage (Schüring et al., 2015). Similarly, Whitehead et al. (2015) found 2.72 – 19.67 ng/mL concentration among males and 2.96 – 26.36 ng/mL among females. The results of these two studies are identical in males; with males recording a LL of 2.72 ng/mL between them. However, no such result is recorded among Ghanaians. On the other hand, the female results reported by Schüring et al. (2015) is similar to what is found among Ghanaian females (4.3 - 41.1 ng/mL). The results of no prolactin concentration among Ghanaian males is not too surprising as prolactin is a predominantly female hormone. On the other hand, however, sometimes it is possible to observe some predominant features of males in females and vice versa. What is obvious in these two studies that produced RIs in males is that the concentration levels of prolactin are very low compared to females. A more revealing finding is the narrower observation of prolactin RI among Chinese non-pregnant women who were found to have a prolactin concentration of 8.41 – 35.60 ng/mL (Hu et al., 2017). Again, this varies with the Ghanaian results because it is apparently narrower compared to what this study reports. In a multicenter study in Asia involving over seven countries including Japan and China, the estradiol concentration was found to be 50 – 840 pmol/L (Ichihara et al., 2013). However, this result varies significantly from what is found among Ghanaians as the UL of the Ghanaian results is approximately 50% higher than the UL of this previous study. The Beckman Coulter stand shows the variation of the concentration levels with respect to the menstrual cycle. Here, the lowest RI is reported at the follicular phase with the LL 190 University of Ghana http://ugspace.ug.edu.gh of 45.4 pmol/L and the highest UL reported as 1461 pmol/L during ovulation. Hence, the overall range for the Beckman Coulter standard is 45.4 - 1461 pmol/L which has a UL that is slightly higher than what is found among Ghanaian women (32 – 1262 pmol/L). Racial differences have been adjudged as one of the main causes of this sex hormone differences (Rohrmann et al., 2007; Pinheiro et al., 2005; Kim et al., 2012). Hence, in these contrasting findings between the Ghanaian and other populations results, the possible explanation could be the racial differences. Like prolactin, there is also no RI for males regarding estradiol among Ghanaians even though some RI is reported among Asian males albeit very low (66 -140 pmol/L). Progesterone, one of the commonest hormones among females recorded RI of 0.1 – 79 nmol/L in this study. Again, the study among Asians generally have the same LL as that of Ghanaians, yet the UL of Ghanaians is considerably higher than that of the Asians (0.1 – 66 nmol/L) (Ichihara et al., 2013). Moreover, the finding in this study varies with a study conducted by Schuring et al. (2016) which used the Advia Centaur and Immulite 2000XP platforms and found RI of 0 – 71.6 nmol/L and 0 - 60.67 nmol/L respectively. While the Ghanaian value is significantly raised above the two RIs reported by Schuring et al. (2016) it is not too far away from the Beckman Coulter standard (0.18 – 75.9 nmol/L). While the predominant reason that has been attributed to variations in sex hormones concentration levels in different populations has been a racial factor, a question may be asked whether the analyzer platform used has influence in RIs? The testosterone concentration level in Ghanaians was found to be 10 – 41.9 nmol/L. Interestingly, this sharply contrasts with previous studies among Asians that reports the same LL of testosterone concentration among Asian males but significantly lower UL 191 University of Ghana http://ugspace.ug.edu.gh (28.4 nmol/L) than what is found among Ghanaians. Similarly, other studies among Korean, American and European populations have reported RIs of testosterone that are considerably lower (Kim et a., 2019; Friedrich et al., 2008; Travison et al., 2017) compared with what this study found. Elsewhere, while no differences in testosterone concentration were observed between black and white Americans, yet, Mexican-Americans presented a higher concentration of testosterone than whites (Rohrmann et al., 2007). Again, racial factors may be the fundamental reason for these variations observed among various populations. Sex-related variations were observed in cortisol as revealed in this study, with males presenting considerably higher RI (47 -183 ng/mL) than in females (31 – 141 ng/mL). This finding is consistent with what is reported among other populations (Ichihara et al., 2013; Sofer et al., 2016). The contradictory part is the relatively higher RI that is reported among Asians (males = 51 -197 ng/mL and females = 41 -190 ng/mL) than Ghanaians. 5.2.9 Reference Intervals for Tumour Markers. 5.2.9.1 Alpha 1-Fetoprotein (AFP). In this study, the RI of AFP was revealed to be ranging from 1.17 ng/mL to 19.1 ng/mL in males while females had RI ranging from 1.21 ng/mL to 14.9 ng/mL. This finding shows that Ghanaian males have a higher concentration of AFP than their female counterparts. This finding is consistent with Zhang and his colleagues’ study among Chinese aged 20 - 90 years where males recorded higher RIs of AFP than their female counterparts (Zhang et al., 2016). On the other hand, the specific RIs recorded among Ghanaians are significantly higher than what are reported in Zhang et al study. Thus, while the Ghanaian males have RI concentration levels of 1.17 - 19.1 ng/mL, their Chinese counterparts have 192 University of Ghana http://ugspace.ug.edu.gh AFP concentration of 1.31–7.89 ng/mL. Similarly, the females presented AFP levels (1.01–7.10 ng/ml) that vary significantly from that of Ghanaians. Moreover, other studies did not show any distinction between RI of AFP for sex, yet, the ULs of these studies were all significantly lower compared with the UL of Ghanaians (Qin et al., 2011; Christiansen et al., 2001; Bjerner et al., 2008). What is more interesting is the contrasting values between Ghanaians and what is currently in use by some laboratories in the country (Beckman Coulter). In fact, the UL of the manufacturer’s standard is  10 ng/ml which again falls short of the UL among Ghanaians. Even though there is no explicit reason attributed to such variations, it is possible race may be the main factor accounting for this. 5.2.9.2 Carcinoembryonic Antigen (CEA). In previous studies, RIs of CEA have been reported as 0.51 – 4.86 ng/ml in Chinese males and 0.35 – 3.45 ng/ml in Chinese females (Zhang et al., 2016). This previous study supports the finding among Ghanaians which reports a higher concentration of CEA levels in males (0.41 – 4.0 ng/ml) as against 0.46 – 2.8 ng/ml in females. It is interesting to emphasize that the males’ results are relatively comparable to the females’ results which shows slight variations in the ULs. In addition, a study among Chinese males, Qin et al. (2011) found RI values significantly higher than that of Ghanaian males. Specifically, the authors found a UL of CEA to be 5.34 ng/ml and 7.36 ng/ml among different age categories; thus age 20 – 44 years and 45 – 70 years respectively. Interestingly, these values are significantly higher than those found among other Chinese population in Zhang et al. (2016) study. Other studies have also reported UL values that are different from the Ghanaian population, reporting values that range from 3.6 to 6.9 (Bjerner et al., 2008; 193 University of Ghana http://ugspace.ug.edu.gh Behbehani et al., 2002; Kratz et al., 2004). It is important to point out that, the inconsistencies reported may largely be attributed to population differences and background. 5.2.9.3 Carbohydrate Antigen 125 (CA125). Sex-related differences was observed in the RI of CA125 between Ghanaian males and females. Thus, females had higher concentration levels (42 U/mL) of CA125 that was almost three times that of males (15 U/mL). Specifically, in females, a study conducted among Asians reports a UL of 38.3 U/mL (Park et al., 2011); a value that is not too distant from the UL of Ghanaian women. Compared with the Beckman Coulter standard that is currently used in Ghana (see appendix III), the same ULs are used for both males and females contrary to what is found in this study. It is worth pointing out that, such variations and their consequences cannot be underestimated in diagnosis and patient management as well as resource use in the health system. 5.2.9.4 Prostate-specific antigen (PSA) Prostate-specific antigen (PSA) is a male-specific tumour marker which is usually examined with respect to age. In this study, it was found that the Ghanaian male has an average concentration of PSA that range from 0.24 ng/mL to 2.82 ng/mL. Previous studies have concentrated on deriving RI of PSA for male populations that are 40 years and above. For example, in a study among Chinese males, Liu and his colleagues (2009) found the age-specific PSA reference ranges were as follows: 40–49 years, 2.15 ng/mL; 50–59 years, 3.20 ng/mL; 60–69 years, 4.10 ng/mL; and 70–79 years, 5.37 ng/mL. Similarly, Oesterling and colleagues (1993) recommended reference range for serum PSA for men aged 40 to 194 University of Ghana http://ugspace.ug.edu.gh 49 years as 0.0 to 2.5 ng/mL; for 50 to 59 years, 0.0 to 3.5 ng/mL; 60 to 69 years, 0.0 to 4.5 ng/mL; and 70 to 79 years, 0.0 to 6.5 ng/mL. In these previous studies, the ULs consistently increase with ageing going as high as 6.5 ng/mL. Even though this study did not derive age-specific RI of PSA among Ghanaians, it is possible that the apparently higher values derived in these previous studies are explained by the population of 40+ years used. This makes the population slightly distinct from the combined range of 18 to 70 years used in this study. It is important to emphasize that the Ghanaian finding is congruent with what is reported by Gupta et al. (2014) among Indian males where the reference interval was found to be 0.71 ng/mL in males younger than 40 years and up to 2.35 ng/ml men older than 80 years. 5.3 Determination of Sources of Variation The study examined the sources of variation that are important to the various analytes derived. Thus, it was important that other factors that impact the individual analytes to be explored. Therefore, in consonance with what is reported in the literature, this study specifically explored the effect of age, BMI, SBP, DBP, exercise level, alcohol and ethnicity. 5.3.1 Age and Clinical chemistry Analytes. 5.3.1.1 Relationship between Age and Kidney Function Analytes. Over the years several studies have established the negative relationship between age and eGFR (van der Velde et al., 2010; Pottel et al., 2016; Douville et al., 2008; Denic et al., 2016; Abdulkader et al., 2017). The finding of this study supports existing literature as the Ghanaian population is mirroring the global trend. Thus, it has been observed that GFR 195 University of Ghana http://ugspace.ug.edu.gh declines, perhaps inexorably, with normal ageing, usually beginning after 30–40 years of age, with the decrease being more rapid after 50 years. This drop appears to be a part of the normal physiologic process of cellular and organ senescence and is linked with structural changes in the kidney (Glassock, 2009; Denic et al., 2016). It is also worth mentioning that, chronic kidney disease (CKD) is associated with decreased serum eGFR concentration, yet this is distinguishable from decline caused by normal ageing. Thus, it is observed that the preservation with ageing of a normal urinalysis, normal serum urea and creatinine values, erythropoietin synthesis, and normal phosphorus, calcium and magnesium tubular handling helps in drawing this distinction (Musso & Oreopoulos, 2011). Moreover, this current study found that there is a positive causal effect of age on creatinine and urea in females. This is consistent with what is reported among Indians (Bandebuche et al., 2017). Accordingly, Tiao and colleagues (2002) confirmed this finding in their study where the authors emphasize “Advancing age affects serum creatinine levels; the changes in serum creatinine concentration that occur with age is relevant in the interpretation of the results of renal monitoring after intervention”. In general, it obvious that age is the single most important determinant of kidney function as this study has proved that almost all the various analytes used to test kidney function are affected by age. Particularly, special emphasis is placed on how ageing affects the physiological function of the kidney and the fact many of the analytes are interlinked. It is also important to note that over time this association of age and kidney function has led to many old people being diagnosed as having CKD. Interestingly, this has led to a debate in the literature (Dousdampanis et al., 2012). Some contend that the increased recognition of 196 University of Ghana http://ugspace.ug.edu.gh CKD is a positive development that will improve care in older adult populations (Levey et al., 2011; Eckardt et al., 2009). On the other hand, others argue that this development is causing unnecessary labelling of far too many older populations as diseased without any proven clinical benefit (Moynihan et al., 2013; Delanaye et al., 2016). 5.3.1.2 Relationship between Age and Liver Function Analytes. In the current study, the findings revealed that total protein and albumin are significantly influenced by age; specifically, as age increases total protein and albumin both decrease. This causal relationship is confirmed in other studies (Tian et al., 2014; Weaving et al., 2016). It is worth pointing out that, among the Chinese the rate of decline of total protein with age was faster in females than in males; a revelation that contradicts what is reported among Ghanaians. Thus, in Ghanaian males the rp = -0.45 while in Ghanaian females rp = -0.25. The interesting revelation is that there is inconsistency in the literature regarding the effect of age on albumin. For instance, no age-related changes were observed in albumin concentration (Manolio et al., 1992). Nonetheless, a study found that there was about 10% reduction in albumin concentration over the age range of 18 to 65 years; with values observed to be about 2 g/L in females until those ages associated with the menopause (McPherson et al.,1978). Several studies have also acknowledged the phenomenon of ageing and decreasing liver function (Kim et al., 2016; Cieslak et al., 2016; Frith et al., 2009; Sheedfar et al., 2013; Gan et al., 2011). Similar, to the relationship between age and kidney diseases, it is vital that physicians and public health professionals recognize that ageing is a potential risk 197 University of Ghana http://ugspace.ug.edu.gh factor for liver diseases, particularly non-alcoholic fatty liver diseases (NAFLD). It is also important that for albumin and total protein, particular emphasis be put on the reference intervals to be used for them in clinical applications since they are age-dependent. 5.3.1.3 Relationship between Age and Lipids Analytes. In this study lipid analytes such as TC, TG, and LDL-C were positively influenced by age. The effect of age on these analytes have been reported elsewhere (Marhoum et al., 2013). In previous cross-sectional (Carroll et al., 2005; Kuzuya, et al., 2002; Abbott et al., 1998) and prospective (Anderson et al., 1987; Yamada et al., 1997; Ferrara et al., 1997) studies, total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) gradually increase after adolescence until the age of 60–65 years in men and 70–75 years in women, and begin to decline thereafter. This is confirmed in a recent study (Upmeier et al., 2011). This trend of increasing concentration of lipids with age and the subsequent decline in the elderly have led many to believe there is inconsistency in the literature. Yet, the fundamental observation here is that it is important that authors clearly delineate the age categories so that their results will not be misinterpreted. Consequently, a synthesis of previous reports and the findings from this current study are congruous. Even though, this study did not analyze lipids specifically for the elderly. 5.3.1.4 Relationship between Age and Endocrine Analytes. In both males and females, this study found that age significantly influences the concentration level of PTH by causing it to rise. This is congruent with previous studies (Need et al., 2004; Carrivick et al., 2015; Arabi et al., 2010; Blocki et al., 2012). The increasing concentration of PTH resulting from increasing in age has been attributed to the potential of the decline in creatinine clearance with ageing (Arabi et al., 2010). It is also 198 University of Ghana http://ugspace.ug.edu.gh noteworthy that calcium and serum 25-OHD have been identified to be closely linked with PTH concentration level (Haden et al., 2000; Carrivick et al., 2015; Arabi et al., 2010; Blocki et al., 2012). For example, Patel et al. (2007) indicated that the glomerular filtration rate is an essential determinant of the serum PTH level for low levels of 25-OHD. In addition, Arabi et al. (2010) established PTH correlated with creatinine level but emphasized that the relationship between age and PTH disappeared when creatinine was adjusted. Hence, the authors suggest the effect of age on PTH is mediated by the decline in kidney function (Arabi et al., 2010). The increase in PTH with age in this study suggests that consideration should be given to the establishment of age-related PTH reference ranges. Consequently, this is explored with SDRage in the succeeding pages (page 36). This will generally improve diagnostic accuracy and patient management. One of the interesting findings concerns the group of female-specific hormones that were studied and their relationship with age. These hormones including estradiol, progesterone and prolactin significantly declined with increasing age. This is arguably unsurprising following the fact that the physiological and biological makeup of females are distinct are goes through several changes before, during and after the reproductive stage. Several studies have confirmed this inverse relationship between age and these analytes in women. Specifically, most of these group categorize the ages into pre-menopause and post- menopause. For example, according to Roelfsema and colleagues (2012), postmenopausal women exhibit a 40% decrease in prolactin secretion compared with premenopausal women in the follicular phase of their menstrual cycle. Similarly, Vekemans and Robyn 199 University of Ghana http://ugspace.ug.edu.gh (1976) avow that serum prolactin levels in women reduce progressively with age with a significant decline at the menopause which continues even further during the next decade. Elsewhere this study found that increasing age caused a rise in the concentration level of LH and FSH. This is consistent with previous studies (Lee et al., 1988; Reame et al., 1996). According to Reame et al. (1996), the FSH and LH are interlinked and that “1) the age- related increase in FSH concentrations in ovulatory women, although more pronounced, is associated with phase-dependent enhancement of pulsatile LH secretion, 2) the higher LH concentrations are brought about by changes in both pulse frequency and amplitude, and 3) these age effects preempt overt declines in cyclic estradiol or progesterone concentrations”. At this point, and based on the foregoing discussion above, it is important to mention that amongst other all other variables that were tested in this study, age remains the most important determinant of female sex hormones – estradiol, progesterone, prolactin, LH and FSH. It therefore gives direction as to how reference intervals for these analytes must be interpreted by paying critical attention to age. In this regard, the current study further analyzed these analytes with the SDRage to ascertain the potential of deriving age-specific reference intervals for these analytes. 5.3.1.5 Relationship between Age and Tumour markers. Prostate-specific Antigen (PSA) is a male-specific tumour marker, apparently the concentration level of this analyte and its dependence on age is well documented in literature (Casey et al., 2012; Heidegger et al., 2015; Loeb et al., 2012; Bergers & Oelke, 2011; Vickers et al., 2014; Luboldt et al., 2007). The finding in this study agrees with the global trend that PSA is highly influenced by age. To illustrate this, Casey et al. (2012), 200 University of Ghana http://ugspace.ug.edu.gh revealed that in males, the velocity of PSA is 0.024 ng/mL/year. The authors emphasized that there is a near flat line of PSA concentration values from age 20 to 50 years but this steadily rise after age 50 years. As an important determinant of prostate cancer, there has been a lot of publicity about prostate cancer in the public health circle. This has come with advocacy that males should pay close attention to their prostate health and go for prostate cancer screening after age 45 years. Moreover, the cognizance of this effect of age on PSA has led many authors to establish age-specific reference intervals (Bergers & Oelke, 2011; Na et al., 2013; Lin et al., 2010; Gupta et al., 2014). This is critical to clinical decision making and will eventually lead to improved clinical outcomes. In this study, such partitioning is further explored with the standard deviation ratio (SDRage) which is discussed subsequently elsewhere. In females, ovarian cancer is tested with the CA125 biomarker. In this study, it was revealed that the concentration of CA125 decreases with increasing age. Previous studies are consistent with this finding (Park et al., 2011; Yang et al., 2013; Johnson et al., 2008; Chao et al., 2013). Given that the elevated levels of CA125 are observed in older women, the role of menopause has been explored by other authors. For example, Park et al. (2011) attribute the possibility of the lower concentrations of CA125 recorded among the oldest age group in their study to menopausal status. Despite the claim, what is fundamental here and for the benefit of general public health is that age is a potential risk factor for ovarian cancer. 