REGIONAL INSTITUTE FOR POPULATION STUDIESAT THEUNIVERSITY OF GHANA, LEGON FEMALE ADOLESCENT SEXUALITY, CONTRACEPTIVE USE AND REPRODUCTIVE HEALTH IN THE CENTRAL REGION OF GHANA: A CASE STUDY OF CAPE COAST ANDMANKRONG BY STEPHEN OWUSU KWANKYE THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AW ARD OF PH.D DEGREE INPOPULATION STUDIES APRIL 2003 i6>7? • «*(<* i W! { - c - ( ACCEPTANCE Accepted by the Faculty o f Social Studies, University o f Ghana, Legon, in fulfilment o f the requirement for Degree in Ph.D (Population Studies). Supervisors o f Thesis: .......Naa Prof. John S. Nabila DECLARA TION I hereby declare that, except for references to other people’s work which have been duly acknowledged, this work is the result o f my own research and that it has neither in part nor in whole been presented elsewhere for another degree. Stephen Owusu Kwankye Date I dedicate this thesis to my very understanding children, Kwame, Akos, Nana and Mercy. DEDICATION ACKNOWLEDGEMENTS A number o f individuals and institutions have contributed a great deal to ensure the completion o f this thesis. In fact, it has first and foremost been the abundant grace o f God that has seen me this far. Doing a Ph.D on part-time in Ghana while a full-time lecturer is not easy and but for God’s provision o f good health, I might not have completed the thesis. To God be all the glory, honour and thanksgiving. I wish to also thank the University o f Ghana for first giving me the permission to do this course on part-time and also granting me fellowship under the A.G. Leventis facility for a three-year period. I also would like to register my appreciation to CODESRIA for granting me some fellowship that assisted me take care o f part o f the cost o f data collection for the study. Furthermore, I express my profound gratitude to my supervisors namely Dr. E.O.Tawiah, Dr. Tesfay Teklu, Prof. J.K. Anarfi and Naa Prof. John S. Nabila for their guidance and invaluable suggestions as well as providing the needed encouragement especially in very difficult times. To all these kind-hearted persons, I owe a great deal of gratitude not only on account o f this thesis but my entire academic training. Prof. George Benneh, former Vice-Chancellor o f the University o f Ghana, Legon, Prof. A.F. Aryee and the entire management and staff o f the Population Impact Project (PIP), both past and present have also been a great source o f inspiration and encouragement and I cannot forget to thank them. The same goes to all colleagues and staff o f RIPS who co-operated with me in diverse ways while I doubled as lecturer and student at the same time at the Institute. I must particularly mention Mr. Eric August and Ms. Rebecca Atisu for their assistance in the data cleaning and analysis as well as providing secretarial services at very crucial periods of this project. v To my District Pastor, Godwin Osei Bonsu, Odorkor District o f The Church of Pentecost and his entire executive, I owe much gratitude for standing solidly with me during very critical moments o f my life. I shall never forget their wise counselling and prayer support. I cannot forget my immediate family namely, my wife Vida and my children Kwame, Akos, Nana and Mercy for bearing with me when I had to combine academic, church duties and domestic chores. Their sacrifices especially during these five and half years I undertook this study, will ever be remembered. I also would wish to thank my mother, brothers and sisters whose support for me has never been in doubt. I further wish to mention my nephew, Charles and his wife Esther for their immense contribution during very critical moments when I was doing this study. To all these persons and others who in diverse ways assisted me complete this research, I pray that God will bless them. I must, however, emphasize that I am solely responsible for any errors and criticisms that may arise as a result of this study. April 2003 S.O. Kwankye TABLE OF CONTENTS TITLE Title Page Acceptance Note Declaration Dedication Acknowledgements Table o f Contents List of Tables List o f Figures Appendices Abstract CHAPTER ONE: INTRODUCTION 1.1 Background to the Study 1.2 Statement o f the Problem 1.3 Rationale of the Study 1.4 Objectives 1.5 Literature Review 1.5.1 Adolescent Fertility 1.5.2 Determinants of Adolescent Sexuality and Reproduction 1.5.3 Adolescent’s Reproductive Health Knowledge, Attitudes and Practices 1.5.4 Knowledge of HIV/AIDS and STIs Among Adolescents 1.5.5 Contraceptive Education and Practice 1.5.4 Adolescent Pregnancy Risks and Consequences 1.6 Conceptual Framework PAGE in iv v vii xii xx xxi xxii 1 1 2 6 8 9 10 13 18 22 24 27 32 vii 1.7 Hypotheses ... ... 34 1.8 Sources o f Data and Methodology ... ... 35 1.8.1 Sources o f Data ... ... 35 1.8.2 Sample Design ... ... 37 1.8.3 Techniques o f Analysis ... ... ... 39 1.9 Definition o f Concepts ... ... ... 39 1.10 Limitations of the Study ... ... ... 41 1.11 Organisation of the Study ... ... ... 42 CHAPTER TWO: HOUSEHOLD CHARACTERISTICS OF THESTUDYAREAS ... ... ... ... 