201 University of Ghana http://ugspace.ug.edu.gh 5.3.2 Body Mass Index and Clinical Chemistry Analytes. In this study, it was found that amylase had an inverse relationship with BMI, with an increasing BMI causing a decrease in the concentration levels of amylase in females. The role of BMI in low serum amylase is emphasized elsewhere in both a univariate and multivariate model. Specifically, obesity has been identified as an important determinant of low serum amylase in the general population (Nakajima 2016). In another study, Nakajima and colleagues found no significant difference in the level of serum amylase between lean and obese subjects with overt non-alcoholic fatty liver diseases (NAFLD). This finding actually buttresses the fact that the impact of BMI on amylase is readily observed in subjects that are healthy rather than subjects that may have other conditions such as NAFLD. The impact of BMI on uric acid has seen mixed results in literature. Nonetheless, this study found that BMI impacts UA, as increasing BMI is associated with an elevated level of serum UA in both males and females. This is supported by the finding in the Tromso study in Norway which demonstrated that higher BMI was associated with an elevated level of serum UA (Norvik et al., 2016). Of course, there is an assertion that obesity and UA may resonate one another by activating the renin-angiotensin system (Goossens et al., 2003; Reaven, 2011; Corry et al., 2008). While the relationship between BMI and UA is confirmed in this by other studies, Yokoi and colleagues (2016) did not observe an interaction between BMI and UA among Japanese males. A finding that contradicts with what is reported among Ghanaian males. 202 University of Ghana http://ugspace.ug.edu.gh 5.3.2.1 Body Mass Index and Liver Function In both males and females BMI was positively associated with ALT even though the impact was more in males (rp= 0.31) than in females (rp = 0.23). This finding agrees with previous studies (Ruhl & Everhart, 2003; Kim & Jo, 2010; Adams et al., 2008; Stranges et al., 2004; Carter et al., 2018; Hsieh et al., 2009). In contrast, however, Das et al. (2015) in a study among Indians found no significant relationship between BMI and ALT when BMI was categorized into three groups of normal, overweight and obese individuals. While this finding may be valid, its external validity may be questioned as the 156 of the sample size used may be too small to detect any significant relationship in such study. Nonetheless, Marchesini et al. (2008) emphasize that higher BMI, particularly the presence of obesity increases the risk of elevated liver enzymes by two to three-fold. Thus, the elevated level of ALT with increasing BMI is a common phenomenon in both males and females globally. This is an important finding for efforts to reduce or prevent liver- related diseases as maintenance of normal weight is protective to liver diseases. 5.3.2.2 Body Mass Index and Lipid Analytes. One of the interesting findings was the consistency in the effect of BMI on the various analytes used for testing lipids. For instance, TC, TG, and CHOL/HDL were all positively impacted by BMI. In essence, increasing BMI led to increased levels of these analytes, indicating the need to consider this source of variation. Several studies have confirmed this association (Shamai et al., 2011; Ichihara et al., 2017; Ali et al., 2011; Chang et al., 2018; Gostynski et al., 2004). Meanwhile, in this study BMI was found to be inversely related with HDL-C; a finding that is consistent with what is reported elsewhere (Shamai et al., 2011; Chang et al., 2018). 203 University of Ghana http://ugspace.ug.edu.gh It is also worth mentioning that Chang et al. (2018) did not find any statistically significant association between BMI and LDL-C. This contradicts what was found in this study partially because in males, a statistically significant association was found but in females, no association was established. It is important to emphasize that the most essential explanatory factor that impacts lipid analytes is the BMI. Hence, in this reference interval derivation study it provides direction as a source of variation that should be further analysed to determine partitioning where necessary. 5.3.2.3 Body Mass Index and Endocrine Analytes (Hormones). Testosterone was found to be inversely related to BMI, with testosterone concentration decreasing with increasing BMI. This finding corroborates several previous studies elsewhere (Shamim et al., 2015; Osuna et al., 2006; DeFina et al., 2015; Diaz-Arjonilla et al., 2009; Eriksson et al., 2017). For example, among Indian males, it was found that a statistically significant decrease in serum testosterone concentration was recorded with increasing BMI. Specifically, those classified as obese presented significantly decreased serum testosterone concentration compared with their normal-weight counterparts (Devi et al., 2018). It is however not always a simple relationship particularly in the presence of sex-hormone binding globulin (SHBG). According to Diaz-Arjonilla et al. (2009), it is acknowledged that BMI is contrariwise proportional to serum total testosterone concentrations; low serum sex-hormone binding globulin (SHBG) levels in obesity contribute to the low serum total testosterone. This is supported by Eriksson et al. (2017) claim that serum SHBG is a well-known determinant of total serum testosterone. In their study, they confirmed that approximately 26% variance in serum testosterone was explained by SHBG. This further explains the complexity of the 204 University of Ghana http://ugspace.ug.edu.gh relationship between BMI, SHBG and serum testosterone. Despite this, it is important that this study in support of previous studies highlight paying attention to obesity as a risk factor for low serum testosterone in men. Thus, health promotion activities and strategies can consider this relationship. 5.3.3 Systolic Blood Pressure and Kidney Function This study further revealed the impact of systolic blood pressure on some kidney function tests. Thus, it was found that among males, increasing systolic blood pressure caused an increase in ANGAP but a decline in TCO2. On the other hand, among females, elevated systolic blood pressure contributes to a decrease in the concentration of K and Cl. These important analytes and their role in the determination of kidney health means that individual should pay attention to maintaining normal systolic blood pressure level. Moreover, for public health officers and health promotion advocates, systolic blood pressure is an important risk factor for kidney diseases as this study has revealed. Therefore, controlling systolic blood pressure in both males and females should be included in kidney health education in Ghana and other regions. 5.3.4 Hours of Standing and Lipids Profile. One of the striking findings was the revelation that hours of standing inversely impact triglyceride concentration. Thus, among Ghanaian males, standing contributed to a decline in TG level. This is confirmed in a study among Australians where the authors measured the effect of standing on the lipids profile by replacing sitting with standing. It was revealed that replacing 2 hours of sitting with standing led to 11% lower TG concentration (Healy et al., 2015). This finding comes at the most critical point when the prevalence of cardiovascular diseases is on the ascendancy. Individuals who are seeking to improve their 205 University of Ghana http://ugspace.ug.edu.gh cardio-metabolic health should consider standing as an important strategy. Intuitively, this may be an important aspect of physical activity. 5.4 Standard Deviation Ratios (Sex, Age, Ethnicity). Proper interpretation of results requires an understanding of the sources of variation which influence laboratory tests. Determination of these sources of variation was an important part of this study which provided guidance for sex, age and ethnic partitioning. 5.4.1 Standard Deviation Ratios and Haematology. This study found that SDRsex for RBC, Hb and Ht exceeded the 0.4 threshold for partitioning. This finding corroborates previous studies in Kenya (Omuse et al., 2018) and Japan (Yamakado et al., 2015). Furthermore, while this study found Plt required sex partition (SDRsex =0.45) it contradicts what is reported in Japan (SDRsex = 0.14) (Yamakado et al., 2015). Interestingly, mono% was found to require sex-specific RIs for males and females according to what Omuse et al. (2018) found among Kenyans; a finding that is in sharp contrast with our finding. The sex variation that was observed in RBC, Hb, and Ht could be as a result of the effect of androgens on erythropoiesis. Androgen stimulates haematopoietic system by various mechanisms, including stimulation of erythropoietin, increasing bone marrow activities and iron incorporation into the RBC (Shahani et al., 2009). Also, the significant sex difference of platelet count in females (157- 402 x109/L) showing higher values than males (115 – 339 x109/L) may be due to the effect of endogenous female sex hormones, mainly the estrogen on the activation of platelet formation. Megakaryocytes use 3-hydroxysteroid dehydrogenase to synthesize 17b- estradiol (E2), which regulates platelet formation (Leng & Bray, 2005; Moro et al., 2005). 206 University of Ghana http://ugspace.ug.edu.gh 5.4.2 Standard Deviation Ratios and Chemistry Analytes. In chemistry analytes, the study revealed that UA, CK, and C3 required partitioning of their RIs according to sex. This is confirmed in other multicenter studies in China, Japan, Saudi Arabia and other six Asian countries which derived RI for their respective populations (Xia et al., 2016; Yamakado et al., 2015; Ichihara et al., 2013; Borai et al., 2016). It is worth emphasizing that these analytes and their between-sex differences have been confirmed in this study, where males presented significantly higher RI values than females. In addition, this found no need for partitioning LDH according to sex because its SDRsex = 0.00 which does not meet the SDRsex > 0.4 criteria. Again, this is consistent with what is reported among Saudi Arabians and Turkish (Borai et al., 2016; Ozarda et al., 2013). Calcium presented both sex and age-specific direction among Ghanaians with the SDRageF =0.45; an indicator that partitioning of RIs according to age is required among females. This finding reflects the established trend of decreasing calcium concentration level in older age. However, for the purpose of this study, it is important to mention that deriving RIs and partitioning them according to age must only be done among Ghanaian females. This means that even though separate RIs calcium should be used for males and females, the females RIs should further be distributed with respect to age. 5.4.2.1 Standard Deviation Ratios and Kidney Function Analytes. Regarding SDRs for kidney function tests, it was found that creatinine and chloride required sex-specific RIs. Specific to creatinine, previous studies among Chinese, Saudi Arabians, and Japanese support this finding that males have a significantly higher concentration of creatinine than females (Xia et al., 2016; Yamakado et al., 2015; Borai et al., 2016). Hence, the use of a combined reference interval of creatinine for both males and females will be misleading. For instance, the RIs of creatinine that are presented in 207 University of Ghana http://ugspace.ug.edu.gh previous studies considerably differ between males and females (Abebe et al. 2018; Yamakado et al., 2015; Ozarda et al., 2014). Therefore, using males’ RIs for females and vice versa will be clinically disastrous. Also, among Saudi Arabians Borai et al. (2016) confirmed that males and females require separate RIs to guide diagnosis and patient management. In contrast however, this finding is not parallel to what is found among Ghanaians. The variation in these two studies may be accounted for by simply population differences. Furthermore, among Ghanaians age-specific differences were observed for eGFR and Na. Here, while there is a need to partition RIs with respect to age among both males (SDRageM = 0.48) and females (SDRageF = 0.70) for eGFR, such partitioning is necessary among only females (SDRageF = 0.81) for Na. The finding for eGFR is confirmed by Yamakado et al. (2015) whereas the finding for Na contrasts with what is reported by Borai et al. (2016). The need for partitioning eGFR with respect to age is confirmed in the SDR (Figure 4.2) which shows a steady decline of eGFR with increasing age. Similarly, among females the Na declines with age, therefore the need to have narrower RI for advancing is appropriate for improved clinical decision making. Ethnicity-related variation was observed for ANGAP, and pointed to the need to partition the RIs for females between Akans and non-Akans. This between-ethnicity differences in ANGAP is not readily explicable, however, the possibility of this being attributable to nutrition have been explained elsewhere. According to Adeva and Souto (2011), human diet with excessive animal products and cereal grains consist of sulfur-containing amino acids, whose oxidation generates sulfate that is a major determinant of acidosis, whiles fruits and vegetable-based foods contain potassium salts of metabolized anions, which 208 University of Ghana http://ugspace.ug.edu.gh consumes hydrogen ions when metabolized causing an alkalinizing effect. Consequently, it is believed that the differences in the diet consumed by Akans and non-Akans account for these differences in ANGAP. 5.4.2.2 Standard Deviation Ratios andLiver Function Analytes. Several analytes under the kidney function test such as albumin, Tbil, Dbil, AST, ALT, and GGT all required partitioning by sex. The conspicuous differences between RIs for males and females are shown in the Figure 4.3. Other studies have confirmed the need to partition for sex regarding albumin, Tbil, ALT, and GGT (Xia et al., 2016; Yamakado et al., 2015; Ichihara et al., 2013; Borai et al., 2016; Ozarda et al., 2013). The required partitioning of these analytes is confirmed in the various RIs that are derived among Ghanaian males and females which show significant variations between sex. Generally, the sex-related differences may largely be explained by the biological, physiological and hormonal make up of males and females. It is also interesting to mention that albumin showed age-related differences in this study; a finding that is also consistent with previous studies elsewhere (Ozarda et al., 2013; Ichihara et al., 2012; Xia et al., 2016). What is more striking here is that among Ghanaians the age partitioning is required in only males (SDRageM = 0.48). Interestingly, this is consistent with what is reported in Chinese, and Turkish (Ozarda et al., 2013; Xia et al., 2016), but in a multicenter study among Asian countries the partitioning was required in both males and females (Ichihara et al., 2012). Moreover, the age-related changes that are observed in Ghanaian males showed that as age increases albumin level declines. The MRA explains why the partitioning is required in males only because in the MRA the effect is observed only in males (rp =-45). 209 University of Ghana http://ugspace.ug.edu.gh 5.4.2.3 Standard Deviation Ratios and Lipids Profile. Amongst the analytes for lipids profile, age-specific partitioning was required for only TG. Thus, as evidenced in the MRA, increasing age corresponded with increasing concentration of TG in females. This finding is consistent with other studies that found the need to partition for females, age-specific RIs (Ozarda et al., 2013; Borai et al., 2016). Contrary to other studies, partitioning was restricted to sex only without consideration for age (Ichihara et al., 2012; Yamakado et al., 2015). Moreover, HDL-C was found not to require partitioning for sex, age, and ethnicity among Ghanaians. Contrastingly, among all Asian populations studied, there is a need to partition HDL-C for sex (Xia et al., 2016; Yamakado et al., 2015; Ichihara et al., 2012). Thus, the RIs of males and females in these populations are not identical. The interesting thing about the lipids profile tests is that at one point the Ghanaian results are identical with those among Asian, Saudi Arabians and Turkish and at another point, they contradict each other. Notwithstanding this, it is evident that it is always necessary to compute in multiple ways and interpret the results appropriately with reference to age and sex. 5.4.2.4 Standard Deviation Ratios and Iron Studies. The ferritin and iron analytes were identified in this study to require partitioning for age and sex. In the case of ferritin, both sex-specific RI needed to be established yet, for females, further partitioning was required with respect to age. This corroborates a previous study conducted among the Turkish population (Ozarda et al., 2013). There are two main parts to this finding. First, the need for sex partitioning is the significantly higher concentration of ferritin in males than females. Secondly, the females have an age-related variation that is associated with reproductive age particularly lower in this age group. 210 University of Ghana http://ugspace.ug.edu.gh According to Casale and colleagues (1981), this could be explained as a result of the activation of the reticuloendothelial system and of the surge of iron storage with ageing. More so, the need for iron to be partitioned according to sex is confirmed with previous studies in China, Turkey, and other Asian countries (Xia et al., 2016; Ozarda et al., 2013; Ichihara et al., 2012). In the RIs derived in this study for Ghanaian adults, the reference values for males are significantly higher than females, confirming the need to use separate and sex-specific RIs for males and females. Like the case of ferritin, blood loss due to birth and menstruation have been the main factors that have been attributed to the lower reference values for females. 5.4.2.5 Standard Deviation Ratios and Tumour Markers. Prostate-specific antigen (PSA) is a male-specific tumour marker that is associated with age. There is widespread evidence that this analyte is positively related to age (Casey et al., 2012; Heidegger et al., 2015; Loeb et al., 2012; Bergers & Oelke, 2011; Vickers et al., 2014). Therefore, as males advance in age, their PSA concentration also increases. The SDRageM = 0.47 proves that there is a need to have specific age RIs that will guide clinical decision making for testing PSA. Elsewhere, other authors have established age-specific RIs for PSA (Bergers & Oelke, 2011; Na et al., 2013; Lin et al., 2010; Gupta et al., 2014). 5.4.2.6 Standard Deviation Ratios and Endocrinoly Consistent with the MRA, the SDR for all the female-specific hormones in the endocrine studies are heavily dependent on age. Apparently, the SDRageF for all of these analytes – estradiol, progesterone, LH, FSH, and PRL are > 0.4 which means they need to be partitioned according to age. This is actually confirmed in a previous study for estradiol in 211 University of Ghana http://ugspace.ug.edu.gh a multicenter study among Asians (Ichihara et al., 2012). Yet, in that same study, the SDRageF < 0.4 for progesterone did not meet the criteria to warrant partitioning for age (Ichihara et al., 2012). This contrasting finding with respect to progesterone is particularly interesting because it affirms the mixed results in literature as to the relationship between age and progesterone concentration. The LH, FSH, and PRL are all found to require age- specific RIs that will improve diagnosis and patient management. This is particularly important for the premenopausal and postmenopausal stages as their linkage with reproduction is equally vital. Furthermore, this study found that the cortisol, FT3, and PTH all required sex-specific RIs; with PTH requiring further partitioning for age in males and females. The need for partitioning RIs by sex for cortisol contrasts with that which is reported in the Asian multicenter study (Ichihara et al., 2012). Thus, in that study the authors found that the cortisol concentration was higher in males than in females, their SDRsex analysis was not big enough to support partitioning. Meanwhile, in this study, among Ghanaians cortisol level was considerably higher in males than in females particularly the UL and the need for partitioning RIs for sex is confirmed with the SDRsex = 0.60. The PTH which shows that age-related partitioning is required for both males and females reflects the impact age has on this analyte. This is confirmed with the MRA where age causes a significant increase in the concentration level of PTH in both males and females. According to Arabi et al. (2010), one of the main reasons why PTH concentration increases with age is because of the decline in kidney function, against the backdrop that creatinine level correlates with PTH. 212 University of Ghana http://ugspace.ug.edu.gh 5.4.2.7 Standard Deviation Ratios (Sex, Age, Ethnicity) on Immunoglobulins. The analytes IgG, IgA, and IgM produced SDR results that are consistent with the findings in the Asian multicenter study. In a large collaborative study involving 10 countries in Asian that aimed to establish standardized reference intervals, the authors found neither sex nor age-related variations in IgG and IgA (Ichihara et al., 2012). However, for IgM the authors reported an SDRsex = 0.71 which is a pointer to the need to have male and female- specific RIs for IgM. These results are corroborated with the findings in the current study which found no sex and age-related differences in IgG and IgA but for IgM. This consonance in results albeit different populations may be as a result of the global trend in immunoglobulins distribution. 5.4.2.8 Standard Deviation Ratios (Sex, Age, Ethnicity) on Vitamins Studies. This study examined the role of ethnicity as a source of variation on analytes that were derived. For vitamins, it was observed that between Akans and non-Akans, partitioning was required and this was necessary only in females. The reason for this variation between Akans and non-Akans may be attributed to nutrition. Traditionally, the kinds of food that are consumed by non-Akans differ to some extent from that of Akans and that may be a possible source of folate variations between these two groups. 5.5 Latent Abnormal Values Exclusion Method on both parametric and non- parametric approaches In deriving reference intervals, it is important to adopt strict criteria to eliminate all subjects with abnormal results due to latent diseases. The latent abnormal values exclusion method was applied to aid in excluding all the apparently healthy individuals with latent diseases such as diabetes mellitus, metabolic syndrome, inflammatory diseases and 213 University of Ghana http://ugspace.ug.edu.gh anaemia which is common in Ghana. The results revealed a conspicuous difference between the RIs derived with and without the application of LAVE in both parametric and non-parametric methods. By the application of the LAVE method, the RIs of most analytes had lower and narrowed reference intervals compared with the ones without LAVE. The haematological analytes, mostly the erythrocytes (RBC, Hb, Ht, MCV, MCH, MCHC & RDW) depicts an evident variation of application of LAVE (+) showing narrow lower limits in both male and female as shown in Figure 4.10. This finding corroborates research conducted in Kenya which adopted the use of LAVE method in deriving haematological analytes among healthy Kenyans (Omuse et al., 2018). The narrower intervals observed in the analytes as a result of the effect of LAVE application may be due to the elimination of individuals with potential latent diseases such as anaemia and inflammations whose results could have influenced the reference intervals derived. Interestingly, the analytes used as reference tests in the application of LAVE, which had direct association with metabolic syndrome, inflammation and anaemia such as aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (GGT), creatinine (CRE), total cholesterol (TC), triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C), low density lipoprotein (LDL-C), Haemoglobin, Haematocrit, ferritin, transferrin, iron, etc. RI results were obviously narrowed. These findings confirm with other studies conducted using the LAVE method as a means of eliminating latent diseases (Borai et al., 2016; Ozarda et al., 2014; Ichihara and Boyd, 2010; Omuse, 2018). This obvious change among these reference analytes may be due to the exclusion of individuals with potential latent diseases such as diabetes, inflammation 214 University of Ghana http://ugspace.ug.edu.gh and iron deficiency syndrome whose results could have influenced the reference intervals derived. Although, the application of LAVE was prominent in some analytes but it did not show any change in the RIs of other analytes, mainly the electrolytes (Na, P, Cl). This may be because these analytes are mostly not associated with metabolic syndrome and anaemia which are latent diseases. This affirms the importance of the LAVE as a secondary exclusion of latent diseases among apparently healthy population used in RI derivation. Also, the study derived RIs using both the parametric method and non-parametric method. It is noteworthy that RIs derived by the non-parametric method was only used for comparison. The main advantage of the parametric method over the non-parametric is that it enables identification of outlying points from the profile of the central portion of reference distribution. Both methods used in the RI derivation depicted variations in the RIs of the studied analytes. The findings of the study confirmed with other studies, that RIs derived by parametric method have narrow RIs compared to that of the non-parametric method, especially for those analytes which are associated with the metabolic syndrome, anaemia and inflammation (Hb, Ht, Folate, Ferritin, CRP, LDL-C, TG, TC, CK, Glu, TG, AST, ALT, and GGT) as shown in Figures 4.10-4.13. This variation observed between the two methods with non-parametric exhibiting wider RIs among most analytes may be the effect of individuals with latent abnormal diseases influencing the results (Borai et al., 2016; Ozarda et al., 2014; Ichihara and Boyd, 2010; Omuse, 2018). Moreover, it is worth noting that some of the analytes derived by the two methods did not show any variations among the RIs of some analytes (TP, Alb, UN, Na, K, Ca, and AMY) 215 University of Ghana http://ugspace.ug.edu.gh for which may be the prevalence of abnormal results in the healthy population for such analytes were very low. The only difference between the two methods in this study is that the parametric method gives narrower confidence intervals for the lower and upper limits of the RI and this affirms other studies which used the parametric method in the derivation of RI (Ichihara & Boyd, 2010). 216 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE CONCLUSIONS AND RECOMMENDATIONS 6.0 Introduction This concluding chapter summarizes the thesis and draws conclusions based on the findings. Recommendations, contributions of the study and suggestions for further studies are also presented. 6.1 Conclusions The findings of this research highlight the sources of variations such as age, sex, body mass index, blood pressure, and ethnicity on the reference intervals of the haematological and clinical chemistry analytes derived among Ghanaian population. It also underscores the variation of the reference intervals of the haematology and clinical chemistry analytes derived among Ghanaian population and the manufacturer analyzers’ reference interval currently in use in the country. The use of robust statistical technique and the application of latent abnormal value exclusion (LAVE) method in the derivation of the reference intervals aided in excluding all the reference individuals who had latent diseases such as iron-deficiency anaemia, metabolic syndrome, diabetes mellitus and inflammations. The exclusion of such subjects led to a refined and appropriate UL and LL of RIs derived. The haematological analytes derived among the Ghanaian adult population depicted both sex and age variations. Most of the erythrocytes (RBC, Hb, Ht, MCV) were higher in 217 University of Ghana http://ugspace.ug.edu.gh males than in females whiles Platelet count was the only analyte that showed significantly higher values in females compared to males. None of the leukocytes showed sex differences except eosinophil. Also, with age variation with regards to haematological analytes, only red blood cells in males depict decreased age variation. This signifies that haematological reference intervals derived among the Ghanaian population require sex- specific RIs for erythrocytes while with the leukocyte analytes the combined male and female RIs could be used. Furthermore, the study found significant variations between the haematology RIs derived in this study for Ghanaians and the haematology RIs from the manufacturer analyzer (Sysmex XN 1000) currently in use by some diagnostic laboratories in Ghana. This is an indication for the need to effectively drop these foreign-derived RIs in use and adopt locally derived RIs. More so, majority of the clinical chemistry analytes showed sex, age, BMI and ethnicity variations. Sex-related variation existed in the kidney and liver analytes, but more prominent in the endocrine analytes due to the hormonal disparities in the physiology of males and females. This is an indication that interpretation of laboratory results should factor sex variation, because most of the clinical chemistry analytes reference intervals derived among Ghanaian adults shows differences in males and females. Age-related variations was observed in some of the clinical chemistry analytes. Most of the RIs of the kidney (eGFR), liver (Tp, Alb), tumour marker (CA125) and endocrine analytes were age- dependent; thus, RIs of these analytes decreases with increasing age. In addition, age- related variations were observed in ferritin among females; ferritin concentration increase with increasing age. This signifies that interpretation of laboratory results should factor age because RIs of some analytes are interpreted more accurately when they are age- specific. 218 University of Ghana http://ugspace.ug.edu.gh Furthermore, the study found ethnicity variations among Akan and the other ethnic groups. Folate, anion gap and phosphate were higher in Akans than the non-Akan ethnic group. This indicates that within-country variations exist between different ethnicity, which could be attributable to diet; hence the need for further research on RIs among ethnic groups in Ghana. Regarding the effect of BMI on the clinical chemistry analytes, most of the analytes had a positive association with BMI. One interesting finding was the consistency of the effect of BMI on the various lipid analytes; thus, increasing BMI led to increased levels of these lipid analytes, indicating the need to consider this source of variation, whether its magnitude warrants the need for partitioning. It should be noted also that, the consistently strong association of BMI on these analytes indicates that, BMI is a potential risk factor for several lipid-related diseases. Moreover, there were significant variations observed between reference intervals derived among Ghanaian adults and the current manufacturers standard in use. This indicates that the continuous use of such foreign-derived RIs can lead to misdiagnosis and affect clinical decision making. Interestingly, within-country differences was observed among other studies conducted in Ghana. This indicates that ethnicity and environment as confirmed by literature could be some factors of such variation studies. More importantly, it is observed that the application of the LAVE significantly improved the RIs derived which contributed to the considerable variations between this current study and other previous studies conducted in Ghana. The study affirmed the advantage of the parametric method over the non-parametric and the need for the application of LAVE method for analytes, which are easily influenced by the presence of prevalent latent anaemia, metabolic syndrome and inflammation. Sex, age 219 University of Ghana http://ugspace.ug.edu.gh and BMI differences were mainly the sole determinant of variations among the studied analytes. This indicates that sex and age-specific RIs are required for the effective interpretation of some haematological and clinical chemistry analytes. The robust statistical technique used in this study makes the RI derived more representative of the healthy Ghanaian population, which will lead to more reliable clinical decision making and patient care. The findings from the present study therefore indicate that adopting these haematological and biochemical RIs for clinical use will be beneficial to healthcare systems in Ghana and sub-Saharan Africa. 6.2 Recommendations In reference to the findings of this study, the following recommendations are made to aid in improving clinical decision making with reference to the interpretation of clinical laboratory results. 6.2.1 Ministry of Health, Ghana Health Service, Ghana Standards Authority (Medium-term) Foremost, this Ghanaian population-based derived reference intervals should be adopted for use in the country. Using locally derived reference intervals will improve clinical and policy decisions that can help save scarce economic resources in the healthcare sector. In essence, the government, healthcare providers and households will benefit from improved accuracy of clinical diagnosis and disease management. The Ministry of Health (MOH), Ghana Health Service (GHS), Ghana Standards Authority (GSA), Ghana Medical and Dental Council, Private laboratory diagnostic services, and other stakeholders can come into consensus to adopt these newly derived RIs for nationwide use. This will be in the bid to improve healthcare clinical decision making and save scarce resources including lives. 220 University of Ghana http://ugspace.ug.edu.gh The adoption of this new findings, means, there should be increased public education and awareness campaigns for clinicians, healthcare providers and laboratory technicians on enhancing the interpretation of diagnostic test results as well as the need to use Ghanaian derived reference intervals as standards in interpreting laboratory results. This can be achieved through campaigns on radio, televisions and during conferences organized by the MoH, GSA, and other relevant stakeholders. 6.2.2 Researchers and Clinical Research Laboratories (Short-term) Secondly, this study proved an invaluable contribution that LAVE brings to RI derivation. Against this backdrop, this study recommends that establishment of any population-based reference intervals that will be subsequently conducted in Ghana or elsewhere should adopt the parametric statistical technique with application of LAVE method. This is because the LAVE method excludes all individuals with latent diseases and imflammation, whose values could influence the RI values unduly and it helped to increase the precision of calculating the RI. It was found that Ghanaian RIs for lipid analytes were considerably higher compared with the current RI in use, hence this study recommends that further nationwide research should be done to investigate RI of lipase in Ghanaian adult to confirm the cause of the higher concentration of lipids among Ghanaian adult. The findings from further studies will help ascertain whether the high concentration of lipids is genetic or lifestyle dependent. 6.2.3 Healthcare Providers and Medical Practitioners (Short-term) Furthermore, proper interpretation of results requires an understanding of the sources of variation which influence laboratory tests. Hence clinicians and healthcare providers 221 University of Ghana http://ugspace.ug.edu.gh should consider sources of variations specifically sex, age and BMI when interpreting laboratory results. In this study, sex and age-specific RIs are provided and clinicians will be better positioned when they use these sex and age-specific RIs for the respective analytes. This means sex and age-specific RIs of some analytes must be available for clinicians’ referral. 6.2.4 General Population/Ministry of Health (Long-term) Ghanaian women generally have low levels of haemoglobin and ferritin compared with the WHO standards; hence the need for nutritional education intervention programmes for the general public. Public education on the need to consume folate-rich food and its contribution to healthy living. Also, folic acid supplementation should be carried out nationwide for all women, since this study observed the low Hb and folate level in women. In addition, government through MoH, GHS, Food and Drugs Authority (FDA) should ensure that most of our food products such as cooking oils, dairy products, etc should be Vitamin D fortified. More so, there should be public education on the effect of vitamin D deficiency on health as well as the importance of the morning sunlight as the main source of vitamin D; and encourage the public to make efficient use of it. 6.3 Limitation of Study One of the limitations of this study was that volunteers’ recruitment comprised largely of the urban population, majority of the volunteers were from the capital city of Greater Accra and Northern region, whose socioeconomic status and lifestyles are different from the rural population. Hence some of these RIs for certain analytes might not be applicable to the rural setting. Another limitation of the study was that we relied on self-reporting of chronic illnesses including those of an infectious aetiology and metabolic syndrome. It is likely 222 University of Ghana http://ugspace.ug.edu.gh that some volunteers may not have provided accurate information and as such may have ended up being included inadvertently. However, the application of LAVE as a secondary exclusion criterion excluded most volunteers whose test results suggested the possibility of sub-clinical disease. 6.4 Future Research Directions Similar research should be carried out in all other regions in Ghana, particularly, future research should include rural communities. Moreover, the extension of the study to cover all regions will provide vital information on ethnicity and possible explanations to the variations of RIs of some analytes among the Ghanaian population. Also, these population- based reference interval projects to be conducted in other regions in Ghana will provide information to drive policy intervention. 6.5 Contribution to Knowledge This study fills in the knowledge gap in haematology, clinical chemistry, hormones, thyroid, immunoglobin and tumour markers reference intervals for Ghanaian population. This is the first study, to the best of my knowledge, to derive a comprehensive reference interval for several clinical chemistry analytes, vitamin, iron, hormones, thyroid, immunoglobulin and tumour markers for Ghanaian adults. Again, this is the first study in Ghana to use the parametric method with the application of LAVE as a statistical tool to derive reference intervals for haematology, clinical chemistry, hormones, tumour markers and immunoglobin analytes for Ghanaian adults. The application of robust statistical methods including the use of LAVE for secondary exclusion helped to reduce apparent healthy individuals with latent anaemia, metabolic 223 University of Ghana http://ugspace.ug.edu.gh syndrome and inflammation whose results might have influenced the derived RIs unduly. This is a major strength of this study which enhances the external validity of the results. The study has also discovered high lipids and low vitamin D concentration in Ghanaian adults, which is a key public health issue, which needs intervention.. The study has also reinforced to existing knowledge on the sex, age, BMI and racial variations of reference intervals in literature. 224 University of Ghana http://ugspace.ug.edu.gh REFERENCES Abbott, R., Yano, K., Hakim, A., Burchfiel, C., Sharp, D., Rodriguez, B., & Curb, J. D. (1998). Changes in total and high-density lipoprotein cholesterol over 10-and 20- year periods (the Honolulu Heart Program). The American journal of cardiology, 82(2), 172-178. Abdulkader, R. C., Burdmann, E. A., Lebrao, M. L., Duarte, Y. A., & Zanetta, D. M. (2017). Aging and decreased glomerular filtration rate: An elderly population- based study. PloS one, 12(12), e0189935. Abebe, M., Melku, M., Enawgaw, B., Birhan, W., Deressa, T., Terefe, B., & Baynes, H. W. (2018). Reference intervals of routine clinical chemistry parameters among apparently healthy young adults in Amhara National Regional State, Ethiopia. PLOS ONE, 13(8), e0201782. https://doi.org/10.1371/journal.pone.0201782 Addai-Mensah, O., Gyamfi, D., Duneeh, R. V., Danquah, K. O., Annani-Akollor, M. E., Boateng, L., ... & Ofosu, D. N. (2019). Determination of haematological reference ranges in healthy adults in three regions in Ghana. BioMed research international, 2019. Adetifa, I. M. O., Hill, P. C., Jeffries, D. J., Jackson-Sillah, D., Ibanga, H. B., Bah, G., … Adegbola, R. A. (2009). Haematological values from a Gambian cohort - possible reference range for a West African population. International Journal of Laboratory Hematology, 31(6), 615–622. https://doi.org/10.1111/j.1751- 553X.2008.01087.x Adeva, M. M., & Souto, G. (2011). Diet-induced metabolic acidosis. Clinical nutrition, 30(4), 416-421. Al-Mawali, A., Pinto, A. D., Al Busaidi, R., & Al-Zakwani, I. (2013). Lymphocyte subsets: Reference ranges in an age- and gender-balanced population of Omani healthy adults. Cytometry Part A, 83A(8), 739–744. https://doi.org/10.1002/cyto.a.22322 Ali, Z. A. U. A., & Al-Zaidi, M. S. (2011). The association between body mass index, lipid profile and serum estradiol levels in a sample of iraqi diabetic premenopausal women. Oman medical journal, 26(4), 263. Allin, K. H., & Nordestgaard, B. G. (2011). Elevated C-reactive protein in the diagnosis, prognosis, and cause of cancer. Critical Reviews in Clinical Laboratory Sciences, 48(4), 155–170. https://doi.org/10.3109/10408363.2011.599831 225 University of Ghana http://ugspace.ug.edu.gh Ambayya, A., Su, A. T., Osman, N. H., Nik-Samsudin, N. R., Khalid, K., Chang, K. M., ... & Yegappan, S. (2014). Haematological reference intervals in a multiethnic population. PloS one, 9(3), e91968. Ameijeiras, A. H., Paz, J. E. L., Alende, M. J., Gonzalez, G. C., Durán, V. M., Montes, A. P., … Gomez, C. C. (2016). [PP.32.12] GENDER DIFFERENCES IN URIC ACID AND CARDIOVASCULAR RISK. Journal of Hypertension, 34, e321. https://doi.org/10.1097/01.hjh.0000492282.81683.53 American Diabetes Association. (2010). Diagnosis and classification of diabetes mellitus. Diabetes care, 33(Supplement 1), S62-S69. Anifandis, G., Koutselini, E., Louridas, K., Liakopoulos, V., Leivaditis, K., Mantzavinos, T., … Vamvakopoulos, N. (2005). Estradiol and leptin as conditional prognostic IVF markers. Reproduction, 129(4), 531–534. https://doi.org/10.1530/rep.1.00567 Arabi, A., Baddoura, R., El-Rassi, R., & Fuleihan, G. E. H. (2010). Age but not gender modulates the relationship between PTH and vitamin D. Bone, 47(2), 408-412. Araujo, A. B., O’Donnell, A. B., Brambilla, D. J., Simpson, W. B., Longcope, C., Matsumoto, A. M., & McKinlay, J. B. (2004). Prevalence and Incidence of Androgen Deficiency in Middle-Aged and Older Men: Estimates from the Massachusetts Male Ageing Study. The Journal of Clinical Endocrinology & Metabolism, 89(12), 5920–5926. https://doi.org/10.1210/jc.2003-031719 Asare, G. A., Ezekiel, N. N., Nkumpoi, T., & Amoah, A. G. B. (2012). Reference intervals for common biochemical analytes in serum and plasma of a random adult population in the Greater Accra Region of Ghana. Clinical Laboratory, 58(7–8), 687–93. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2299 7969 Atzmon, G., Barzilai, N., Hollowell, J. G., Surks, M. I., & Gabriely, I. (2009). Extreme Longevity Is Associated with Increased Serum Thyrotropin. The Journal of Clinical Endocrinology & Metabolism, 94(4), 1251–1254. https://doi.org/10.1210/jc.2008-2325 Bais, R., & Philcox, M. (1994). IFCC methods for the measurement of catalytic concentration of enzymes. Part 8. IFCC method for lactate dehydrogenase (L- lactate: NAD. Journal of Analytical Methods in Chemistry, 16(5), 167-182. Bakan, E., Polat, H., Ozarda, Y., Ozturk, N., Baygutalp, N. K., Umudum, F. Z., & Bakan, N. (2016). A reference interval study for common biochemical analytes in Eastern Turkey: a comparison of a reference population with laboratory data 226 University of Ghana http://ugspace.ug.edu.gh mining. Biochemia Medica, 210–223. https://doi.org/10.11613/BM.2016.023 Bakrim, S., Motiaa, Y., Benajiba, M., Ouarour, A., & Masrar, A. (2018). Establishment of the hematology reference intervals in a healthy population of adults in the Northwest of Morocco (Tangier-Tetouan region). The Pan African Medical Journal, 29, 169. https://doi.org/10.11604/pamj.2018.29.169.13042 Bansal, V. K. (1990). Serum Inorganic Phosphorus. Clinical Methods: The History, Physical, and Laboratory Examinations. Butterworths. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/21250152 Basit, S. (2013). Vitamin D in health and disease: a literature review. British Journal of Biomedical Science, 70(4), 161–72. Retrieved from http://www.ncbi.nlm.nih. gov/ pubmed/ 24400428 Basu, A., Seth, S., Chauhan, A. K., Bansal, N., Arora, K., & Mahaur, A. (2016). Comparative study of tumour markers in patients with colourectal carcinoma before and after chemotherapy. Annals of Translational Medicine, 4(4), 71. https://doi.org/10.3978/j.issn.2305-5839.2016.02.02 Basu, G., & Mohapatra, A. (2012). Interactions between thyroid disorders and kidney disease. Indian Journal of Endocrinology and Metabolism, 16(2), 204–13. https://doi.org/10.4103/2230-8210.93737 Beckman Coulter. (2012). Procedure: ASPARTATE AMINOTRANSFERASE (AST) OSR6109, OSR6209, and OSR6509. Behbehani, A. I., Mathew, A., Farghaly, M., & Van Dalen, A. (2002). Reference levels of the tumor markers carcinoembryonic antigen, the carbohydrate antigens 19-9 and 72-4, and cytokeratin fragment 19 using the® Elecsys Relecsys 1010 analyzer in a normal population in Kuwait. The importance of the determination of local reference levels. The International journal of biological markers, 17(1), 67-70. Beier, K., Eppanapally, S., Bazick, H. S., Chang, D., Mahadevappa, K., Gibbons, F. K., & Christopher, K. B. (2011). Elevation of blood urea nitrogen is predictive of long- term mortality in critically ill patients independent of "normal" creatinine. Critical Care Medicine, 39(2), 305–13. https://doi.org/10.1097/CCM.0b013e3181ffe22a Berges, R., & Oelke, M. (2011). Age-stratified normal values for prostate volume, PSA, maximum urinary flow rate, IPSS, and other LUTS/BPH indicators in the German male community-dwelling population aged 50 years or older. World journal of urology, 29(2), 171-178. 227 University of Ghana http://ugspace.ug.edu.gh Biondi, B. (2012). Natural history, diagnosis and management of subclinical thyroid dysfunction. Best Practice & Research Clinical Endocrinology & Metabolism, 26(4), 431–446. https://doi.org/10.1016/j.beem.2011.12.004 Biondi, B. (2013). The Normal TSH Reference Range: What Has Changed in the Last Decade? The Journal of Clinical Endocrinology & Metabolism, 98(9), 3584– 3587. https://doi.org/10.1210/jc.2013-2760 Bjerner, J., Høgetveit, A., Wold Akselberg, K., Vangsnes, K., Paus, E., Bjøro, T., ... & Nustad, K. (2008). Reference intervals for carcinoembryonic antigen (CEA), CA125, MUC1, Alfa‐foeto‐protein (AFP), neuron‐specific enolase (NSE) and CA19. 9 from the NORIP study. Scandinavian journal of clinical and laboratory investigation, 68(8), 703-713. Bonini, P., Plebani, M., Ceriotti, F., & Rubboli, F. (2002). Errors in Laboratory Medicine. Retrieved from http://clinchem.aaccjnls.org/content/clinchem/48/5/ 691.full.pdf Borai, A., Ichihara, K., Al Masaud, A., Tamimi, W., Bahijri, S., Armbuster, D., … Committee on Reference Intervals and Decision Limits, International Federation for Clinical Chemistry and Laboratory Medicine. (2016). Establishment of reference intervals of clinical chemistry analytes for the adult population in Saudi Arabia: a study conducted as a part of the IFCC global study on reference values. Clinical Chemistry and Laboratory Medicine (CCLM), 54(5), 843–55. https://doi.org/10.1515/cclm-2015-0490 Borghi, C., Rosei, E. A., Bardin, T., Dawson, J., Dominiczak, A., Kielstein, J. T., … Mancia, G. (2015). Serum uric acid and the risk of cardiovascular and renal disease. Journal of Hypertension, 33(9), 1729–1741. https://doi.org/10.1097 /HJH.0000000000000701 Boyd, J. C. (2008). Cautions in the Adoption of Common Reference Intervals. Clinical Chemistry, 54(2), 238–239. https://doi.org/10.1373/clinchem.2007.098228 Boyd, J. C. (2010a). Defining laboratory reference values and decision limits: populations, intervals, and interpretations. Asian Journal of Andrology, 12(1), 83–90. https://doi.org/10.1038/aja.2009.9 Boyd, J. C. (2010b). Defining laboratory reference values and decision limits: Populations, intervals, and interpretations. Asian Journal of Andrology, 12(1), 83–90. https://doi.org/10.1038/aja.2009.9 Brancaccio, P., Lippi, G., & Maffulli, N. (2010). Biochemical markers of muscular damage. Clinical Chemistry and Laboratory Medicine, 48(6), 757–67. 228 University of Ghana http://ugspace.ug.edu.gh https://doi.org/10.1515/CCLM.2010.179 Brewster, L. M., Coronel, C. M. D., Sluiter, W., Clark, J. F., & van Montfrans, G. A. (2012). Ethnic Differences in Tissue Creatine Kinase Activity: An Observational Study. PLoS ONE, 7(3), e32471. https://doi.org/10.1371/journal. pone.0032471 Brodin, P., Jojic, V., Gao, T., Bhattacharya, S., Angel, C. J. L., Furman, D., … Davis, M. M. (2015). Variation in the Human Immune System Is Largely Driven by Non- Heritable Influences. Cell, 160(1–2), 37–47. https://doi.org/10.1016/J.CELL. 2014.12.020 Brody, T. (1999). Nutritional Biochemistry (2nd ed.). London: Academic Press. Retrieved from https://books.google.com.gh/books?hl=en&lr=&id=n2fgyh DUaTEC&oi=fnd&pg=PP1&dq=Brody+T.+Nutritional+Biochemistry.+2nd+E dition.+San+Diego+CA:+Academic+Press%3B+1999.+&ots=Cu23NLRRE1& sig=33Ck0JH4Hftvv7PnCyP7kZIrAcM&redir_esc=y#v=onepage&q&f=false Brown, S. J., Bremner, A. P., Hadlow, N. C., Feddema, P., Leedman, P. J., O’Leary, P. C., & Walsh, J. P. (2016). The log TSH-free T4 relationship in a community-based cohort is nonlinear and is influenced by age, smoking and thyroid peroxidase antibody status. Clinical Endocrinology, 85(5), 789–796. https://doi.org/10.1111/cen.13107 Calderón, B., Hevia, V., Vega-Piñero, B., Martín-Hidalgo, A., Mendez-del Sol, H., Escobar-Morreale, H. F., & Botella-Carretero, J. I. (2018). Serum retinol, folic acid, and copper are associated with sperm abnormalities in men with obesity. Journal of the American College of Nutrition, 37(3), 194-200. Cappola, A. R., & Ladenson, P. W. (2003). Hypothyroidism and Atherosclerosis. The Journal of Clinical Endocrinology & Metabolism, 88(6), 2438–2444. https://doi.org/10.1210/jc.2003-030398 Carmel, R. (2008). How I treat cobalamin (vitamin B12) deficiency. Blood, 112(6), 2214– 2221. https://doi.org/10.1182/blood-2008-03-040253 Carrivick, S. J., Walsh, J. P., Brown, S. J., Wardrop, R., & Hadlow, N. C. (2015). Brief report: does PTH increase with age, independent of 25-hydroxyvitamin D, phosphate, renal function, and ionized calcium? The Journal of Clinical Endocrinology & Metabolism, 100(5), 2131-2134. Carroll, M. D., Lacher, D. A., Sorlie, P. D., Cleeman, J. I., Gordon, D. J., Wolz, M., ... & Johnson, C. L. (2005). Trends in serum lipids and lipoproteins of adults, 1960- 2002. Jama, 294(14), 1773-1781. 229 University of Ghana http://ugspace.ug.edu.gh Casey, R. G., Hegarty, P. K., Conroy, R., Rea, D., Butler, M. R., Grainger, R., ... & Thornhill, J. A. (2012). The distribution of PSA age-specific profiles in healthy Irish men between 20 and 70. ISRN oncology, 2012. Cauthen, C. A., Lipinski, M. J., Abbate, A., Appleton, D., Nusca, A., Varma, A., … Vetrovec, G. W. (2008). Relation of Blood Urea Nitrogen to Long-Term Mortality in Patients With Heart Failure. The American Journal of Cardiology, 101(11), 1643–1647. https://doi.org/10.1016/j.amjcard.2008.01.047 Centor, R. M. (1990). Serum Total Carbon Dioxide. Clinical Methods: The History, Physical, and Laboratory Examinations. Butterworths. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/21250150 Ceriotti, F. (2007). Prerequisites for use of common reference intervals. The Clinical Biochemist. Reviews, 28(3), 115–21. Ceriotti, F., Boyd, J. C., Klein, G., Henny, J., Queraltó, J., Kairisto, V., … IFCC Committee on Reference Intervals and Decision Limits (C-RIDL). (2008). Reference intervals for serum creatinine concentrations: assessment of available data for global application. Clinical Chemistry, 54(3), 559–66. https://doi.org/10.1373/clinchem.2007.099648 Ceriotti, F., Henny, J., Queraltó, J., Ziyu, S., Özarda, Y., Chen, B., … null, null. (2010). Common reference intervals for aspartate aminotransferase (AST), alanine aminotransferase (ALT) and γ-glutamyl transferase (GGT) in serum: results from an IFCC multicenter study. Clinical Chemistry and Laboratory Medicine, 48(11), 1593–1601. https://doi.org/10.1515/CCLM.2010.315 Ceriotti, F., Hinzmann, R., & Panteghini, M. (2009). Reference intervals: the way forward. Annals of Clinical Biochemistry, 46(1), 8–17. https://doi.org/10.1258/ acb.2008.008170 Chaker, L., Korevaar, T. I. ., Medici, M., Uitterlinden, A. G., Hofman, A., Dehghan, A., … Peeters, R. P. (2016). Thyroid Function Characteristics and Determinants: The Rotterdam Study. Thyroid, 26(9), 1195–1204. https://doi.org/10.1089/thy. 2016.0133 Chao, A., Tang, Y. H., Lai, C. H., Chang, C. J., Chang, S. C., Wu, T. I., ... & Chang, T. C. (2013). Potential of an age-stratified CA125 cut-off value to improve the prognostic classification of patients with endometrial cancer. Gynecologic oncology, 129(3), 500-504. 230 University of Ghana http://ugspace.ug.edu.gh Chen, K. T., Malo, M. S., Moss, A. K., Zeller, S., Johnson, P., Ebrahimi, F., … Hodin, R. A. (2010). Corrigendum. American Journal of Physiology-Renal Physiology, 299(2), F467–F467. https://doi.org/10.1152/ajprenal.zh2-5986-corr.2010 Cheong, E., Ryu, S., Lee, J.-Y., Lee, S. H., Sung, J.-W., Cho, D.-S., … Sung, K.-C. (2017). Association between serum uric acid and cardiovascular mortality and all-cause mortality. Journal of Hypertension, 35, S3–S9. https://doi.org/10.1097/ HJH.0000000000001330 Chien, S.-C., Chen, C.-Y., Lin, C.-F., & Yeh, H.-I. (2017). Critical appraisal of the role of serum albumin in cardiovascular disease. Biomarker Research, 5(1), 31. https://doi.org/10.1186/s40364-017-0111-x Chiras, D. (2013). Human biology. Retrieved from https://books.google.com/ books?hl=en&lr=&id=6QZVAgAAQBAJ&oi=fnd&pg=PR3&dq=Chiras+DD, +Chiras+DEGREE+2013+Human+Biology.+Jones+%26+Bartlett+Learning&o ts=PigZHntZ9S&sig=MWT9ugUriPGOuItNR-QYrj39iic Chisale, M. R., Kumwenda, P., Ngwira, M., M’baya, B., Chosamata, B. I., & Mwapasa, V. (2015a). A pilot study to determine the normal haematological indices for young Malawian adults in Blantyre, Malawi. Malawi Medical Journal: The Journal of Medical Association of Malawi, 27(3), 96–100. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/26715954 Chisale, M. R., Kumwenda, P., Ngwira, M., M’baya, B., Chosamata, B. I., & Mwapasa, V. (2015b). A pilot study to determine the normal haematological indices for young Malawian adults in Blantyre, Malawi. Malawi Medical Journal: The Journal of Medical Association of Malawi, 27(3), 96–100. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/26715954 Cholongitas, E., Shusang, V., Marelli, L., Nair, D., Thomas, M., Patch, D., … Burroughs, A. K. (2007). Review article: renal function assessment in cirrhosis - difficulties and alternative measurements. Alimentary Pharmacology & Therapeutics, 26(7), 969–978. https://doi.org/10.1111/j.1365-2036.2007. 03443.x Christiansen, M., Høgdall, C. K., Andersen, J. R., & Nørgaard-Pedersen, B. (2001). Alpha- fetoprotein in plasma and serum of healthy adults: preanalytical, analytical and biological sources of variation and construction of age-dependent reference intervals. Scandinavian journal of clinical and laboratory investigation, 61(3), 205-215. 231 University of Ghana http://ugspace.ug.edu.gh Cieslak, K. P., Baur, O., Verheij, J., Bennink, R. J., & van Gulik, T. M. (2016). Liver function declines with increased age. HPB, 18(8), 691-696. Ciloglu, F., Peker, I., Pehlivan, A., Karacabey, K., İlhan, N., Saygin, O., & Ozmerdivenli, R. (2005). Exercise intensity and its effects on thyroid hormones. Neuroendocrinology letters, 26(6), 830-834. Clement E. Zeh, C. O. O. and L. A. M. (2012). Laboratory Reference Intervals in Africa. Intechopen.Com, 303–316. https://doi.org/10.1074/jbc.M114.633156 Clinical and Laboratory Standards Institute. (2008). Defining, establishing, and verifying reference intervals in the clinical laboratory; approved guideline. CLSI Document C28-A3 (3rd Edition). Clinical and Laboratory Standards Institute. (2010). Defining, establishing, and verifying reference intervals in the clinical laboratory; Approved Guideline – CLSI Document EPC28-A3c. Third Edition. Clinical and Laboratory Standards Institute. (2010). EP28-A3c Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory; Approved Guideline- Third Edition. Retrieved from www.clsi.org. CLSI and IFCC. (2008). C28-A3 document; Defining, establishing and verifying reference intervals in the clinical laboratory: approved guideline-third edition (Vol. 28). Colombini, A., Corsetti, R., Marco, M., Graziani, R., Lombardi, G., Lanteri, P., … Banfi, G. (2012). Serum Creatine Kinase Activity and Its Relationship With Renal Function Indices in Professional Cyclists During the Giro dʼItalia 3-Week Stage Race. Clinical Journal of Sport Medicine, 22(5), 408–413. https://doi.org/10.1097/JSM.0b013e31825e66cc Concordet, D., Geffré, A., Braun, J. P., & Trumel, C. (2009). A new approach for the determination of reference intervals from hospital-based data. Clinica Chimica Acta, 405(1–2), 43–48. https://doi.org/10.1016/J.CCA.2009.03.057 Cooper, D. S., & Biondi, B. (2012). Subclinical thyroid disease. Lancet (London, England), 379(9821), 1142–54. https://doi.org/10.1016/S0140-6736(11)60276- 6 Corry, D. B., Eslami, P., Yamamoto, K., Nyby, M. D., Makino, H., & Tuck, M. L. (2008). Uric acid stimulates vascular smooth muscle cell proliferation and oxidative stress via the vascular renin–angiotensin system. Journal of hypertension, 26(2), 269-275. 232 University of Ghana http://ugspace.ug.edu.gh Craig, C., & Stitzel, R. (2004). Modern pharmacology with clinical applications. Retrieved from https://books.google.com/books?hl=en&lr=&id=KqA29hQ- m3AC&oi=fnd&pg=PR4&dq=Craig+CR,+Stitzel+RE+2004+Modern+Pharma cology+with+Clinical+Applications.+Lippincott+Williams+%26+Wilkins&ots =DJC7o1I4Ww&sig=ngkSqcFt8HssndyABGXQvVfdozQ Cunningham, G. R., Stephens-Shields, A. J., Rosen, R. C., Wang, C., Ellenberg, S. S., Matsumoto, A. M., … Snyder, P. J. (2015). Association of Sex Hormones With Sexual Function, Vitality, and Physical Function of Symptomatic Older Men With Low Testosterone Levels at Baseline in the Testosterone Trials. The Journal of Clinical Endocrinology & Metabolism, 100(3), 1146–1155. https://doi.org/10.1210/jc.2014-3818 Czarnywojtek, A., Owecki, M., Zgorzalewicz-Stachowiak, M., Woliński, K., Szczepanek- Parulska, E., Budny, B., … Ruchała, M. (2014). The Role of Serum C-Reactive Protein Measured by High-Sensitive Method in Thyroid Disease. Archivum Immunologiae et Therapiae Experimentalis, 62(6), 501–509. https://doi.org/10.1007/s00005-014-0282-1 Dale, J. C., Burritt, M. F., & Zinsmeister, A. R. (2002). Clinical Chemistry / DIURNAL VARIATION OF SERUM IRON LEVELS Diurnal Variation of Serum Iron, Iron- Binding Capacity, Transferrin Saturation, and Ferritin Levels. Am J Clin Pathol (Vol. 117). Retrieved from https://academic.oup.com/ajcp/article- abstract/117/5/802/1758685 Dalton, R. N., Kilpatrick, E. S., Nichols, S. P., & Maylor, P. W. (2010). Serum creatinine and glomerular filtration rate: perception and reality. Clinical Chemistry, 56(5), 687–9. https://doi.org/10.1373/clinchem.2010.144261 Das, A. K., Chandra, P., Gupta, A., & Ahmad, N. (2015). Obesity and the levels of liver enzymes (ALT, AST & GGT) in East Medinipur, India. Asian Journal of Medical Sciences, 6(1), 40-42. Dasgupta, A., & Dasgupta, A. (2015). Liver Enzymes as Alcohol Biomarkers. Alcohol and Its Biomarkers, 121–137. https://doi.org/10.1016/B978-0-12-800339-8.00005-5 DeFina, L., Radford, N., Leonard, D., Howard, E., Wilson, R., Cooper, T., ... & Gruntmanis, U. (2015). Testosterone level in men correlates with BMI and cardiorespiratory fitness but is not related to age. In 17th European Congress of Endocrinology (Vol. 37). BioScientifica. 233 University of Ghana http://ugspace.ug.edu.gh Delanaye, P., Glassock, R. J., Pottel, H., & Rule, A. D. (2016). An age-calibrated definition of chronic kidney disease: rationale and benefits. The Clinical Biochemist Reviews, 37(1), 17. Delanaye, P., Schaeffner, E., Ebert, N., Cavalier, E., Mariat, C., Krzesinski, J.-M., & Moranne, O. (2012). Normal reference values for glomerular filtration rate: what do we really know? Nephrology Dialysis Transplantation, 27(7), 2664–2672. https://doi.org/10.1093/ndt/gfs265 Delanaye, P., Schaeffner, E., Ebert, N., Cavalier, E., Mariat, C., Krzesinski, J. M., & Moranne, O. (2012). Normal reference values for glomerular filtration rate: what do we really know? Nephrology Dialysis Transplantation, 27(7), 2664-2672. Denic, A., Glassock, R. J., & Rule, A. D. (2016). Structural and functional changes with the aging kidney. Advances in chronic kidney disease, 23(1), 19-28. Devaraj, S., Singh, U., Chemistry, I. J.-C., & 2009, U. (2009). The evolving role of C- reactive protein in atherothrombosis. Clinchem.Aaccjnls.Org, 55, 229–238. Retrieved from http://clinchem.aaccjnls.org/content/55/2/229.short Devi, S., Garg, N., Verma, P., & Jain, N. (2018). Association between Serum Total Testosterone and Obesity in Healthy Adults. International Journal of Physiology, 6(1), 117-122. Diana Nicoll, C. (2007). Appendix: Therapeutic drug monitoring and laboratory reference ranges. Current medical diagnosis and treatment. Stephen JM, Maxine AP. 46th edition, Mc Graw hill, 1767-1775. Diaz-Arjonilla, M., Schwarcz, M., Swerdloff, R. S., & Wang, C. (2009). Obesity, low testosterone levels and erectile dysfunction. International journal of impotence research, 21(2), 89. Dosoo, D. K., Kayan, K., Adu-Gyasi, D., Kwara, E., Ocran, J., Osei-Kwakye, K., … Owusu-Agyei, S. (2012). Haematological and biochemical reference values for healthy adults in the middle belt of Ghana. PloS One, 7(4), e36308. https://doi.org/10.1371/journal.pone.0036308 Dousdampanis, P., Trigka, K., & Fourtounas, C. (2012). Diagnosis and management of chronic kidney disease in the elderly: a field of ongoing debate. Aging and disease, 3(5), 360. Douville, P., Martel, A. R., Talbot, J., Desmeules, S., Langlois, S., & Agharazii, M. (2008). Impact of age on glomerular filtration estimates. Nephrology Dialysis Transplantation, 24(1), 97-103. 234 University of Ghana http://ugspace.ug.edu.gh Du, Y., Chen, H., Sun, H., Wang, C., Wang, C., & Li, Y. (2016). Investigation and Analysis of Reference Intervals for Blood Cell Parameters in a Healthy Population from Daxingan Region Inner Mongolia. Clinical Laboratory, 62(1– 2), 129–34. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/27012042 Eckardt, K. U., Berns, J. S., Rocco, M. V., & Kasiske, B. L. (2009). Definition and classification of CKD: the debate should be about patient prognosis—a position statement from KDOQI and KDIGO. American journal of kidney diseases, 53(6), 915-920. Egbuna, O. I., & Brown, E. M. (2011). Hypoparathyroidism. In Immunoendocrinology: Scientific and Clinical Aspects (pp. 501-517). Humana Press, Totowa, NJ. Eller, L. A., Eller, M. A., Ouma, B., Kataaha, P., Kyabaggu, D., Tumusiime, R., ... & Kibuuka, H. (2008). Reference intervals in healthy adult Ugandan blood donors and their impact on conducting international vaccine trials. PLOS one, 3(12), e3919. Emilsson, V., Thorleifsson, G., Zhang, B., Leonardson, A. S., Zink, F., Zhu, J., … Stefansson, K. (2008). Genetics of gene expression and its effect on disease. Nature, 452(7186), 423–428. https://doi.org/10.1038/nature06758 Emmett, M. (2006). Anion-gap interpretation: the old and the new. Nature Clinical Practice Nephrology, 2(1), 4–5. https://doi.org/10.1038/ncpneph0073 Eriksson, J., Haring, R., Grarup, N., Vandenput, L., Wallaschofski, H., Lorentzen, E., ... & Lorentzon, M. (2017). Causal relationship between obesity and serum testosterone status in men: A bi-directional mendelian randomization analysis. PloS one, 12(4), e0176277. Filippatos, G., Rossi, J., Lloyd-Jones, D. M., Stough, W. G., Ouyang, J., Shin, D. D., … Gheorghiade, M. (2007). Prognostic Value of Blood Urea Nitrogen in Patients Hospitalized with Worsening Heart Failure: Insights From the Acute and Chronic Therapeutic Impact of a Vasopressin Antagonist in Chronic Heart Failure (ACTIV in CHF) Study. Journal of Cardiac Failure, 13(5), 360–364. https://doi.org/10.1016/j.cardfail.2007.02.005 Fontes, R., Coeli, C., Aguiar, F., & Vaisman, M. (2013). Reference interval of thyroid stimulating hormone and free thyroxine in a reference population over 60 years old and in very old subjects (over 80 years): comparison to young subjects. Thyroid Research, 6(1), 13. https://doi.org/10.1186/1756-6614-6-13 235 University of Ghana http://ugspace.ug.edu.gh Freedman, N. D., Everhart, J. E., Lindsay, K. L., Ghany, M. G., Curto, T. M., Shiffman, M. L., … Sinha, R. (2009). Coffee intake is associated with lower rates of liver disease progression in chronic hepatitis C. Hepatology, 50(5), 1360–1369. https://doi.org/10.1002/hep.23162 Friedrich, N., Völzke, H., Rosskopf, D., Steveling, A., Krebs, A., Nauck, M., & Wallaschofski, H. (2008). Reference ranges for serum dehydroepiandrosterone sulfate and testosterone in adult men. Journal of andrology, 29(6), 610-617. Frith, J., Day, C. P., Henderson, E., Burt, A. D., & Newton, J. L. (2009). Non-alcoholic fatty liver disease in older people. Gerontology, 55(6), 607-613. Fuentes-Arderiu, X. (2006). Biological reference intervals and ISO 15189. Clinica Chimica Acta, 364(1–2), 365–366. https://doi.org/10.1016/j.cca.2005.07.014 Fuentes-Arderiu, X., Ferré-Masferrer, M., Gonz%aacute;lez-Alba, J. M., Escolà-Aliberas, J., Balsells-Rosello, D., Blanco-Cristobal, C., … Vicens-Manero, M. (2001). Multicentric reference values for some quantities measured with Tina-Quant® reagents systems and RD/Hitachi analysers. Scandinavian Journal of Clinical and Laboratory Investigation, 61(4), 273–276. https://doi.org/10.1080/ 00365510152378987a Fukui, M., Tanaka, M., Shiraishi, E., Harusato, I., Hosoda, H., Asano, M., … Nakamura, N. (2008). Relationship between serum bilirubin and albuminuria in patients with type 2 diabetes. Kidney International, 74(9), 1197–1201. https://doi.org/10.1038/KI.2008.398 Furusyo, N., Ai, M., Okazaki, M., Ikezaki, H., Ihara, T., Hayashi, T., … Hayashi, J. (2013). Serum cholesterol and triglyceride reference ranges of twenty lipoprotein subclasses for healthy Japanese men and women. Atherosclerosis, 231(2), 238– 245. https://doi.org/10.1016/j.atherosclerosis.2013.09.008 Galukande, M., Jombwe, J., Fualal, J., Baingana, R., & Gakwaya, A. (2011). Reference values for serum levels of folic acid and vitamin B12 in a young adult Ugandan population. African health sciences, 11(2). Gan, L., Chitturi, S., & Farrell, G. C. (2011). Mechanisms and implications of age-related changes in the liver: nonalcoholic fatty liver disease in the elderly. Current gerontology and geriatrics research, 2011. Gao, C., Yang, B., Guo, Q., Wei, L., & Tian, L. (2014). High Thyroid-Stimulating Hormone Level is Associated with the Risk of Developing Atherosclerosis in Subclinical Hypothyroidism. Hormone and Metabolic Research, 47(03), 220– 236 University of Ghana http://ugspace.ug.edu.gh 224. https://doi.org/10.1055/s-0034-1394370 Geffré, A., Friedrichs, K., Harr, K., Concordet, D., Trumel, C., & Braun, J.-P. (2009). Reference values: a review. Veterinary Clinical Pathology, 38(3), 288–298. https://doi.org/10.1111/j.1939-165X.2009.00179.x Giannini, E., Botta, F., Fasoli, A., Romagnoli, P., Mastracci, L., Ceppa, P., … Testa, R. (2001). Increased levels of gammaGT suggest the presence of bile duct lesions in patients with chronic hepatitis C: absence of influence of HCV genotype, HCV-RNA serum levels, and HGV infection on this histological damage. Digestive Diseases and Sciences, 46(3), 524–9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11318526 Giroux Leprieur, E., Lavole, A., Ruppert, A.-M., Gounant, Val., Wislez, M., Cadranel, J., & Milleron, B. (2012). Factors associated with long-term survival of patients with advanced non-small cell lung cancer. Respirology, 17(1), 134–142. https://doi.org/10.1111/j.1440-1843.2011.02070.x Glassock, R. J., & Winearls, C. (2009). Ageing and the glomerular filtration rate: truths and consequences. Transactions of the American Clinical and Climatological Association, 120, 419. Gledhill, R. F., Van der Merwe, C. A., Greyling, M., & Van Niekerk, M. M. (1988). Race- gender differences in serum creatine kinase activity: a study among South Africans. Journal of Neurology, Neurosurgery & Psychiatry, 51(2), 301-304. Goossens, G. H., Blaak, E. E., & Van Baak, M. A. (2003). Possible involvement of the adipose tissue renin‐angiotensin system in the pathophysiology of obesity and obesity‐related disorders. Obesity Reviews, 4(1), 43-55. Gostynski, M., F. Gutzwiller, K. Kuulasmaa, A. Döring, M. Ferrario, D. Grafnetter, and A. Pajak. "Analysis of the relationship between total cholesterol, age, body mass index among males and females in the WHO MONICA Project." International journal of obesity 28, no. 8 (2004): 1082. Gowda, S., Desai, P. B., Hull, V. V, Math, A. A. K., Vernekar, S. N., & Kulkarni, S. S. (2009). A review on laboratory liver function tests. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2984286/pdf/pamj-03-17.pdf Gräsbeck, R. (2004). The evolution of the reference value concept. Clinical Chemistry and Laboratory Medicine (CCLM), 42(7), 692–7. https://doi.org/10 .1515/CCLM.2004.118 237 University of Ghana http://ugspace.ug.edu.gh Green, R. M., & Flamm, S. (2002). AGA technical review on the evaluation of liver chemistry tests. Gastroenterology, 123(4), 1367–84. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12360498 Greg Miller, W., Horowitz, G. L., Ceriotti, F., Fleming, J. K., Greenberg, N., Katayev, A., … Professor, A. (2016). Reference Intervals: Strengths, Weaknesses, and Challenges. Clinical Chemistry, 62(7), 916–923. Retrieved from http://clinchem.aaccjnls.org/content/clinchem/62/7/916.full.pdf Gupta, A., Gupta, D., Raizada, A., Gupta, N. P., Yadav, R., Vinayak, K., & Tewari, V. (2014). A hospital-based study on reference range of serum prostate specific antigen levels. The Indian journal of medical research, 140(4), 507. Gür, T., Demir, H., & Kotan, M. Ç. (2011). Tumour markers and biochemical parameters in colon cancer patients before and after chemotherapy. Asian Pacific Journal of Cancer Prevention : APJCP, 12(11), 3147–50. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/22394005 Haden, S. T., Brown, E. M., Hurwitz, S., Scott, J., & Fuleihan, G. E. H. (2000). The effects of age and gender on parathyroid hormone dynamics. Clinical endocrinology, 52(3), 329-338. Hallworth, M. J. (2011). The “70% claim”: what is the evidence base? Annals of Clinical Biochemistry, 48(6), 487–488. https://doi.org/10.1258/acb.2011.011177 Hammerling, J. A. (2012). A Review of Medical Errors in Laboratory Diagnostics and Where We Are Today: Table 1. Laboratory Medicine, 43(2), 41–44. https://doi.org/10.1309/LM6ER9WJR1IHQAUY Harm, S., Schildböck, C., & Hartmann, J. (2018). Removal of stabilizers from human serum albumin by adsorbents and dialysis used in blood purification. PLOS ONE, 13(1), e0191741. https://doi.org/10.1371/journal.pone.0191741 Harris, E. K., & Boyd, J. C. (1990). On Dividing Reference Data into Subgroups to Produce Separate Reference Ranges. CLIN. CHEM. CLINICAL CHEMISTRY, 362(2), 265–270. Harris, E. K., Wong, E. T., & Shaw, S. T. (1991). Statistical criteria for separate reference intervals: race and gender groups in creatine kinase. Clinical Chemistry, 37(9), 1580–2. Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/1893593 Harris, E., & Boyd, J. (1995). Statistical bases of reference values in laboratory medicine. CRC Press. Retrieved from https://content.taylorfrancis.com/books/ download?dac=C2006-0-04182- 238 University of Ghana http://ugspace.ug.edu.gh 6&isbn=9781482273151&format=googlePreviewPdf Hawkins, R. (2012). Manageing the pre- and post-analytical phases of the total testing process. Annals of Laboratory Medicine, 32(1), 5–16. https://doi.org/10.3343/ alm.2012.32.1.5 He, D., Wang, M., Chen, X., Gao, Z., He, H., Zhau, H. E., … Nan, X. (2004). Ethnic differences in distribution of serum prostate-specific antigen: a study in a healthy chinese male population. Urology, 63(4), 722–726. https://doi.org/ 10.1016/J. UROLOGY.2003.10.066 Headstrom, P. D., Rulyak, S. J., & Lee, S. D. (2008). Prevalence of and risk factors for vitamin B12 deficiency in patients with Crohnʼs disease. Inflammatory Bowel Diseases, 14(2), 217–223. https://doi.org/10.1002/ibd.20282 Healy, G. N., Eakin, E. G., Owen, N., LaMontagne, A. D., Moodie, M., Winkler, E. A., ... & Dunstan, D. W. (2016). A cluster randomized controlled trial to reduce office workers' sitting time: impact on activity outcomes. Medicine & science in sports & exercise, 48(9), 1787-1797. Heidegger, I., Skradski, V., Steiner, E., Klocker, H., Pichler, R., Pircher, A., ... & Bektic, J. (2015). High risk of under-grading and-staging in prostate cancer patients eligible for active surveillance. PLoS One, 10(2), e0115537. Heikkilä, K., Ebrahim, S., & Lawlor, D. A. (2007). A systematic review of the association between circulating concentrations of C reactive protein and cancer. Journal of Epidemiology and Community Health, 61(9), 824–33. https://doi.org/10.1136/jech.2006.051292 Henny, J., Petitclerc, C., Fuentes-Arderiu, X., Petersen, P. H., Queraltó, J. M., Schiele, F., & Siest, G. (2000). Need for Revisiting the Concept of Reference Values. Clinical Chemistry and Laboratory Medicine, 38(7), 589–95. https://doi.org/10.1515/CCLM.2000.085 Herman, W. H., & Cohen, R. M. (2012). Racial and Ethnic Differences in the Relationship between HbA1c and Blood Glucose: Implications for the Diagnosis of Diabetes. https://doi.org/10.1210/jc.2011-1894 Hoffmann, J. J. M. L., Nabbe, K. C. A. M., & van den Broek, N. M. A. (2015). Effect of age and gender on reference intervals of red blood cell distribution width (RDW) and mean red cell volume (MCV). Clinical Chemistry and Laboratory Medicine (CCLM), 53(12), 2015–9. https://doi.org/10.1515/cclm-2015-0155 239 University of Ghana http://ugspace.ug.edu.gh Horn, P. S., & Pesce, A. J. (2003). Reference intervals: an update. Clinica Chimica Acta, 334(1–2), 5–23. https://doi.org/10.1016/S0009-8981(03)00133-5 Horn, P. S., Feng, L., Li, Y., & Pesce, A. J. (2001). Effect of Outliers and Nonhealthy Individuals on Reference Interval Estimation. Clinical Chemistry, 47(12). Hsieh, M. H., Ho, C. K., Hou, N. J., Hsieh, M. Y., Lin, W. Y., Yang, J. F., ... & Dai, C. Y. (2009). Abnormal liver function test results are related to metabolic syndrome and BMI in Taiwanese adults without chronic hepatitis B or C. International journal of obesity, 33(11), 1309. Ichihara, K. (2014). Statistical considerations for harmonization of the global multicenter study on reference values. Clinica Chimica Acta, 432, 108–118. https://doi.org/10.1016/J.CCA.2014.01.025 Ichihara, K., & Boyd, J. C. (2010). An appraisal of statistical procedures used in derivation of reference intervals. Clinical Chemistry and Laboratory Medicine, 48(11), 1537–1551. https://doi.org/10.1515/CCLM.2010.319 Ichihara, K., Ceriotti, F., Kazuo, M., Huang, Y. Y., Shimizu, Y., Suzuki, H., … Okubo, Y. (2013b). The Asian project for collaborative derivation of reference intervals: (2) results of non-standardized analytes and transference of reference intervals to the participating laboratories on the basis of cross-comparison of test results. Clinical Chemistry and Laboratory Medicine, 51(7), 1443–1457. https://doi.org/10.1515/cclm-2012-0422 Ichihara, K., Ceriotti, F., Tam, T. H., Sueyoshi, S., Poon, P. M. K., Thong, M. L., … Committee on Reference Intervals and Decision Limits, International Federation for Clinical Chemistry and Laboratory Medicine, and the Science Committee for the Asia-Pacific Federation of Clinical Biochemistry. (2013a). The Asian project for collaborative derivation of reference intervals: (1) strategy and major results of standardized analytes. Clinical Chemistry and Laboratory Medicine, 51(7), 1429–42. https://doi.org/10.1515/cclm-2012-0421 Ichihara, K., Itoh, Y., Lam, C. W. K., Poon, P. M. K., Kim, J.-H., Kyono, H., … Science Committee for the Asian-Pacific Federation of Clinical Biochemistry. (2008). Sources of variation of commonly measured serum analytes in 6 Asian cities and consideration of common reference intervals. Clinical Chemistry, 54(2), 356–65. https://doi.org/10.1373/clinchem.2007.091843 Ichihara, K., Itoh, Y., Min, W.-K., Yap, S. F., Lam, C. W. K., Kong, X. T., … Committee on Plasma Proteins, International Federation of Clinical Chemistry and 240 University of Ghana http://ugspace.ug.edu.gh Laboratory Medicine. (2004). Diagnostic and epidemiological implications of regional differences in serum concentrations of proteins observed in six Asian cities. Clinical Chemistry and Laboratory Medicine (CCLM), 42(7), 800–9. https://doi.org/10.1515/CCLM.2004.133 Ichihara, K., Ozarda, Y., Barth, J. H., Klee, G., Qiu, L., Erasmus, R., … Yadav, B. K. (2017a). A global multicenter study on reference values: 1. Assessment of methods for derivation and comparison of reference intervals. Clinica Chimica Acta, 467, 70–82. https://doi.org/10.1016/j.cca.2016.09.016 Ichihara, K., Ozarda, Y., Barth, J. H., Klee, G., Qiu, L., Erasmus, R., … Yadav, B. K. (2017b). A global multicenter study on reference values: 1. Assessment of methods for derivation and comparison of reference intervals. Clinica Chimica Acta, 467, 70–82. https://doi.org/10.1016/j.cca.2016.09.016 Ichihara, K., Ozarda, Y., Barth, J. H., Klee, G., Shimizu, Y., Xia, L., … Takahashi, A. (2017). A global multicenter study on reference values: 2. Exploration of sources of variation across the countries. Clinica Chimica Acta, 467, 83–97. https://doi.org/10.1016/j.cca.2016.09.015 Ichihara, K., Ozarda, Y., Klee, G., Straseski, J., Baumann, N., & Ishikura, K. (2013c). Utility of a panel of sera for the alignment of test results in the worldwide multicenter study on reference values. Clinical Chemistry and Laboratory Medicine, 51(5). https://doi.org/10.1515/cclm-2013-0248 Ichihara, K., Saito, K., & Itoh, Y. (2007). Sources of variation and reference intervals for serum cystatin C in a healthy Japanese adult population. Clinical Chemical Laboratory Medicine, 45(9), 1232-1236. International Federation of Clinical Chemistry (IFCC), & International Committee for Standardization in Haematology (ICSH). (1987). Approved Recommendation (1986) on the Theory of Reference Values. Journal of Clin. Chem. Clin. Biochem., 25, 337–342. Islam, S., Choudhury, K. N., Mainuddin, A., & Wahiduzzaman, M. (2014). Serum lipid profile and its association with hypertension in Bangladesh. Vascular Health and Risk Management, 327. https://doi.org/10.2147/VHRM.S61019 Ittermann, T., Roser, M., Wood, G., Preez, H., Lüdemann, J., Völzke, H., & Nauck, M. (2010). Reference intervals for eight measurands of the blood count in a large population-based study. Clinical laboratory, 56(1-2), 9-19. 241 University of Ghana http://ugspace.ug.edu.gh Jacobs, I., & Bast, R. C. (1989). The CA 125 tumour-associated antigen: a review of the literature. Human Reproduction, 4(1), 1–12. https://doi.org/10.1093/ oxfordjournals.humrep.a136832 Jalali, M. T., Honomaror, A. M., Rekabi, A., & Latifi, M. (2013). Reference Ranges for Serum Total Cholesterol, HDL-Cholesterol, LDL-Cholesterol, and VLDL- Cholesterol and Triglycerides in Healthy Iranian Ahvaz Population. Indian Journal of Clinical Biochemistry, 28(3), 277. https://doi.org/10.1007/S12291- 012-0268-X Jannie, W., Treuting, J. J., & Donald, C. C. (1979). Metabolic intermediates and inorganic ions. Clinincal diagnosis and management; 16th edition; Philadelphia; London and Toronto; WB Saunders company, 259-304. Jialal, I., Devaraj, S., & Venugopal, S. K. (2004). C-Reactive Protein: Risk Marker or Mediator in Atherothrombosis? Hypertension, 44(1), 6–11. https://doi.org/10.1161/01.HYP.0000130484.20501.df Jirtle, R. L., & Skinner, M. K. (2007). Environmental epigenomics and disease susceptibility. Nature Reviews Genetics, 8(4), 253–262. https://doi.org/10.1038 /nrg2045 Johnson, A. M., Petersen, P. H., Whicher, J. T., Carlström, A., MacLennan, S., & International Federation of Clinical Chemistry and Laboratory Medicine, Committee on Plasma Proteins. (2004). Reference intervals for plasma proteins: similarities and differences between adult Caucasian and Asian Indian males in Yorkshire, UK. Clinical Chemistry and Laboratory Medicine (CCLM), 42(7), 792–9. https://doi.org/10.1515/CCLM.2004.132 Johnson, C. C., Kessel, B., Riley, T. L., Ragard, L. R., Williams, C. R., Xu, J. L., & Buys, S. S. (2008). The epidemiology of CA-125 in women without evidence of ovarian cancer in the Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) Screening Trial. Gynecologic oncology, 110(3), 383-389. Johnson, P. J. (1999). Role of alpha-fetoprotein in the diagnosis and management of hepatocellular carcinoma. Journal of Gastroenterology and Hepatology, 14 Suppl, S32-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10382636 Jones, G., & Barker, A. (2008). Reference intervals. The Clinical Biochemist. Reviews, 29 Suppl 1, S93-7. Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/18852866 Joo, E.-Y., Kim, Y. J., Go, Y., & Song, J.-G. (2018). Relationship between perioperative thyroid function and acute kidney injury after thyroidectomy. Scientific Reports, 242 University of Ghana http://ugspace.ug.edu.gh 8(1), 13539. https://doi.org/10.1038/s41598-018-31946-w Kalantar-Zadeh, K., Streja, E., Molnar, M. Z., Lukowsky, L. R., Krishnan, M., Kovesdy, C. P., & Greenland, S. (2012). Mortality prediction by surrogates of body composition: an examination of the obesity paradox in hemodialysis patients using composite ranking score analysis. American Journal of Epidemiology, 175(8), 793–803. https://doi.org/10.1093/aje/kwr384 Kallistratos, M., Giannitsi, S., Poulimenos, L. E., Iliou, K., Koukouzeli, A., Kouremenos, N., … Manolis, A. J. (2017). [PP.21.07] SERUM URIC ACID LEVELS AND 10-YEAR RISK OF FATAL CARDIOVASCULAR DISEASE. A PROSPECTIVE STUDY. Journal of Hypertension, 35, e271. https://doi.org/10.1097/01.hjh.0000523785.70737.db Kallistratos, M., Giannitsi, S., Poulimenos, L., Miaris, N., Koukouzeli, A., Khashlok, L. A., … Manolis., A. J. (2018). URIC ACID AS A RISK FACTOR FOR CARDIOVASCULAR DISEASE. A PROSPECTIVE OBSERVATIONAL STUDY. Journal of Hypertension, 36, e138. https://doi.org/10.1097/01.hjh. 0000539359.83742.00 Kamath, P. S., Wiesner, R. H., Malinchoc, M., Kremers, W., Therneau, T. M., Kosberg, C. L., ... & Kim, W. R. (2001). A model to predict survival in patients with end- stage liver disease. Hepatology, 33(2), 464-470. Kansui, Y., Matsumura, K., Morinaga, Y., Inoue, M., Kiyohara, K., Ohta, Y., … Kitazono, T. (2018). Impact of serum uric acid on incident hypertension in a worksite population of Japanese men. Journal of Hypertension, 36(7), 1499–1505. https://doi.org/10.1097/HJH.0000000000001743 Kaptoge, S., Di Angelantonio, E., Lowe, G., Pepys, M. B., Thompson, S. G., Collins, R., & Danesh, J. (2010). C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. The Lancet, 375(9709), 132–140. https://doi.org/10.1016/S0140-6736(09)61717-7 Karim, S. M. F., Rahman, M. R., Shermin, S., & Sultana, R. (2015). Correlation between Aminotransferase Ratio (AST/ALT) and Other Biochemical Parameters in Chronic Liver Disease of Viral Origin. Delta Medical College Journal, 3(1), 13– 17. https://doi.org/10.3329/dmcj.v3i1.22234 Karita, E., Ketter, N., Price, M. A., Kayitenkore, K., Kaleebu, P., Nanvubya, A., … Kamali, A. (2009). CLSI-Derived Hematology and Biochemistry Reference Intervals for Healthy Adults in Eastern and Southern Africa. PLoS ONE, 4(2), 243 University of Ghana http://ugspace.ug.edu.gh e4401. https://doi.org/10.1371/journal.pone.0004401 Katayev, A., Balciza, C., & Seccombe, D. W. (2010). Establishing Reference Intervals for Clinical Laboratory Test Results Is There a Better Way? Am J Clin Pathol, 133, 180–186. https://doi.org/10.1309/AJCPN5BMTSF1CDYP Kato, I., Dnistrian, A., Schwartz, M., Toniolo, P., Koenig, K., Shore, R., ... & Riboli, E. (2000). Risk of iron overload among middle-aged women. International journal for vitamin and nutrition research, 70(3), 119-125. Khan, S. (2006). Evaluation of hyperbilirubinemia in acute inflammation of appendix: a prospective study of 45 cases. Kathmandu University Medical Journal (KUMJ), 4(3), 281–9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/18603920 Kibaya, R. S., Bautista, C. T., Sawe, F. K., Shaffer, D. N., & Sateren, W. B. (2008). Reference Ranges for the Clinical Laboratory Derived from a Rural Population in Kericho. PLoS ONE, 3(10), 3327. https://doi.org/10.1371/journal.pone. 0003327 Kim, C., Golden, S. H., Mather, K. J., Laughlin, G. A., Kong, S., Nan, B., ... & Diabetes Prevention Program Research Group. (2012). Racial/ethnic differences in sex hormone levels among postmenopausal women in the diabetes prevention program. The Journal of Clinical Endocrinology & Metabolism, 97(11), 4051- 4060. Kim, H., Kisseleva, T., & Brenner, D. A. (2015). Aging and liver disease. Current opinion in gastroenterology, 31(3), 184. Kim, J. H., & Park, J. Y. (2017). The significance of preoperative serum cancer antigen 125 in malignant ovarian germ cell tumours. Gynecologic Oncology, 145, 215– 216. https://doi.org/10.1016/j.ygyno.2017.03.496 Kim, J., & Jo, I. (2010). Relationship between body mass index and alanine aminotransferase concentration in non-diabetic Korean adults. European journal of clinical nutrition, 64(2), 169. Kim, M., Moon, K. H., Kim, T. H., & Ahn, T. Y. (2019). 135 Reference Intervals for Serum Testosterone Levels in Korean: Results from Large Health Examination Data. The Journal of Sexual Medicine, 16(4), S70. Koerbin, G., Cavanaugh, J. A., Potter, J. M., Abhayaratna, W. P., West, N. P., Glasgow, N., … Hickman, P. E. (2015). ‘Aussie normals’: an a priori study to develop clinical chemistry reference intervals in a healthy Australian population. Pathology, 47(2), 138–144. https://doi.org/10.1097/PAT.0000000000000227 244 University of Ghana http://ugspace.ug.edu.gh Kone, B., Maiga, M., Baya, B., Sarro, Y., Coulibaly, N., Kone, A., … Siddiqui, S. (2017). Establishing Reference Ranges of Haematological Parameters from Malian Healthy Adults. Journal of Blood & Lymph, 7(1). https://doi.org/10.4172/2165- 7831.1000154 Koram, K. A., Addae, M. M., Ocran, J. C., Adu-Amankwah, S., Rogers, W. O., & Nkrumah, F. K. (2007). Number 4 GHANA MEDICAL JOURNAL (Vol. 41). Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2350114/ pdf/GMJ4104-0160.pdf Kratz, A., & Lewandrowski, K. B. (1998). Normal reference laboratory values. New England Journal of Medicine, 339(15), 1063-1072. Kratz, A., Ferraro, M., Sluss, P. M., & Lewandrowski, K. B. (2004). Laboratory reference values. New England Journal of Medicine, 351, 1548-1564. Kraut, J. A., & Madias, N. E. (2007). Serum anion gap: its uses and limitations in clinical medicine. Clinical Journal of the American Society of Nephrology : CJASN, 2(1), 162–74. https://doi.org/10.2215/CJN.03020906 Kraut, J. A., & Nagami, G. T. (2013). The serum anion gap in the evaluation of acid-base disorders: what are its limitations and can its effectiveness be improved? Clinical Journal of the American Society of Nephrology : CJASN, 8(11), 2018–24. https://doi.org/10.2215/CJN.04040413 Kueviakoe, I. M., Segbena, A. Y., Jouault, H., Vovor, A., & Imbert, M. (2011). Haematological Reference Values for Healthy Adults in Togo. ISRN Hematology, 2011, 1–5. https://doi.org/10.5402/2011/736062 Kulasingam, V., & Diamandis, E. P. (2008). Strategies for discovering novel cancer biomarkers through utilization of emerging technologies. Nature Clinical Practice Oncology, 5(10), 588–599. https://doi.org/10.1038/ncponc1187 Kumar, D., & Sharma, C. B. (2017). A Comparative Study of Serum Alt &Ast Level And Ast/Alt Ratio In Alcoholic And Non-Alcoholic Acute Uncomplicated Falciparum Malaria Without Clinical Jaundice. IOSR Journal of Dental and Medical Sciences (IOSR-JDMS) e-ISSN, 16(1), 6-08. https://doi.org/10.9790/0853-1601090608 Kumar, V., & Gill, K. D. (2018). To Estimate the Activity of Alkaline Phosphatase in Serum. In Basic Concepts in Clinical Biochemistry: A Practical Guide (pp. 107– 109). Singapore: Springer Singapore. https://doi.org/10.1007/978-981-10-8186- 6_26 245 University of Ghana http://ugspace.ug.edu.gh Kunutsor, S. K., Kieneker, L. M., Burgess, S., Bakker, S. J. L., & Dullaart, R. P. F. (2017). Circulating Total Bilirubin and Future Risk of Hypertension in the General Population: The Prevention of Renal and Vascular End‐Stage Disease (PREVEND) Prospective Study and a Mendelian Randomization Approach. Journal of the American Heart Association, 6(11), e006503. https://doi.org/ 10.1161/JAHA.117.006503 Kushner, I., Rzewnicki, D., & Samols, D. (2006). What Does Minor Elevation of C- Reactive Protein Signify? The American Journal of Medicine, 119(2), 166.e17- 166.e28. https://doi.org/10.1016/j.amjmed.2005.06.057 Kussmaul, T., Greiser, K. H., Haerting, J., Werdan, K., Thiery, J., & Kratzsch, J. (2014). Thyroid analytes TSH, FT3 and FT4 in serum of healthy elderly subjects as measured by the Roche modular system: do we need age and gender dependent reference levels? Clinical Laboratory, 60(9), 1551–9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/25291952 Kuster, N., Bargnoux, A.-S., Pageaux, G.-P., & Cristol, J.-P. (2012). Limitations of compensated Jaffe creatinine assays in cirrhotic patients. Clinical Biochemistry, 45(4–5), 320–325. https://doi.org/10.1016/j.clinbiochem.2011.11.008 Kuzuya, M., Ando, F., Iguchi, A., & Shimokata, H. (2002). Changes in serum lipid levels during a 10 year period in a large Japanese population: a cross-sectional and longitudinal study. Atherosclerosis, 163(2), 313-320. Kyriakides, T., Angelini, C., Schaefer, J., Sacconi, S., Siciliano, G., Vilchez, J. J., & Hilton-Jones, D. (2010). EFNS guidelines on the diagnostic approach to pauci- or asymptomatic hyperCKemia. European Journal of Neurology, 17(6), 767– 773. https://doi.org/10.1111/j.1468-1331.2010.03012.x Lahiri, K. D., Baruah, M., Ghosh, J., & Sengupta, S. (2014). Establishment of reference interval of serum prolactin in an Indian population. Journal of Clinical and Diagnostic Research : JCDR, 8(7), CC08-10. https://doi.org/10.7860/JCDR/ 2014/8400.4599 Lala, V., & Minter, D. A. (2018). Liver Function Tests. In StatPearls [Internet]. StatPearls Publishing. Lanteri, P., Lombardi, G., Colombini, A., & Banfi, G. (2013). Vitamin D in exercise: Physiologic and analytical concerns. Clinica Chimica Acta, 415, 45–53. https://doi.org/10.1016/j.cca.2012.09.004 246 University of Ghana http://ugspace.ug.edu.gh Laurberg, P., Andersen, S., Carlé, A., Karmisholt, J., Knudsen, N., & Pedersen, I. B. (2011). The TSH upper reference limit: where are we at? Nature Reviews Endocrinology, 7(4), 232–239. https://doi.org/10.1038/nrendo.2011.13 Lawrie, D., Coetzee, L. M., Becker, P., Mahlangu, J., Stevens, W., & Glencross, D. K. (2009). Local reference ranges for full blood count and CD4 lymphocyte count testing. South African Medical Journal, 99(4), 243–8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/19588777 Lee, S. J., Lenton, E. A., Sexton, L., & Cooke, I. D. (1988). The effect of age on the cyclical patterns of plasma LH, FSH, oestradiol and progesterone in women with regular menstrual cycles. Human reproduction, 3(7), 851-855. Leng, X. H., & Bray, P. F. (2005). Hormone therapy and platelet function. Drug Discovery Today: Disease Mechanisms, 2(1), 85-91. Leverenz, D., Zaha, O., Crofford, L. J., & Chung, C. P. (2016). Causes of creatine kinase levels greater than 1000 IU/L in patients referred to rheumatology. Clinical Rheumatology, 35(6), 1541–7. https://doi.org/10.1007/s10067-016-3242-9 Levey, A. S., De Jong, P. E., Coresh, J., Nahas, M. E., Astor, B. C., Matsushita, K., ... & Eckardt, K. U. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney international, 80(1), 17-28. Levitt, D. G., & Levitt, M. D. (2016). Human serum albumin homeostasis: a new look at the roles of synthesis, catabolism, renal and gastrointestinal excretion, and the clinical value of serum albumin measurements. International Journal of General Medicine, 9, 229–55. https://doi.org/10.2147/IJGM.S102819 Lewandowski, K. (2015). Reference ranges for TSH and thyroid hormones. Thyroid Research, 8(Suppl 1), A17. https://doi.org/10.1186/1756-6614-8-S1-A17 Li, P., Wang, S.-S., Liu, H., Li, N., McNutt, M. A., Li, G., & Ding, H.-G. (2011). Elevated serum alpha fetoprotein levels promote pathological progression of hepatocellular carcinoma. World Journal of Gastroenterology, 17(41), 4563–71. https://doi.org/10.3748/wjg.v17.i41.4563 Lilleng, H., Abeler, K., Johnsen, S. H., Stensland, E., Løseth, S., Jorde, R., … Bekkelund, S. I. (2011). Variation of serum creatine kinase (CK) levels and prevalence of persistent hyperCKemia in a Norwegian normal population. The Tromsø Study. Neuromuscular Disorders, 21(7), 494–500. https://doi.org/10.1016/j.nmd.2011. 04.007 247 University of Ghana http://ugspace.ug.edu.gh Lim, E., Miyamura, J., & Chen, J. J. (2015). Racial/Ethnic-Specific Reference Intervals for Common Laboratory Tests: A Comparison among Asians, Blacks, Hispanics, and White. Hawai’i Journal of Medicine & Public Health : A Journal of Asia Pacific Medicine & Public Health, 74(9), 302–10. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/26468426 Lin, K. J., Pang, S. T., Chang, Y. H., Wu, C. T., Chuang, K. L., Chuang, H. C., & Chuang, C. K. (2010). Age-related reference levels of serum prostate-specific antigen among Taiwanese men without clinical evidence of prostate cancer. Chang Gung Med J, 33(2), 182-187. Lin, K.-J., Pang, S.-T., Chang, Y.-H., Wu, C.-T., Chuang, K.-L., Chuang, H.-C., & Chuang, C.-K. (2010). Age-related Reference Levels of Serum Prostate-specific Antigen among Taiwanese Men without Clinical Evidence of Prostate Cancer. Chang Gung Med J (Vol. 33). Retrieved from https://pdfs.semanticscholar.org/ 10cb/16353d963266134ddf3517b7318de06f908e.pdf Ling Hao, Jing Ma, Meir J. Stampfer, Aiguo Ren, Yihua Tian, Yi Tang, Walter C. Willett, Zhu Li, Geographical, Seasonal and Gender Differences in Folate Status among Chinese Adults, The Journal of Nutrition, Volume 133, Issue 11, November 2003, Pages 3630–3635, https://doi.org/10.1093/jn/133.11.3630 Ling, H., L., Ma, J., Stampfer, M. J., Ren, A., Tian, Y., Tang, Y., ... & Li, Z. (2003). Geographical, seasonal and gender differences in folate status among Chinese adults. The Journal of nutrition, 133(11), 3630-3635. Ling, Y., Jiang, J., Gui, M., Liu, L., Aleteng, Q., Wu, B., … Gao, X. (2015). Thyroid Function, Prevalent Coronary Heart Disease, and Severity of Coronary Atherosclerosis in Patients Undergoing Coronary Angiography. International Journal of Endocrinology, 2015, 1–9. https://doi.org/10.1155/2015/708272 Lippi, G., Lima-Oliveira, G., Salvagno, G. L., Montagnana, M., Gelati, M., Picheth, G., … Guidi, G. C. (2010). Influence of a light meal on routine haematological tests. Blood Transfusion, 8(2), 94–9. https://doi.org/10.2450/2009.0142-09 Lippi, G., Schena, F., Montagnana, M., Salvagno, G. L., & Guidi, G. C. (2008). Influence of acute physical exercise on emerging muscular biomarkers. Clinical Chemistry and Laboratory Medicine, 46(9), 1313–8. https://doi.org/10.1515/ CCLM.2008.250 248 University of Ghana http://ugspace.ug.edu.gh Liu, Z. Y., Sun, Y. H., Xu, C. L., Gao, X., Zhang, L. M., & Ren, S. C. (2009). Age-specific PSA reference ranges in Chinese men without prostate cancer. Asian journal of andrology, 11(1), 100. Liu, Z.-Y., Sun, Y.-H., Xu, C.-L., Gao, X., Zhang, L.-M., & Ren, S.-C. (2009). Age- specific PSA reference ranges in Chinese men without prostate cancer. Asian Journal of Andrology, 11(1), 100–3. https://doi.org/10.1038/aja.2008.17 Loeb, S., Carter, H. B., Catalona, W. J., Moul, J. W., & Schroder, F. H. (2012). Baseline prostate-specific antigen testing at a young age. European urology, 61(1), 1-7. López-Delgado, L., Riancho-Zarrabeitia, L., García-Unzueta, M. T., Tenorio, J. A., García-Hoyos, M., Lapunzina, P., … Riancho, J. A. (2018). Abnormal bone turnover in individuals with low serum alkaline phosphatase. Osteoporosis International, 29(9), 2147–2150. https://doi.org/10.1007/s00198-018-4571-0 Lott, J. (1992). Persistence of Increased Amylase and Lipase Concentrations in Acute- Pancreatitis-Response. Clinical Chemistry, 38(4), 609-609. Luboldt, H. J., Schindler, J. F., & Rübben, H. (2007). Age-specific reference ranges for prostate-specific antigen as a marker for prostate cancer. EAU-EBU update series, 5(1), 38-48. Luz, P. L. da, Favarato, D., Junior, J. R. F.-N., Lemos, P., & Chagas, A. C. P. (2008). High Ratio of Triglycerides to HDL-Cholesterol Predicts Extensive Coronary Disease. Clinics (Sao Paulo, Brazil), 63(4), 427. https://doi.org/10.1590/S1807- 59322008000400003 Males, S., Gad, R. R., Esmat, G., Abobakr, H., Anwar, M., Rekacewicz, C., … Fontanet, A. (2007). Serum alpha-foetoprotein level predicts treatment outcome in chronic hepatitis C. Antiviral Therapy, 12(5), 797–803. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/17713163 Malo, M. S. (2015). A High Level of Intestinal Alkaline Phosphatase Is Protective Against Type 2 Diabetes Mellitus Irrespective of Obesity. EBioMedicine, 2(12), 2016– 23. https://doi.org/10.1016/j.ebiom.2015.11.027 Malo, M. S., Alam, S. N., Mostafa, G., Zeller, S. J., Johnson, P. V., Mohammad, N., … Hodin, R. A. (2010). Intestinal alkaline phosphatase preserves the normal homeostasis of gut microbiota. Gut, 59(11), 1476–1484. https://doi.org/10.1136/ gut.2010.211706 Mancini, G. B. J., Tashakkor, A. Y., Baker, S., Bergeron, J., Fitchett, D., Frohlich, J., … Pope, J. (2013). Diagnosis, Prevention, and Management of Statin Adverse 249 University of Ghana http://ugspace.ug.edu.gh Effects and Intolerance: Canadian Working Group Consensus Update. Canadian Journal of Cardiology, 29(12), 1553–1568. https://doi.org/10.1016/j.cjca.2013. 09.023 Manolio, T. A., Burke, G. L., Savage, P. J., Jacobs, D. R., Sidney, S., Wagenknecht, L. E., ... & Tracy, R. P. (1992). Sex-and race-related differences in liver-associated serum chemistry tests in young adults in the CARDIA study. Clinical chemistry, 38(9), 1853-1859. Marchesini, G., Moscatiello, S., Di Domizio, S., & Forlani, G. (2008). Obesity-associated liver disease. The Journal of Clinical Endocrinology & Metabolism, 93(11_supplement_1), s74-s80. Marhoum, T. A., Abdrabo, A. A., & Lutfi, M. F. (2013). Effects of age and gender on serum lipid profile in over 55 years-old apparently healthy Sudanese individuals. Asian Journal of Biomedical and Pharmaceutical Sciences, 3(19), 10. Matsubara, A., Ichihara, K., & Fukutani, S. (2008). Determination of reference intervals for 26 commonly measured biochemical analytes with consideration of long- term within-individual variation. Clinical chemistry and laboratory medicine, 46(5), 691-698. Mauro P, Renze B, Wouter W. (2006). In: Tietz text book of clinical chemistry and molecular diagnostics. 4th edition. Carl AB, Edward R, David EB, editors. Elsevier;Enzymes; pp. 604–616. McCullough, A. J. (2002). Update on nonalcoholic fatty liver disease. Journal of clinical gastroenterology, 34(3), 255-262. McPherson, K., Healy, M. J. R., Flynn, F. V., Piper, K. A. J., & Garcia-Webb, P. (1978). The effect of age, sex and other factors on blood chemistry in health. Clinica Chimica Acta, 84(3), 373-397. Mekenkamp, L. J. M., Heesterbeek, K. J., Koopman, M., Tol, J., Teerenstra, S., Venderbosch, S., … Nagtegaal, I. D. (2012). Mucinous adenocarcinomas: Poor prognosis in metastatic colourectal cancer. European Journal of Cancer, 48(4), 501–509. https://doi.org/10.1016/j.ejca.2011.12.004 Mekonnen, Z., Amuamuta, A., Mulu, W., Yimer, M., Zenebe, Y., Adem, Y., … Gebregziabher, Y. (2017a). Clinical chemistry reference intervals of healthy adult populations in Gojjam Zones of Amhara National Regional State, Northwest Ethiopia. PLOS ONE, 12(9), e0184665. https://doi.org/10.1371/ 250 University of Ghana http://ugspace.ug.edu.gh journal. pone.0184665 Mekonnen, Z., Amuamuta, A., Mulu, W., Yimer, M., Zenebe, Y., Adem, Y., … Gebregziabher, Y. (2017b). Clinical chemistry reference intervals of healthy adult populations in Gojjam Zones of Amhara National Regional State, Northwest Ethiopia. PloS one, 12(9), e0184665. https://doi.org/10.1371 /journal.pone. Meng, Z., Liu, M., Zhang, Q., Liu, L., Song, K., Tan, J., … Zhang, J. (2015). Gender and Age Impacts on the Association Between Thyroid Function and Metabolic Syndrome in Chinese. Medicine, 94(50), e2193. https://doi.org/10.1097/MD.0000000000002193 Miller, W. G., Bruns, D. E., Hortin, G. L., Sandberg, S., Aakre, K. M., McQueen, M. J., … Narva, A. S. (2008). Current Issues in Measurement and Reporting of Urinary Albumin Excretion. Clinical Chemistry, 55(1), 24–38. https://doi.org/10.1373/clinchem.2008.106567 Miller, W. G., Seegmiller, J. C., Lieske, J. C., Narva, A. S., & Bachmann, L. M. (2017). Standardization of Urine Albumin Measurements: Status and Performance Goals. https://doi.org/10.1373/jalm.2017.023614 Miller, R., Stalder, T., Jarczok, M., Almeida, D. M., Badrick, E., Bartels, M., ... & Fischer, J. E. (2016). The CIRCORT database: Reference ranges and seasonal changes in diurnal salivary cortisol derived from a meta-dataset comprised of 15 field studies. Psychoneuroendocrinology, 73, 16-23. Miri-Dashe, T., Osawe, S., Tokdung, M., Daniel, N., Choji, R. P., Mamman, I., ... & Abimiku, A. L. (2014). Comprehensive reference ranges for hematology and clinical chemistry laboratory parameters derived from normal Nigerian adults. PLoS One, 9(5), e93919. Mitchell, A. J., Roediger, B., & Weninger, W. (2014). Monocyte homeostasis and the plasticity of inflammatory monocytes. Cellular Immunology, 291(1–2), 22–31. https://doi.org/10.1016/j.cellimm.2014.05.010 Moghadam-Kia, S., Oddis, C. V, & Aggarwal, R. (2016). Approach to asymptomatic creatine kinase elevation. Cleveland Clinic Journal of Medicine, 83(1), 37–42. https://doi.org/10.3949/ccjm.83a.14120 Mohd Nasir Mbbs, N., Sthaneshwar Mbbs, P., Junaidah MEGAT YUNUS, P., & Yap Mbbs, S.-F. (2010). Comparing measured total carbon dioxide and calculated bicarbonate. Malaysian J Pathol (Vol. 32). Retrieved from 251 University of Ghana http://ugspace.ug.edu.gh http://mjpath.org.my/2010.1/Active_carbon.pdf Molloy, J. W., Calcagno, C. J., Williams, C. D., Jones, F. J., Torres, D. M., & Harrison, S. A. (2012). Association of coffee and caffeine consumption with fatty liver disease, nonalcoholic steatohepatitis, and degree of hepatic fibrosis. Hepatology, 55(2), 429–436. https://doi.org/10.1002/hep.24731 Moro, L., Reineri, S., Piranda, D., Pietrapiana, D., Lova, P., Bertoni, A., ... & Sinigaglia, F. (2005). Nongenomic effects of 17β-estradiol in human platelets: potentiation of thrombin-induced aggregation through estrogen receptor β and Src kinase. Blood, 105(1), 115-121. Motzer, R. J., Escudier, B., Bukowski, R., Rini, B. I., Hutson, T. E., Barrios, C. H., … Gore, M. E. (2013). Prognostic factors for survival in 1059 patients treated with sunitinib for metastatic renal cell carcinoma. British Journal of Cancer, 108(12), 2470–2477. https://doi.org/10.1038/bjc.2013.236 Moyer, V. A., & U.S. Preventive Services Task Force. (2012). Screening for Prostate Cancer: U.S. Preventive Services Task Force Recommendation Statement. Annals of Internal Medicine, 157(2), 120. https://doi.org/10.7326/0003-4819- 157-2-201207170-00459 Moynihan, R., Glassock, R., & Doust, J. (2013). Chronic kidney disease controversy: how expanding definitions are unnecessarily labelling many people as diseased. Bmj, 347, f4298. Mulu, W., Abera, B., Mekonnen, Z., Adem, Y., Yimer, M., Zenebe, Y., … Gebeyehu, W. (2017). Haematological and CD4 + T cells reference ranges in healthy adult populations in Gojjam zones in Amhara region, Ethiopia. https://doi.org/10.1371/journal.pone.0181268 Murray, R., Rodwell, V. , Bender, D. , Botham, K. M. , Weil, P., Anthony, , & Kennelly, P. J. (2009). Harper’s Illustrated Biochemistry 28/e. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.477.1890&rep=rep1 &type=pdf Musa, H. H., Tyrab, E. M. A., Hamid, M. M. A., Elbashir, E. A. R., Yahia, L. M., & Salih, N. M. (2013). Characterization of lipid profile in coronary heart disease patients in Sudan. Indian Heart Journal, 65(2), 232–233. https://doi.org/10.1016/j.ihj. 2013.03.007 252 University of Ghana http://ugspace.ug.edu.gh Musso, C. G., & Oreopoulos, D. G. (2011). Aging and physiological changes of the kidneys including changes in glomerular filtration rate. Nephron Physiology, 119(Suppl. 1), p1-p5. Na, R., Wu, Y., Xu, J., Jiang, H., & Ding, Q. (2013). Age-specific prostate specific antigen cutoffs for guiding biopsy decision in Chinese population. PloS one, 8(6), e67585. Nagai, K., Tsuchida, K., Hirose, D., Michiwaki, H., Hann, M., Kanayama, H., … Minakuchi, J. (2016). The effect of albumin leakage in hemodialysis patients on redox status of serum albumin. Journal of Artificial Organs, 19(3), 310–314. https://doi.org/10.1007/s10047-016-0900-2 Nakajima, K. (2016). Low serum amylase and obesity, diabetes and metabolic syndrome: A novel interpretation. World journal of diabetes, 7(6), 112. National Center for Health Statistics. (2015). Laboratory Procedure Manual Alkaline Phosphatase (ALP) Refrigerated Serum Beckman UniCel ® DxC 800 Synchron & Beckman UniCel ® DxC 660i Synchron Access Clinical Systems (Identical Method). Retrieved from https://wwwn.cdc.gov/ nchs/data/nhanes/2015- 2016/labmethods/BIOPRO_I_MET_ALP_DXC800and660i.pdf Neal, R. C., Ferdinand, K. C., Yčas, J., & Miller, E. (2009). Relationship of ethnic origin, gender, and age to blood creatine kinase levels. The American journal of medicine, 122(1), 73-78. Need, A. G., O’Loughlin, P. D., Morris, H. A., Horowitz, M., & Nordin, B. C. (2004). The effects of age and other variables on serum parathyroid hormone in postmenopausal women attending an osteoporosis center. The Journal of Clinical Endocrinology & Metabolism, 89(4), 1646-1649. Ngowi, B. J., Mfinanga, S. G., Bruun, J. N., & Morkve, O. (2009). Immunohaematological reference values in human immunodeficiency virus-negative adolescent and adults in rural northern Tanzania. BMC Infectious Diseases, 9(1), 1. https://doi.org/10.1186/1471-2334-9-1 Nicholson, J. K., & Wilson, I. D. (2003). Opinion: Understanding “Global” Systems Biology: Metabonomics and the Continuum of Metabolism. Nature Reviews Drug Discovery, 2(8), 668–676. https://doi.org/10.1038/nrd1157 Noori, N., Kopple, J. D., Kovesdy, C. P., Feroze, U., Sim, J. J., Murali, S. B., … Kalantar- Zadeh, K. (2010). Mid-arm muscle circumference and quality of life and survival 253 University of Ghana http://ugspace.ug.edu.gh in maintenance hemodialysis patients. Clinical Journal of the American Society of Nephrology : CJASN, 5(12), 2258–68. https://doi.org/10.2215/CJN.02080310 Noori, N., Kovesdy, C. P., Bross, R., Lee, M., Oreopoulos, A., Benner, D., … Kalantar- Zadeh, K. (2011). Novel equations to estimate lean body mass in maintenance hemodialysis patients. American Journal of Kidney Diseases: The Official Journal of the National Kidney Foundation, 57(1), 130–9. https://doi.org/10.1053/j.ajkd.2010.10.003 Nordin, G., Mårtensson, A., Swolin, B., Sandberg, S., Christensen, N. J., Thorsteinsson, V., … Savolainen, E. ‐R. (2004). A multicentre study of reference intervals for haemoglobin, basic blood cell counts and erythrocyte indices in the adult population of the Nordic countries. Scandinavian Journal of Clinical and Laboratory Investigation, 64(4), 385–398. https://doi.org/10.1080/ 00365510410002797 Norris, K. C., Smoyer, K. E., Rolland, C., Van der Vaart, J., & Grubb, E. B. (2018). Albuminuria, serum creatinine, and estimated glomerular filtration rate as predictors of cardio-renal outcomes in patients with type 2 diabetes mellitus and kidney disease: a systematic literature review. BMC Nephrology, 19(1), 36. https://doi.org/10.1186/s12882-018-0821-9 Norvik, J. V., Storhaug, H. M., Ytrehus, K., Jenssen, T. G., Zykova, S. N., Eriksen, B. O., & Solbu, M. D. (2016). Overweight modifies the longitudinal association between uric acid and some components of the metabolic syndrome: The Tromsø Study. BMC cardiovascular disorders, 16(1), 85. Odhiambo, C., Oyaro, B., Odipo, R., Otieno, F., Alemnji, G., Williamson, J., & Zeh, C. (2015a). Evaluation of locally established reference intervals for hematology and biochemistry parameters in Western Kenya. PloS One, 10(4), e0123140. https://doi.org/10.1371/journal.pone.0123140 Odhiambo, C., Oyaro, B., Odipo, R., Otieno, F., Alemnji, G., Williamson, J., & Zeh, C. (2015b). Evaluation of Locally Established Reference Intervals for Hematology and Biochemistry Parameters in Western Kenya. PLOS ONE, 10(4), e0123140. https://doi.org/10.1371/journal.pone.0123140 Ojima, T., Iwahashi, M., Nakamura, M., Matsuda, K., Nakamori, M., Ueda, K., … Yamaue, H. (2007). Successful cancer vaccine therapy for carcinoembryonic antigen (CEA)-expressing colon cancer using genetically modified dendritic cells that express CEA and T helper-type 1 cytokines in CEA transgenic mice. 254 University of Ghana http://ugspace.ug.edu.gh International Journal of Cancer, 120(3), 585–593. https://doi.org/10.1002/ ijc.22298 Omuse, G., Maina, D., Mwangi, J., Wambua, C., Radia, K., Kanyua, A., ... & Ichihara, K. (2018). Complete blood count reference intervals from a healthy adult urban population in Kenya. PloS one, 13(6), e0198444. Osei‐Bimpong, A., McLean, R., Bhonda, E., & Lewis, S. M. (2012). The use of the white cell count and haemoglobin in combination as an effective screen to predict the normality of the full blood count. International journal of laboratory hematology, 34(1), 91-97. Osuna C, J. A., Gomez-Perez, R., Arata-Bellabarba, G., & Villaroel, V. (2006). Relationship between BMI, total testosterone, sex hormone-binding-globulin, leptin, insulin and insulin resistance in obese men. Archives of andrology, 52(5), 355-361. Ozarda, Y. (2016). Reference intervals: current status, recent developments and future considerations. Biochemia Medica, 5–16. https://doi.org/10.11613 /BM.2016.001 Ozarda, Y., Ichihara, K., Aslan, D., Aybek, H., Ari, Z., Taneli, F., … Degirmen, E. (2014). A multicenter nationwide reference intervals study for common biochemical analytes in Turkey using Abbott analyzers. Clinical Chemistry and Laboratory Medicine (CCLM), 52(12), 1823–1833. https://doi.org/10.1515 /cclm-2014-0228 Ozarda, Y., Ichihara, K., Bakan, E., Polat, H., Ozturk, N., Baygutalp, N. K., … Eker, P. (2017). A nationwide multicentre study in Turkey for establishing reference intervals of haematological parameters with novel use of a panel of whole blood. Biochemia Medica, 27(2), 350–377. https://doi.org/10.11613/ BM.2017.038 Ozarda, Y., Ichihara, K., Barth, J. H., & Klee, G. (2013). Protocol and standard operating procedures for common use in a worldwide multicenter study on reference values. Clinical Chemistry and Laboratory Medicine, 51(5), 1027–1040. https://doi.org/10.1515/cclm-2013-0249 Park, S. H., Park, C. J., Lee, B. R., Kim, M. J., Han, M. Y., Cho, Y. U., & Jang, S. (2016). Establishment of Age- and Gender-Specific Reference Ranges for 36 Routine and 57 Cell Population Data Items in a New Automated Blood Cell Analyzer, Sysmex XN-2000. Annals of Laboratory Medicine, 36(3), 244–9. https://doi.org/10.3343/alm.2016.36.3.244 255 University of Ghana http://ugspace.ug.edu.gh Park, Y., Kim, Y., Lee, E. Y., Lee, J. H., & Kim, H. S. (2012). Reference ranges for HE4 and CA125 in a large Asian population by automated assays and diagnostic performances for ovarian cancer. International journal of cancer, 130(5), 1136- 1144. Park, Y., Lee, J. H., Hong, D. J., Lee, E. Y., & Kim, H. S. (2011). Diagnostic performances of HE4 and CA125 for the detection of ovarian cancer from patients with various gynecologic and non-gynecologic diseases. Clinical biochemistry, 44(10-11), 884-888. Pastuszak, A. W., Rodriguez, K. M., Nguyen, T. M., & Khera, M. (2016). Translational andrology and urology. Translational Andrology and Urology (Vol. 5). Retrieved from http://tau.amegroups.com/article/view/11559/13173 Patel, S. S., Molnar, M. Z., Tayek, J. A., Ix, J. H., Noori, N., Benner, D., … Kalantar- Zadeh, K. (2013). Serum creatinine as a marker of muscle mass in chronic kidney disease: results of a cross-sectional study and review of literature. Journal of Cachexia, Sarcopenia and Muscle, 4(1), 19–29. https://doi.org/10.1007/s13539- 012-0079-1 Patel, S., Hyer, S., & Barron, J. (2007). Glomerular filtration rate is a major determinant of the relationship between 25-hydroxyvitamin D and parathyroid hormone. Calcified tissue international, 80(4), 221-226. Pawlak, R., & Rusher, D. R. (2013). A Review of 89 Published Case Studies of Vitamin B12 Deficiency. J Hum Nutr Food Sci, 1(2), 1008. Retrieved from https://www.jscimedcentral.com/Nutrition/Articles/nutrition-1-1008.pdf Perlstein, T. S., Pande, R. L., Creager, M. A., Weuve, J., & Beckman, J. A. (2008). Serum total bilirubin level, prevalent stroke, and stroke outcomes: National Health and Nutrition Examination Survey. Am J Med, 121(9), 781–788. https://doi.org/10.1016/j.amjmed.2008.03.045 Perrone, R. D., Madias, N. E., & Levey, A. S. (1992). Serum creatinine as an index of renal function: new insights into old concepts. Clinical Chemistry, 38(10), 1933– 53. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1394976 Peters, E., Geraci, S., Heemskerk, S., Wilmer, M. J., Bilos, A., Kraenzlin, B., … Masereeuw, R. (2015). Alkaline phosphatase protects against renal inflammation through dephosphorylation of lipopolysaccharide and adenosine triphosphate. British Journal of Pharmacology, 172(20), 4932–4945. https://doi.org/10.1111/ bph.13261 256 University of Ghana http://ugspace.ug.edu.gh PetitClerc, C., & Solberg, H. E. (1987). International Federation of Clinical Chemistry (IFCC), scientific committee, clinical section, expert panel on theory of reference values, and International Committee for Standardization in Haematology (ICSH), standing committee on reference values. Approved recommendation (1987) on the theory of reference values. Part 2. Selection of individuals for the production of reference values. J Clin Chem Clin Biochem, 25(9), 639-44. Phillips, A., Gerald Shaper, A., & Whincup, P. (1989). Association Between Serum Albumin and Mortality from Cardiovascular Disease, Cancer, And Other Causes. The Lancet, 334(8677), 1434–1436. https://doi.org/10.1016/S0140- 6736(89)92042-4 Pinheiro, S. P., Holmes, M. D., Pollak, M. N., Barbieri, R. L., & Hankinson, S. E. (2005). Racial differences in premenopausal endogenous hormones. Cancer Epidemiology and Prevention Biomarkers, 14(9), 2147-2153. Plant, T. M., Tony M., Zeleznik, A., & Knobil, E. (2014). Knobil and Neill’s physiology of reproduction. Retrieved from https://books.google.com. gh/books?hl=en&lr=&id=I1ACBAAAQBAJ&oi=fnd&pg=PP1&dq=Plant+TM ,+Zeleznik+AJ+2014+Knobil+and+Neill%27s+Physiology+of+Reproduction:+ Two- Volume+Set.+Elsevier+Science&ots=aUN5Y6spCI&sig=7EkIOfRuJ5EHnwHI _qy0P1efhiQ&redir_esc=y#v=onepage& Plebani, M. (2010). The detection and prevention of errors in laboratory medicine. Annals of Clinical Biochemistry, 47(2), 101–110. https://doi.org/10.1258/acb.2009.009222 Poggio, E. D., Rule, A. D., Tanchanco, R., Arrigain, S., Butler, R. S., Srinivas, T., … Schreiber, M. J. (2009). Demographic and clinical characteristics associated with glomerular filtration rates in living kidney donors. Kidney International, 75(10), 1079–1087. https://doi.org/10.1038/ki.2009.11 Ponka, P. (1997). Tissue-specific regulation of iron metabolism and heme synthesis: distinct control mechanisms in erythroid cells. Blood, 89(1), 1–25. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8978272 Pottel, H., Delanaye, P., Weekers, L., Selistre, L., Goffin, K., Gheysens, O., & Dubourg, L. (2017). Age-dependent reference intervals for estimated and measured glomerular filtration rate. Clinical kidney journal, 10(4), 545-551. 257 University of Ghana http://ugspace.ug.edu.gh Pottel, H., Hoste, L., Dubourg, L., Ebert, N., Schaeffner, E., Eriksen, B. O., ... & Glassock, R. J. (2016). An estimated glomerular filtration rate equation for the full age spectrum. Nephrology Dialysis Transplantation, 31(5), 798-806. Poupon, R. (2015). Liver alkaline phosphatase: A missing link between choleresis and biliary inflammation. Hepatology, 61(6), 2080–2090. https://doi.org/10.1002/hep.27715 Powles, T., Bascoul-Mollevi, C., Kramar, A., Lorch, A., & Beyer, J. (2013). Prognostic impact of LDH levels in patients with relapsed/refractory seminoma. Journal of Cancer Research and Clinical Oncology, 139(8), 1311–1316. https://doi.org/10.1007/s00432-013-1442-0 Prince, L. K., Abbott, K. C., Lee, J. J., Oliver, D. K., & Olson, S. W. (2015). Creatine Kinase, Coenzyme Q10, Race, and Risk of Rhabdomyolysis. American Journal of Kidney Diseases, 66(3), 541–542. https://doi.org/10.1053/j.ajkd.2015.04.045 Qiao, R., Yang, S., Yao, B., Wang, H., Zhang, J., & Shang, H. (2014). Complete blood count reference intervals and age- and sex-related trends of North China Han population. Clinical Chemistry and Laboratory Medicine (CCLM), 52(7), 1025– 32. https://doi.org/10.1515/cclm-2012-0486 Qin, X., Lin, L., Mo, Z., Lv, H., Gao, Y., Tan, A., ... & Zhao, J. (2011). Reference intervals for serum alpha-fetoprotein and carcinoembryonic antigen in Chinese Han ethnic males from the Fangchenggang Area Male Health and Examination Survey. The International journal of biological markers, 26(1), 65-71. Qin, X., Tang, G., Qiu, L., Li, P. C., Xia, L., Chen, M., … Ichihara, K. (2015). A Multicenter Reference Intervals Study for Specific Proteins in China. Medicine, 94(49), e2211. https://doi.org/10.1097/MD.0000000000002211 Qiu, L., Wang, D. C., Xu, T., Cheng, X. Q., Sun, Q., Hu, Y. Y., … Wang, Z. J. (2018). [Influence of gender, age and season on thyroid hormone reference interval]. Zhonghua Yi Xue Za Zhi, 98(20), 1582–1587. https://doi.org/10.3760/CMA.J.ISSN.0376-2491.2018.20.011 Ramasamy, S., Nguyen, D. D., Eston, M. A., Nasrin Alam, S., Moss, A. K., Ebrahimi, F., … Hodin, R. A. (2011). Intestinal alkaline phosphatase has beneficial effects in mouse models of chronic colitis. Inflammatory Bowel Diseases, 17(2), 532–542. https://doi.org/10.1002/ibd.21377 Reame, N. E., Kelche, R. P., Beitins, I. Z., Yu, M. Y., Zawacki, C. M., & Padmanabhan, V. A. S. A. N. T. H. A. (1996). Age effects of follicle-stimulating hormone and 258 University of Ghana http://ugspace.ug.edu.gh pulsatile luteinizing hormone secretion across the menstrual cycle of premenopausal women. The Journal of Clinical Endocrinology & Metabolism, 81(4), 1512-1518. Reaven, G. M. (2008). Insulin resistance: the link between obesity and cardiovascular disease. Endocrinology and metabolism clinics of North America, 37(3), 581- 601. Riancho-Zarrabeitia, L., García-Unzueta, M., Tenorio, J. A., Gómez-Gerique, J. A., Ruiz Pérez, V. L., Heath, K. E., … Riancho, J. A. (2016). Clinical, biochemical and genetic spectrum of low alkaline phosphatase levels in adults. European Journal of Internal Medicine, 29, 40–45. https://doi.org/10.1016/J. EJIM.2015.12.019 Richter, B., Sulzgruber, P., Koller, L., Steininger, M., El-Hamid, F., Rothgerber, D. J., … Niessner, A. (2018). Blood urea nitrogen has additive value beyond estimated glomerular filtration rate for prediction of long-term mortality in patients with acute myocardial infarction. European Journal of Internal Medicine. https://doi.org/10.1016/J.EJIM.2018.07.019 Robyn, C., & Vekemans, M. (1976). Influence of low dose oestrogen on circulating prolactin, LH and FSH levels in post-menopausal women. European Journal of Endocrinology, 83(1), 9-14. Roelfsema, F., Pijl, H., Keenan, D. M., & Veldhuis, J. D. (2012). Prolactin secretion in healthy adults is determined by gender, age and body mass index. PloS one, 7(2), e31305. Rohrmann, S., Nelson, W. G., Rifai, N., Brown, T. R., Dobs, A., Kanarek, N., ... & Platz, E. A. (2007). Serum estrogen, but not testosterone, levels differ between black and white men in a nationally representative sample of Americans. The Journal of Clinical Endocrinology & Metabolism, 92(7), 2519-2525. Roizen, J. D., Shah, V., Levine, M. A., & Carlow, D. C. (2013). Determination of reference intervals for serum total calcium in the vitamin D-replete pediatric population. The Journal of Clinical Endocrinology and Metabolism, 98(12), E1946-50. https://doi.org/10.1210/jc.2013-3105 Romeo, J., Wärnberg, J., Gómez-Martínez, S., Díaz, L. E., Moreno, L. A., Castillo, M. J., … Marcos, A. (2009). Haematological reference values in Spanish adolescents: the AVENA study. European Journal of Haematology, 83(6), 586–594. https://doi.org/10.1111/j.1600-0609.2009.01326.x 259 University of Ghana http://ugspace.ug.edu.gh Ruhl, C. E., & Everhart, J. E. (2003). Determinants of the association of overweight with elevated serum alanine aminotransferase activity in the United States. Gastroenterology, 124(1), 71-79. Ruhl, C. E., & Everhart, J. E. (2005). Coffee and caffeine consumption reduce the risk of elevated serum alanine aminotransferase activity in the United States. Gastroenterology, 128(1), 24–32. Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/15633120 Rule, A. D., Amer, H., Cornell, L. D., Taler, S. J., Cosio, F. G., Kremers, W. K., … Stegall, M. D. (2010). The Association Between Age and Nephrosclerosis on Renal Biopsy Among Healthy Adults. Annals of Internal Medicine, 152(9), 561. https://doi.org/10.7326/0003-4819-152-9-201005040-00006 Rustad, P., Felding, P., Franzson, L., Kairisto, V., Lahti, A., Mårtensson, A., … Uldall, A. (2004). The Nordic Reference Interval Project 2000: recommended reference intervals for 25 common biochemical properties. Scandinavian Journal of Clinical and Laboratory Investigation, 64(4), 271–284. https://doi.org/10. 1080/00365510410006324 Ruzagira, E., Abaasa, A., Mulenga, K. E., & Kilembe, J. (2014). Effect of Seasonal Variation on Adult Clinical Laboratory Parameters in Rwanda, Zambia, and Uganda: Implications for HIV Biomedical Prevention Trials. PLoS ONE, 9(8), 105089. https://doi.org/10.1371/journal.pone.0105089 Saab, S., Mallam, D., Cox, G. A., & Tong, M. J. (2014). Impact of coffee on liver diseases: A systematic review. Liver International, 34(4), 495–504. https://doi.org/ 10.1111/ liv.12304 Saathoff, E., Schneider, P., Kleinfeldt, V., Geis, S., Haule, D., Maboko, L., ... & Hoelscher, M. (2008). Laboratory reference values for healthy adults from southern Tanzania. Tropical Medicine & International Health, 13(5), 612-625. Sairam, S., Domalapalli, S., Muthu, S., Swaminathan, J., Ramesh, V. A., Sekhar, L., … Balasubramaniam, U. (2014). Haematological and Biochemical Parameters in Apparently Healthy Indian Population: Defining Reference Intervals. Indian Journal of Clinical Biochemistry, 29(3), 290–297. https://doi.org/10.1007/ s12291-013-0365-5 Sakoh, T., Nakayama, M., Tanaka, S., Yoshitomi, R., Ura, Y., Nishimoto, H., … Kitazono, T. (2015). Association of serum total bilirubin with renal outcome in Japanese patients with stages 3–5 chronic kidney disease. Metabolism, 64(9), 1096–1102. 260 University of Ghana http://ugspace.ug.edu.gh https://doi.org/10.1016/J.METABOL.2015.06.006 Samaneka, W. P., Mandozana, G., Tinago, W., Nhando, N., Mgodi, N. M., Bwakura- Dangarembizi, M. F., … Hakim, J. G. (2016). Adult Hematology and Clinical Chemistry Laboratory Reference Ranges in a Zimbabwean Population. PLOS ONE, 11(11), e0165821. https://doi.org/10.1371/journal.pone.0165821 Sanchis-Gomar, F., & Lippi, G. (2014). Physical activity - an important preanalytical variable. Biochemia Medica, 24(1), 68–79. https://doi.org/10.11613/ BM.2014.009 Sanchis-Gomar, F., Banfi, G., Pareja-Galeano, H., Martinez-Bello, V., & Lippi, G. (2013). Anaemia, heart failure and exercise training. International Journal of Cardiology, 165(3), 587–588. https://doi.org/10.1016/j.ijcard.2012.09.028 Schaeffner, E. (2017). Determining the Glomerular Filtration Rate-An Overview. https://doi.org/10.1053/j.jrn.2017.07.005 Scheepers, L. E. J. M., Boonen, A., Dagnelie, P. C., Schram, M. T., van der Kallen, C. J. H., Henry, R. M. A., … Arts, I. C. W. (2017). Uric acid and blood pressure. Journal of Hypertension, 35(10), 1968–1975. https://doi.org/10.1097/HJH. 0000000000001417 Scheinfeldt, L. B., Soi, S., Thompson, S., Ranciaro, A., Woldemeskel, D., Beggs, W., ... & Tishkoff, S. A. (2012). Genetic adaptation to high altitude in the Ethiopian highlands. Genome biology, 13(1), R1. Schumann, G., Bonora, R., Ceriotti, F., Férard, G., Ferrero, C. A., Franck, P. F., ... & Kessner, A. (2002a). IFCC primary reference procedures for the measurement of catalytic activity concentrations of enzymes at 37 C. Part 4. Reference procedure for the measurement of catalytic concentration of alanine aminotransferase. Clinical Chemistry and Laboratory Medicine, 40(7), 718-724. Schumann, G., Bonora, R., Ceriotti, F., Férard, G., Ferrero, C. A., Franck, P. F., ... & Kessner, A. (2002b). IFCC primary reference procedures for the measurement of catalytic activity concentrations of enzymes at 37°C. Part 5. Reference procedure for the measurement of catalytic activity concentration of aspartate- aminotransferase [L-aspartate: 2-oxoglutarate-aminotransferase (AST). Clinical Chemistry and Laboratory Medicine, 40:725–33 Schüring, A. N., Kelsch, R., Pierściński, G., & Nofer, J. R. (2016). Establishing reference intervals for sex hormones on the analytical platforms Advia Centaur and Immulite 2000XP. Annals of laboratory medicine, 36(1), 55-59. 261 University of Ghana http://ugspace.ug.edu.gh Schwettmann, L., & Berbu, S. (2015). Reference Interval and Status for Serum Folate and Serum Vitamin B12 in a Norwegian Population. Clinical laboratory, 61(8), 1095-1100. Segolodi, T. M., Henderson, F. L., Rose, C. E., Turner, K. T., Zeh, C., Fonjungo, P. N., … Paxton, L. A. (2014). Normal laboratory reference intervals among healthy adults screened for a HIV pre-exposure prophylaxis clinical trial in Botswana. PloS One, 9(4), e93034. https://doi.org/10.1371/journal.pone.0093034 Selhub, J., Morris, M. S., Jacques, P. F., & Rosenberg, I. H. (2009). Folate–vitamin B-12 interaction in relation to cognitive impairment, anaemia, and biochemical indicators of vitamin B-12 deficiency. The American Journal of Clinical Nutrition, 89(2), 702S–706S. https://doi.org/10.3945/ajcn.2008.26947C Selistre, L., De Souza, V., Cochat, P., Antonello, I. C. F., Hadj-Aissa, A., Ranchin, B., … Dubourg, L. (2012). GFR estimation in adolescents and young adults. Journal of the American Society of Nephrology : JASN, 23(6), 989–96. https://doi.org/ 10.1681/ASN.2011070705 Selvakumar, K., Bavithra, S., Suganya, S., Ahmad Bhat, F., Krishnamoorthy, G., & Arunakaran, J. (2013). Effect of Quercetin on Haematobiochemical and Histological Changes in the Liver of Polychlorined Biphenyls-Induced Adult Male Wistar Rats. Journal of Biomarkers, 2013, 1–12. https://doi.org/ 10.1155/2013/960125 Shahani, S., Braga-Basaria, M., Maggio, M., & Basaria, S. (2009). Androgens and erythropoiesis: past and present. Journal of endocrinological investigation, 32(8), 704-716. Shamai, L., Lurix, E., Shen, M., Novaro, G. M., Szomstein, S., Rosenthal, R., ... & Asher, C. R. (2011). Association of body mass index and lipid profiles: evaluation of a broad spectrum of body mass index patients including the morbidly obese. Obesity surgery, 21(1), 42-47. Shamim, M. O., Khan, F. M. A., & Arshad, R. (2015). Association between serum total testosterone and Body Mass Index in middle aged healthy men. Pakistan journal of medical sciences, 31(2), 355. Sharma, U., Pal, D., & Prasad, R. (2014). Alkaline Phosphatase: An Overview. Indian Journal of Clinical Biochemistry, 29(3), 269–278. https://doi.org/10.1007/ s12291-013-0408-y 262 University of Ghana http://ugspace.ug.edu.gh Shcherbinina, M. B. (2007). Low blood bilirubin level: possible diagnostic and prognostic importance. Klinicheskaia meditsina, 85(10), 10-14. Sheedfar, F., Biase, S. D., Koonen, D., & Vinciguerra, M. (2013). Liver diseases and aging: friends or foes?. Aging cell, 12(6), 950-954. Sheehan, M. T. (2016). Biochemical Testing of the Thyroid: TSH is the Best and, Oftentimes, Only Test Needed - A Review for Primary Care. Clinical Medicine & Research, 14(2), 83–92. https://doi.org/10.3121/cmr.2016.1309 Sherpa, L. Y., Deji, Stigum, H., Chongsuvivatwong, V., Luobu, O., Thelle, D. S., … Bjertness, E. (2011). Lipid Profile and Its Association with Risk Factors for Coronary Heart Disease in the Highlanders of Lhasa, Tibet. High Altitude Medicine & Biology, 12(1), 57–63. https://doi.org/10.1089/ham.2010.1050 Shimizu, Y., Ichihara, K., & Kouguchi, K. (2017). Time required for resetting postural effects on serum constituents in healthy individuals. Clinica Chimica Acta, 472, 131-135. Shultz, E., Willard, K., Rich, S., Chemistry, D. C.-…, & 1985, U. (1985). Improved reference-interval estimation. Clinical Chemistry, 31(12), 1974–8. Retrieved from http://clinchem.aaccjnls.org/content/31/12/1974.short Siest, G., Henny, J., Gräsbeck, R., Wilding, P., Petitclerc, C., Queraltó, J. M., & Petersen, P. H. (2013). The theory of reference values: an unfinished symphony. Clinical Chemistry and Laboratory Medicine, 51(1), 47–64. https://doi.org/10.1515/cclm-2012-0682 Siraj, N., Issac, J., Anwar, M., Mehari, Y., Russom, S., Kahsay, S., & Frezghi, H. (2018). Establishment of haematological reference intervals for healthy adults in Asmara. BMC Research Notes, 11, 55. https://doi.org/10.1186/s13104-018- 3142-y Soares, A. A., Prates, A. B., Weinert, L. S., Veronese, F. V., de Azevedo, M. J., & Silveiro, S. P. (2013). Reference values for glomerular filtration rate in healthy Brazilian adults. BMC Nephrology, 14, 54. https://doi.org/10.1186/1471-2369-14-54 Sofer, Y., Osher, E., Limor, R., Shefer, G., Marcus, Y., Shapira, I., ... & Stern, N. (2016). Gender determines serum free cortisol: higher levels in men. Endocrine Practice, 22(12), 1415-1421. Sohn, W., Jun, D. W., Kwak, M. J., Park, Q., Lee, K. N., Lee, H. L., ... & Choi, H. S. (2013). Upper limit of normal serum alanine and aspartate aminotransferase levels in Korea. Journal of gastroenterology and hepatology, 28(3), 522-529. 263 University of Ghana http://ugspace.ug.edu.gh Solberg, H. E. (1987). Approved recommendation (1986) on the theory of reference values. Part 1. The concept of reference values. Clinica Chimica Acta, 165(1), 111–118. https://doi.org/10.1016/0009-8981(87)90224-5 Solberg, H. E. (2006). Establishment and use of reference values. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. Philadelphia: Elsevier Inc, 425-448. Solberg, H. E., & Gräsbeck, R. (1989). Reference Values. Advances in Clinical Chemistry, 27, 1–79. https://doi.org/10.1016/S0065-2423(08)60181-X Sorbi, D., Boynton, J., & Lindor, K. D. (1999). The ratio of aspartate aminotransferase to alanine aminotransferase: potential value in differentiating nonalcoholic steatohepatitis from alcoholic liver disease. The American journal of gastroenterology, 94(4), 1018. Speer, O., Schmugge, M., Metzger, C., & Albisetti, M. (2013). Reference Ranges of Coagulation Tests (pp. 85–96). Humana Press, Totowa, NJ. https://doi.org/10.1007/978-1-62703-339-8_6 Spencer, C. A., Hollowell, J. G., Kazarosyan, M., & Braverman, L. E. (2007). National Health and Nutrition Examination Survey III Thyroid-Stimulating Hormone (TSH)-Thyroperoxidase Antibody Relationships Demonstrate That TSH Upper Reference Limits May Be Skewed by Occult Thyroid Dysfunction. The Journal of Clinical Endocrinology & Metabolism, 92(11), 4236–4240. https://doi.org/10.1210/jc.2007-0287 Storer, T. W., Woodhouse, L., Magliano, L., Singh, A. B., Dzekov, C., Dzekov, J., & Bhasin, S. (2008). Changes in Muscle Mass, Muscle Strength, and Power but Not Physical Function Are Related to Testosterone Dose in Healthy Older Men. Journal of the American Geriatrics Society, 56(11), 1991–1999. https://doi.org/10.1111/j.1532-5415.2008.01927.x Stranges, S., Dorn, J. M., Muti, P., Freudenheim, J. L., Farinaro, E., Russell, M., ... & Trevisan, M. (2004). Body fat distribution, relative weight, and liver enzyme levels: A population‐based study. Hepatology, 39(3), 754-763. Streppel, M. M., Vincent, A., Mukherjee, R., Campbell, N. R., Chen, S.-H., Konstantopoulos, K., … Montgomery, E. A. (2012). Mucin 16 (cancer antigen 125) expression in human tissues and cell lines and correlation with clinical outcome in adenocarcinomas of the pancreas, esophagus, stomach, and colon. Human Pathology, 43(10), 1755–1763. https://doi.org/10.1016/j. 264 University of Ghana http://ugspace.ug.edu.gh humpath.2012.01.005 Sun, X., Shan, Z., & Teng, W. (2014). Effects of Increased Iodine Intake on Thyroid Disorders. Endocrinology and Metabolism, 29(3), 240. https://doi.org/10.3803/ EnM.2014.29.3.240 Sundaram, M., Mohanakrishnan, J., Murugavel, K. G., Shankar, E. M., Solomon, S., Srinivas, C. N., … Balakrishnan, P. (2008a). Ethnic variation in certain haematological and biochemical reference intervals in a south Indian healthy adult population. European Journal of Internal Medicine, 19(1), 46–50. https://doi.org/10.1016/j.ejim.2007.06.010 Sundaram, M., Mohanakrishnan, J., Murugavel, K. G., Shankar, E. M., Solomon, S., Srinivas, C. N., … Balakrishnan, P. (2008b). Ethnic variation in certain haematological and biochemical reference intervals in a south Indian healthy adult population. European Journal of Internal Medicine, 19(1), 46–50. https://doi.org/10.1016/j.ejim.2007.06.010 Švagera, Z., & Šigutová, R. (2016). Pre-Analytical Effects on Laboratory Examinations. In D. Jaroslav Racek & P. D. Daniel Rajdl (Eds.), Clinical Biochemistry (First, pp. 1–419). Prague: Karolinum Press. Retrieved from www.karolinum.cz Tamechika, Y., Iwatani, Y., Tohyama, K., & Ichihara, K. (2006). Insufficient filling of vacuum tubes as a cause of microhemolysis and elevated serum lactate dehydrogenase levels. Use of a data-mining technique in evaluation of questionable laboratory test results. Clinical Chemistry and Laboratory Medicine (CCLM), 44(5), 657-661. Taniguchi, H., & Honnda, Y. (2009). Amylases. Encyclopedia of Microbiology, 159–173. https://doi.org/10.1016/B978-012373944-5.00130-9 Targher, G., Zoppini, G., Cesare Guidi, G., & Lippi, G. (2009). Relationship between serum bilirubin and kidney function in non-diabetic and diabetic individuals. Kidney International, 75(8), 863. https://doi.org/10.1038/KI.2008.677 Taylor, E. H., Fink, L. M., & Pappas, A. A. (1989). Reproducibility of creatine kinase isoenzyme electrophoresis. Clinical chemistry, 35(4), 710-710. Tembe, N., Joaquim, O., Alfai, E., Sitoe, N., Viegas, E., Macovela, E., … Nilsson, C. (2014). Reference Values for Clinical Laboratory Parameters in Young Adults in Maputo, Mozambique. PLoS ONE, 9(5), e97391. https://doi.org/10.1371 /journal.pone.0097391 265 University of Ghana http://ugspace.ug.edu.gh Thapa, B. R., & Walia, A. (2007). Liver function tests and their interpretation. Indian Journal of Pediatrics, 74(7), 663–71. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/17699976 The Emerging Risk Factors Collaboration. (2010). Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. The Lancet, 375(9733), 2215–2222. https://doi.org/10.1016/S0140-6736(10)60484-9 Thiele, J. R., Zeller, J., Bannasch, H., Stark, G. B., Peter, K., & Eisenhardt, S. U. (2015). Targeting C-Reactive Protein in Inflammatory Disease by Preventing Conformational Changes. Mediators of Inflammation, 2015, 372432. https://doi.org/10.1155/2015/372432 Throop, J. L., Kerl, M. E., & Cohn, L. A. (2004). Albumin in Health and Disease: Protein Metabolism and Function*. Retrieved from www.VetLearn.com Tian, C. R., Qian, L., Shen, X. Z., Li, J. J., & Wen, J. T. (2014). Distribution of Serum Total Protein in Elderly Chinese. PloS one, 9(6), e101242. Tirali, R. E., Yalçınkaya Erdemci, Z., & Çehreli, S. B. (2013). Oral findings and clinical implications of patients with congenital neutropenia: a literature review. The Turkish Journal of Pediatrics, 55(3), 241–5. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/24217068 Torner, L. (2016). Actions of Prolactin in the Brain: From Physiological Adaptations to Stress and Neurogenesis to Psychopathology. Frontiers in Endocrinology, 7, 25. https://doi.org/10.3389/fendo.2016.00025 Travison, T. G., Vesper, H. W., Orwoll, E., Wu, F., Kaufman, J. M., Wang, Y., ... & Bhasin, S. (2017). Harmonized reference ranges for circulating testosterone levels in men of four cohort studies in the United States and Europe. The Journal of Clinical Endocrinology & Metabolism, 102(4), 1161-1173. Tsai, W.-N., Wang, Y.-Y., Liang, J.-T., Lin, S.-Y., Sheu, W. H.-H., & Chang, W.-D. (2015). Serum total bilirubin concentrations are inversely associated with total white blood cell counts in an adult population. Annals of Clinical Biochemistry, 52(2), 251–258. https://doi.org/10.1177/0004563214541969 Tugume, S. B., Piwowar, E. M., Lutalo, T., Mugyenyi, P. N., Grant, R. M., Mangeni, F. W., … Katongole-Mbidde, E. (1995). Haematological reference ranges among healthy Ugandans. Clinical and Vaccine Immunology, 2(2). 266 University of Ghana http://ugspace.ug.edu.gh Tuin, A., Poelstra, K., de Jager-Krikken, A., Bok, L., Raaben, W., Velders, M. P., & Dijkstra, G. (2009). Role of alkaline phosphatase in colitis in man and rats. Gut, 58(3), 379–387. https://doi.org/10.1136/gut.2007.128868 Upmeier, E., Lavonius, S., Heinonen, P., Viitanen, M., Isoaho, H., Arve, S., & Lehtonen, A. (2011). Longitudinal changes in serum lipids in older people the Turku elderly study 1991–2006. Age and ageing, 40(2), 280-283. Uribarri, J. (2007). PHOSPHORUS METABOLISM AND MANAGEMENT IN CHRONIC KIDNEY DISEASE: Phosphorus Homeostasis in Normal Health and in Chronic Kidney Disease Patients with Special Emphasis on Dietary Phosphorus Intake. Seminars in Dialysis, 20(4), 295–301. https://doi.org/10.1111/j.1525-139X.2007.00309.x Vadiveloo, T., Donnan, P. T., Murphy, M. J., & Leese, G. P. (2013). Age- and Gender- Specific TSH Reference Intervals in People With No Obvious Thyroid Disease in Tayside, Scotland: The Thyroid Epidemiology, Audit, and Research Study (TEARS). The Journal of Clinical Endocrinology & Metabolism, 98(3), 1147– 1153. https://doi.org/10.1210/jc.2012-3191 Valcour, A., Blocki, F., Hawkins, D. M., & Rao, S. D. (2012). Effects of age and serum 25-OH-vitamin D on serum parathyroid hormone levels. The Journal of Clinical Endocrinology & Metabolism, 97(11), 3989-3995. Van den Bosch, G., Van den Bossche, J., Wagner, C., De Schouwer, P., Van De Vyvere, M., & Neels, H. (2001). Determination of iron metabolism-related reference values in a healthy adult population. Clinical chemistry, 47(8), 1465-1467. van der Velde, M., Bakker, S. J., de Jong, P. E., & Gansevoort, R. T. (2010). Influence of age and measure of eGFR on the association between renal function and cardiovascular events. Clinical Journal of the American Society of Nephrology, 5(11), 2053-2059. Varki, A., Geschwind, D. H., & Eichler, E. E. (2008). Human uniqueness: genome interactions with environment, behaviour and culture. Nature Reviews Genetics, 9(10), 749–763. https://doi.org/10.1038/nrg2428 Venugopal, S. K., Devaraj, S., & Jialal, I. (2005). Effect of C-reactive protein on vascular cells: evidence for a proinflammatory, proatherogenic role. Current Opinion in Nephrology and Hypertension, 14(1), 33–7. Retrieved from http://www.ncbi. nlm.nih.gov/pubmed/15586013 267 University of Ghana http://ugspace.ug.edu.gh Vickers, A. J., Ulmert, D., Sjoberg, D. D., Bennette, C. J., Björk, T., Gerdtsson, A., ... & Scardino, P. T. (2013). Strategy for detection of prostate cancer based on relation between prostate specific antigen at age 40-55 and long-term risk of metastasis: case-control study. Bmj, 346, f2023. Vogiatzoglou, A., Smith, A. D., Nurk, E., Berstad, P., Drevon, C. A., Ueland, P. M., … Refsum, H. (2009). Dietary sources of vitamin B-12 and their association with plasma vitamin B-12 concentrations in the general population: the Hordaland Homocysteine Study. The American Journal of Clinical Nutrition, 89(4), 1078– 1087. https://doi.org/10.3945/ajcn.2008.26598 Vuistiner, P., Rousson, V., Henry, H., Lescuyer, P., Boulat, O., Gaspoz, J.-M., … Guessous, I. (2015). A Population-Based Model to Consider the Effect of Seasonal Variation on Serum 25(OH)D and Vitamin D Status. BioMed Research International, 2015, 1–9. https://doi.org/10.1155/2015/168189 Wahlin, Å., Bäckman, L., Hultdin, J., Adolfsson, R., & Nilsson, L. G. (2002). Reference values for serum levels of vitamin B 12 and folic acid in a population-based sample of adults between 35 and 80 years of age. Public Health Nutrition, 5(3), 505-511. Wan, X., Wei, L., Li, H., Dong, M., Lin, Q., Ma, X., … Hong, M. (2013). High pretreatment serum lactate dehydrogenase level correlates with disease relapse and predicts an inferior outcome in locally advanced nasopharyngeal carcinoma. European Journal of Cancer, 49(10), 2356–2364. https://doi.org/10.1016/j.ejca. 2013.03.008 Wang, G. L., & Semenza, G. L. (1996). Molecular basis of hypoxia-induced erythropoietin expression. Current opinion in hematology, 3(2), 156-162. Wang, X., & Wang, Q. (2018). Alpha-Fetoprotein and Hepatocellular Carcinoma Immunity. Canadian Journal of Gastroenterology & Hepatology, 2018, 9049252. https://doi.org/10.1155/2018/9049252 Weaving, G., Batstone, G. F., & Jones, R. G. (2016). Age and sex variation in serum albumin concentration: an observational study. Annals of clinical biochemistry, 53(1), 106-111. Wei, X., Zhang, D., He, M., Jin, Y., Wang, D., Zhou, Y., … Xu, R. (2016). The predictive value of alkaline phosphatase and lactate dehydrogenase for overall survival in patients with esophageal squamous cell carcinoma. Tumour Biology, 37(2), 1879–1887. https://doi.org/10.1007/s13277-015-3851-y 268 University of Ghana http://ugspace.ug.edu.gh Weide, B., Elsässer, M., Büttner, P., Pflugfelder, A., Leiter, U., Eigentler, T. K., … Garbe, C. (2012). Serum markers lactate dehydrogenase and S100B predict independently disease outcome in melanoma patients with distant metastasis. British Journal of Cancer, 107(3), 422–428. https://doi.org/10.1038/ bjc.2012.306 Weiss, M. K. (1993). Genetic Variation and Human Disease: Principles and Evolutionary Approaches - Kenneth M. Weiss - Google Books. Whicher, J. T., Ritchie, R. F., Johnson, A. M., Baudner, S., Bienvenu, J., Blirup-Jensen, S., … Svendsen, P. J. (1994). New international reference preparation for proteins in human serum (RPPHS). Clinical Chemistry, 40(6). Retrieved from http://clinchem.aaccjnls.org/content/40/6/934.short Whitehouse, J. S., Riggle, K. M., Purpi, D. P., Mayer, A. N., Pritchard, K. A., Oldham, K. T., & Gourlay, D. M. (2010). The Protective Role of Intestinal Alkaline Phosphatase in Necrotizing Enterocolitis. Journal of Surgical Research, 163(1), 79–85. https://doi.org/10.1016/j.jss.2010.04.048 Whitney, A. R., Diehn, M., Popper, S. J., Alizadeh, A. A., Boldrick, J. C., Relman, D. A., & Brown, P. O. (2002). Individuality and variation in gene expression patterns in human blood. Windsor, J. S., & Rodway, G. W. (2007). Heights and haematology: the story of haemoglobin at altitude. Postgraduate medical journal, 83(977), 148-151. Wong, E. T., Cobb, C., Umehara, M. K., Wolff, G. A., Haywood, L. J., Greenberg, T., & Shaw Jr, S. T. (1983). Heterogeneity of serum creatine kinase activity among racial and gender groups of the population. American journal of clinical pathology, 79(5), 582-586. Wong, H. Y., Tan, J. Y. L., & Lim, C. C. (2004). Abnormal liver function tests in the symptomatic pregnant patient: the local experience in Singapore. Annals of the Academy of Medicine, Singapore, 33(2), 204–8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15098635 Xia, L., Chen, M., Liu, M., Tao, Z., Li, S., Wang, L., … Qiu, L. (2016). Nationwide Multicenter Reference Interval Study for 28 Common Biochemical Analytes in China. Medicine, 95(9), e2915. https://doi.org/10.1097/MD.0000000000002915 Yamakado, M., Nagao, K., Imaizumi, A., Tani, M., Toda, A., Tanaka, T., ... & Horimoto, K. (2015). Plasma free amino acid profiles predict four-year risk of developing 269 University of Ghana http://ugspace.ug.edu.gh diabetes, metabolic syndrome, dyslipidemia, and hypertension in Japanese population. Scientific reports, 5, 11918. Yamamoto, Y., Hosogaya, S., Osawa, S., Ichihara, K., Onuma, T., Saito, A., … Iwagami, M. (2013). Nationwide multicenter study aimed at the establishment of common reference intervals for standardized clinical laboratory tests in Japan. Clinical Chemistry and Laboratory Medicine, 51(8), 1663–1672. https://doi.org/10. 1515/cclm-2012-0413 Yan, C., Hu, J., Yang, J., Chen, Z., Li, H., Wei, L., ... & Han, R. (2018). Serum ARCHITECT PIVKA-II reference interval in healthy Chinese adults: Sub- analysis from a prospective multicenter study. Clinical biochemistry, 54, 32-36. Yan, Y.-Q., Dong, Z.-L., Dong, L., Wang, F.-R., Yang, X.-M., Jin, X.-Y., … Chen, Z.-P. (2011). Trimester- and method-specific reference intervals for thyroid tests in pregnant Chinese women: methodology, euthyroid definition and iodine status can influence the setting of reference intervals. Clinical Endocrinology, 74(2), 262–269. https://doi.org/10.1111/j.1365-2265.2010.03910.x Yang, J., Sa, M., Huang, M., Yang, J., Xiang, Z., Liu, B., & Tang, A. (2013). The reference intervals for HE4, CA125 and ROMA in healthy female with electrochemiluminescence immunoassay. Clinical biochemistry, 46(16-17), 1705-1708. Yen, C.-H., Wang, K.-T., Lee, P.-Y., Liu, C.-C., Hsieh, Y.-C., Kuo, J.-Y., … Lam, C. S. P. (2017). Gender-differences in the associations between circulating creatine kinase, blood pressure, body mass and non-alcoholic fatty liver disease in asymptomatic asians. PLOS ONE, 12(6), e0179898. https://doi.org/10.1371/ journal.pone.0179898 Yokoi, Y., Kondo, T., Okumura, N., Shimokata, K., Osugi, S., Maeda, K., & Murohara, T. (2016). Serum uric acid as a predictor of future hypertension: Stratified analysis based on body mass index and age. Preventive medicine, 90, 201-206. Yu, S.-L., Xu, L.-T., Qi, Q., Geng, Y.-W., Chen, H., Meng, Z.-Q., … Chen, Z. (2017). Serum lactate dehydrogenase predicts prognosis and correlates with systemic inflammatory response in patients with advanced pancreatic cancer after gemcitabine-based chemotherapy. Scientific Reports, 7(1), 45194. https://doi.org/10.1038/srep45194 Yuan, X., Dong, Z., Zhang, H., Lin, H., Song, X., Niu, Z., … Lü, J. (2011). Distribution of serum prostate-specific antigen in Chinese healthy men: a population-based 270 University of Ghana http://ugspace.ug.edu.gh study. Chinese Medical Journal, 124(8), 1189–92. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/21542994 Yuji, S., Mio, N., Takaharu, S., Koichiro, K., Shimpei, S., Kazuhiko, A., … Takahiro, M. (2014). Association between Alkaline Phosphatase and Anaemia in Rural Japanese Men: The Nagasaki Islands study. Acta Medica Nagasakiensia, 58(4), 125−130. https://doi.org/10069/34224 Zeh, C. E., Odhiambo, C. O., & Mills, L. A. (2012). Laboratory reference intervals in Africa. Blood Cell: An Overview of Studies in Hematology, 303. Zeh, C., Amornkul, P. N., Inzaule, S., Ondoa, P., Oyaro, B., Mwaengo, D. M., … Laserson, K. (2011). Population-Based Biochemistry, Immunologic and Haematological Reference Values for Adolescents and Young Adults in a Rural Population in Western Kenya. PLoS ONE, 6(6), e21040. https://doi.org/10.1371/journal. pone.0021040 Zhang, J., Yao, Y.-H., Li, B.-G., Yang, Q., Zhang, P.-Y., & Wang, H.-T. (2015). Prognostic value of pretreatment serum lactate dehydrogenase level in patients with solid tumours: a systematic review and meta-analysis. Scientific Reports, 5(1), 9800. https://doi.org/10.1038/srep09800 Zhang, Z., Zhang, Y., Wang, Y., Xu, L., & Xu, W. (2016). Alpha-fetoprotein-L3 and Golgi protein 73 may serve as candidate biomarkers for diagnosing alpha-fetoprotein- negative hepatocellular carcinoma. OncoTargets and therapy, 9, 123. Zhou, B., Liu, J., Wang, Z.-M., & Xi, T. (2012). C-Reactive Protein, Interleukin 6 and Lung Cancer Risk: A Meta-Analysis. PLoS ONE, 7(8), e43075. https://doi.org/10.1371/journal.pone.0043075 Zhou, B., Shu, B., Yang, J., Liu, J., Xi, T., & Xing, Y. (2014). C-reactive protein, interleukin-6 and the risk of colourectal cancer: a meta-analysis. Cancer Causes & Control, 25(10), 1397–1405. https://doi.org/10.1007/s10552-014-0445-8 Zhou, X., Pang, Z., Gao, W., Wang, S., Zhang, L., Ning, F., & Qiao, Q. (2010). Performance of an A1C and fasting capillary blood glucose test for screening newly diagnosed diabetes and pre-diabetes defined by an oral glucose tolerance test in Qingdao, China. Diabetes Care, 33(3), 545–50. https://doi.org/ 10.2337/dc09-1410 271 University of Ghana http://ugspace.ug.edu.gh APPENDICES APPENDIX I: QUESTIONNAIRE 272 University of Ghana http://ugspace.ug.edu.gh 273 University of Ghana http://ugspace.ug.edu.gh 274 University of Ghana http://ugspace.ug.edu.gh 275 University of Ghana http://ugspace.ug.edu.gh APPENDIX II: CONSENT FORM 276 University of Ghana http://ugspace.ug.edu.gh 277 University of Ghana http://ugspace.ug.edu.gh 278 University of Ghana http://ugspace.ug.edu.gh Declaration by Participant By signing below, I ……………………………… agree to take part in a research entitled ESTABLISHING ADULT REFERENCE VALUES FOR SELECTED ANALYTES I declare that:  I have read or had read to me this information and consent form has been written in a language with which I am fluent and comfortable  I have had a chance to ask questions and all my questions have been adequately answered.  I understand that taking part in this study is voluntary and I have not been pressurized to take part.  I may choose to leave the study at any time and will not be penalized or prejudiced in any way  I may be asked to leave the study before it has finished, if the doctor or researcher feels it in my best interests, or if I do not follow the study plan, as agreed to. Signed at (Place)……………………………. On (date) ………………………… Signature of participant……………………………. On (date) ………………… Declaration by Investigator I (name)…………………………………………… declare that:  I explained the information in this document to ………………………………………  I encourage him/ her to ask questions and took adequate time to answer them.  I am satisfied that he/ she adequately understands all aspects of the research, as discussed above.  I did not use a translator. (If a translator is used then the translator must sign the declaration below. Declaration by translator I (name)…………………………………………… declare that:  I assisted the investigator (Name) ……………………………………… to explain the information in this document to (Name of participant) ………………………using the language medium of Twi/ Ga/ Ewe/ Other  We encouraged him/ her to ask questions and took adequate time to answer them.  I conveyed a factually correct version of what was related to me.  I am satisfied that the participant fully understands the content of this informed consent and document and has had all his/her question satisfactorily.  I did not use a translator. (If a translator is used then the translator must sign the declaration below. Signed at (Place)……………………………. On (date) ………………………… 279 University of Ghana http://ugspace.ug.edu.gh APPENDIX III: COMPARISON OF GHANAIAN RIs VS MANUFACTURER’S INSERT THE GHANAIAN DERIVED REFERENCE INTERVAL VS THE BECKMAN COULTER AU480 AND SYSMEX XN1000 STANDARDS HAEMATOLOGY ANALYTES (BLOOD, EDTA) TEST Unit GHANAIANS SYSMEX XN1000 DERIVED RIs MANUFACTURER’S RIs Male Female Male Female RBC 1012/L 4.57-6.50 4.00-5.46 4.29-5.70 3.72-5.06 Hemoglobin g/dL 12.8-17.2 10.7-14.3 13.3-16.6 11.0-14.7 Hematocrit (%) 39.4-52.1 34.0-44.2 41.3-52.1 35.2-46.7 MCV fL 71.9-97.4 71.9-97.4 86.1-101.9 87.1-102.4 MCH pg 23.2-31.8 23.2-31.8 27.5-32.4 26.8-32.4 MCHC g/dL 29.0-35.2 29.0-35.2 30.7-33.2 29.6-32.5 RDW % 11.7-16.0 12.0-17.3 12.2-14.6 12.2-15.0 WBC 109/L 3.08-7.53 3.08-7.53 3.58-8.15 3.17-8.40 Neutrophils % % 29 -62 29 -62 39.6-67.0 39.7-71.2 Neutrophils # 109/L 1.17-3.81 1.17-3.81 1.57-4.59 1.50-5.00 Lymphocytes % % 27-60 27-60 24.0-48.4 21.9-50.3 Lymphocytes # 109/L 1.22-3.38 1.22-3.38 1.13-3.02 1.05-2.87 Monocytes % % 5.0-13.7 5.0-13.7 4.8-10.1 4.2-9.6 Monocytes # 109/L 0.22-0.73 0.22-0.73 0.23-0.63 0.22-0.63 Eosinophils % % 0.5-10.3 0.4-6.5 0.8-5.8 0.6-4.9 Eosinophils # 109/L 0.03-0.53 0.02-0.29 0.04-0.35 0.03-0.27 Basophils % % 0.2-1.5 0.2-1.5 0.4-1.4 0.2-1.4 Basophils # 109/L 0.01-0.08 0.01-0.08 0.02-0.07 0.02-0.07 MPV fL 9.1-12.8 9.1-12.8 9.1-12.0 9.2-12.1 Platelets 109/L 115-339 157-402 172-359 167-390 280 University of Ghana http://ugspace.ug.edu.gh HORMONES & TUMOR MARKERS (SERUM) TEST Unit GHANAIANS RIs MANUFACTURER’S RIs Male Female Male Female ENDOCRINE STUDIES Prolactin ng/mL 4.8-47.2 4.79-23.3 Estradiol pmol/L 40-1242 45.4-854 LH mIU/mL 1.3-47.0 2.4-12.6 FSH mIU/mL 1.1-39 3.5-12.6 Progesterone nmol/L 0.1 – 10 0.181-2.84 (age<45) Testosterone nmol/L 10.0-41.9 6.68-25.7 Cortisol ng/mL 47-183 31-141 43-224 43-225 Parathyriod pg/mL 16.0-80 23-97 14-72 14-72 Hormone Somatotrope ng/mL 0.0-2.5 0.0-4.5 <3 <8 Hormone Insulin uU/mL 2.0-18.4 2.1-17.8 3.0-25.0 3.0-25.0 THYROID STUDIES TSH uIU/mL 0.6-3.4 0.5-4.0 0.27-4.2 0.27-4.2 Free T3 pmol/L 4.1-6.3 3.7-5.8 3.1-6.8 3.1-6.8 Free T4 pmol/L 12.6-20.7 12.8-20.4 12-22 12-22 Thyreoperoxidase U/mL 29-57 28-114 <60 <60 Thyreoglobulin U/mL 15 -33 15.1-91.4 <60 <60 TUMOR MARKERS PSA Total ng/mL 0.2-2.8 0.3-1.0 0.0-4.0 Alpha 1-Fetoprotein ng/mL 1.18-15.6 1.8-15.6 <10 <10 CA125 U/mL 1.97-15 2.94-48 <35 <35 Carcinoembryonic ng/mL 0.41 – 4.4 0.41-4.4 <4.0 <4.0 Antigen (CEA) 281 University of Ghana http://ugspace.ug.edu.gh CLINICAL CHEMISTRY (SERUM) TEST UNITS GHANAIANS DERIVED MANUFACTURER’S RI RI Male Female Male Female Calcium mmol/L 2.21-2.57 2.21-2.57 2.15-2.58 2.15-2.58 Corrected Ca+ mmol/L 2.2-2.47 2.2-2.47 2.02-2.60 2.02-2.60 Magnesium mmol/L 0.75-1.04 0.75-1.04 0.73-1.06 0.77-1.03 Phosphate mmol/L 0.8-1.59 0.98-1.93 0.81-1.45 0.81-1.45 CK U/L 93.0-502 52-276 <171 <145 LDH IU/L 124-278 134-296 <248 <247 AMY U/L 47-177 43-158 28-100 22.0-80.0 Uric Acid umol/L 231-487 149-377 208.3-428.4 154.7-357.0 C-Reactive Prot. mg/dL 0.18-4.94 0.47-9.85 0.0-10.0 0.0-10.0 C3 Complement mg/L 81.7-161 94.5-168 84-160 84-160 C4 Complement mg/dL 14.9-44.6 15.8-50.6 12-36 12-36 LIVER FUNCTION TESTS Total Protein g/L 65.3-83.7 65.3-83.7 66-83 66-83 Albumin g/L 39.2-48.9 36.9-46.2 35.0-52.0 35.0-52.0 Globulin g/L 23.0-39.6 23.0-39.6 20-35 23-35 Total Bilirubin umol/L 6.9-35.8 5.3-23.9 5.0-21.0 5.0-21.0 Direct Bilirubin umol/L 1.4-6.5 0.9-4.8 <3.4 <3.4 AST IU/L 17-39 15-27 <50 <35 ALT IU/L 9 -50 7-26 <50 <35 ALP IU/L 39-111 39-111 30-120 30-120 GGT IU/L 17-87 12-49 <55 <38 KIDNEY FUNCTION TESTS Urea mmol/L 2.17-5.63 2.17-5.63 2.8-7.2 2.8-7.2 Creatinine umol/L 58-109 41-82 64-104 49-90 eGFR mL/min 79.1-179 83-205 >60 >60 Sodium mmol/L 136-143 135-143 136-146 136-146 potassium mmol/L 3.5-5.0 3.5-5.0 3.5-5.1 3.5-5.1 Chloride mmol/L 98-108 98-108 101-109 101-109 Total CO2 mmol/L 18.9-27.9 17.4-26.1 21-31 21-31 ANGAP mmol/L 8.5-21.2 8.5-21.2 4-14 4-14 DIABETES SCREEN Glucose, fasting mmol/L 3.94-5.95 3.94-5.95 4.1-5.9 4.1-5.9 282 University of Ghana http://ugspace.ug.edu.gh HbA1c% % 4.24-6.27 4.24-6.27 4.0-6.2 4.0-6.2 LIPIDS PROFILE Total Cholesterol mmol/L 3.54 -7.51 3.54 -7.51 <5.2 <5.2 HDL-Cholesterol mmol/L 0.83-1.89 0.94-2.03 >1.55 >1.55 Triglycerides mmol/L 0.4-1.9 0.4-1.6 <1.7 <1.7 LDL-Cholesterol Ratio 2.01-5.62 1.9-5.5 <4.1 <4.1 CHOL_HDL mg/L 2.5-6.7 2.6-6.0 <5.60 <5.60 HDL_LDL mg/dL 0.2-0.76 0.2-0.76 >0.24 >0.24 IMMUNOGLOBULINS Immunoglobulin G mg/dL 1165-2164 1165-2164 650-1600 650-1600 Immunoglobulin A mg/dL 114-417 114-417 70-350 70-350 Immunoglobulin M mg/dL 38-195 51-265 50-300 50-300 IRON STUDIES Iron umol/L 8.6-28.5 5.5-22.4 12.5-32.2 10.7-32.2 Ferritin ng/mL 39-502 15-300 30-400 13-150 Transferrin mg/dL 195-308 199-345 215-365 250-380 VITAMINS Vitamin B12 pg/mL 289-1199 332-1330 211-911 211-911 Folate ng/mL 2.1-14.1 2.0-19.3 >5.4 >5.4 Vitamin D ug/L 14.4-38.0 14.4-38.0 30-60 30-60 283