44 2.1 Introduction ... ... ... ... 44 2.2 Age and Sex Distribution o f Household Members ... ... 44 2.3 Household Headship ... ... ... 47 2.4 Household Size ... ... ... ... 49 2.5 Household Conditions ... ... ... 51 2.6 Socio-Economic Status o f Households ... ... 53 CHAPTER THREE: DEMOGRAPHIC AND SOCIO-ECONOMICCHARACTERISTICS OF FEMALE ADOLESCENTS ... 57 3.1 Introduction ... ... ... ... 57 3.2 Age Distribution o f Adolescents ... ... ... 57 3.3 Education ... ... ... ... 59 3.4 Marital Status ... ... ... ... 61 3.5 Migration Status ... ... ... ... 62 3.6 Religious Affiliation ... ... ... ... 68 3.7 Ethnicity ... ... ... .. 69 viii 3.8 Occupation and Economic Activity 3.9 Living Arrangements CHAPTER FOUR: ADOLESCENT SEXUALITY AND MARRIAGE 4.1 Introduction 4.2 Age at First Sex 4.2.1 Living Arrangements 4.2.2 Level o f Education 4.2.3 Religion 4.2.4 Frequency o f Sexual Intercourse ... 4.3 Characteristics o f Sexual Partners o f Adolescents 4.4 Reasons for Adolescents’ Initiation into Sexual Activity 4.5 Adolescents’ Knowledge About Sexuality 4.6 Age at First Marriage 4.7 Decision to Marry CHAPTER FIVE: ADOLESCENT CHILDBEARING AND REPRODUCTIVEKNOWLEDGE 5.1 Introduction 5.2 Age at First Pregnancy 5.3 Age at First Birth 5.4 Adolescent Fertility and Pregnancy Wastage 5.5 Incidence o f Pregnancy and School Attendance 5.6 Desired Fertility Among Adolescents 5.7 Reproductive Knowledge Among Adolescents CHAPTER SIX: CONTRACEPTIVE KNOWLEDGE AND USE 6.1 Introduction ix 70 73 76 76 76 81 83 83 85 88 98 110 115 121 126 126 126 135 137 148 153 158 162 162 6.2 Contraceptive Knowledge ••• ••• 162 6.3 Contraceptive Use at First Sexual Intercourse ... ... 171 6.4 Current Contraceptive Use ... ... ... 178 6.5 Unmarried Persons and Contraception ... ... 187 6.6 Family Life Education and Adolescent Sexual Promiscuity ... 194 6.7 Provision o f Counselling on Contraception for Adolescents ... 199 CHAPTER SEVEN: KNOWLEDGE, INCIDENCE AND PREVENTION OFREPRODUCTIVE HEALTH PROBLEMS AMONG ADOLESCENTS 204 7.1 Introduction ... ... ... ... 204 7.2 Knowledge o f Sexually Transmitted Infections (STIs) ... ... 204 7.3 Incidence and Treatment o f STIs ... ... ... 210 7.4 Knowledge o f HIV/AIDS Transmission and its Avoidance ... 214 CHAPTER EIGHT: RELATIONSHIP BETWEEN ADOLESCENTSEXUALITY, CONTRACEPTION AND REPRODUCTIVE HEALTH 222 8.1 Introduction ... ... ... ... 222 8.2 Adolescent Sexuality and Contraceptive Use ... ... 222 8.3 Adolescent Sexuality and Childbearing ... ... 229 8.4 Contraception and Childbearing ... ... ... 236 8.5 Multiple Regression Analysis o f the Relationship ... ... 242 x CHAPTER NINE: SUMMARY, RECOMMENDATIONS ANDCONCLUSION ... ... ... ... 254 9.1 Summary o f Findings ... ... ... ... 254 9.2 Policy Implications and Recommendations ... ... 269 9.2.1 Abstention From Sex ... ... ... 269 9.2.2 Abortion ... ... ... ... 271 9.2.3 Youth Counselling Centres and Family Planning Education ... 272 9.2.4 Future Research ... ... ... 273 9.3 Conclusion ... ... ... ... 274 REFERENCES ... ... ... ... 275 APPENDICES ... ... ... ... 281 xi LIST OF TABLES TABLE PAGE 2.1 Age and Sex Characteristics o f Household Members, Cape Coastand Mankrong ... • • • ■ 45 2.2 Distribution o f Household Members by Region o f Birth, Cape Coastand Mankrong ... ... • ■ ■ 46 2.3 Distribution o f Households by Sample Area ... ... 47 2.4 Sex Composition o f Household Heads in Cape Coast and Mankrong 2.5 Age Distribution o f Heads o f Household in Cape Coast and Mankrong 2.6 Average Household Size by Study Area 2.7a Type o f Main Material Used in Constructing Outer Walls o fHousehold Premises, Cape Coast and Mankrong ... ... 52 2.7b Type o f Material Used in Roofing Household Premises, Cape Coastand Mankrong ... ... ... ... 52 2.7c Availability o f Electricity Within Household Premises, Cape Coastand Mankrong ... ... ... 53 2.8a Ownership o f Household Premises, Cape Coast and Mankrong ... 54 2.8b Availability o f Radio and Television to Households in Cape Coastand Mankrong ... ... ... 55 2.8c Perception o f Household About their Socio-Economic Status,Cape Coast and Mankrong. ... ... ... 56 3.1 Distribution o f Female Adolescents by Single Years o f Age,Cape Coast and Mankrong ... ... ... 58 3.2. Percentage Distribution o f Female Adolescents by Age and CurrentLevel o f Education, Cape Coast and Mankrong ... ... 60 3.3 Percentage Distribution o f Female Adolescents by Age andCurrent Marital Status, Cape Coast and Mankrong ... ... 61 3.4a Distribution of Adolescent Migrants by Region o f Birth, Cape Coastand Mankrong ... ... ... 64 xii 3.4b Distribution o f Adolescent Migrants by Reason for Migrating,Cape Coast and Mankrong 3.5 Distribution o f Female Adolescents by Type of Economic Activity,Cape Coast and Mankrong 4.1 Percentage Distribution o f Adolescents Who Have Ever Had Sex by Current Age, Cape Coast and Mankrong 4.2 Mean Age at First Sex o f Adolescents by Current Age,Cape Coast and Mankrong 4.3 Mean Age at First Sex of Adolescents by Living Arrangements, Cape Coast and Mankrong 4.4 Mean Age at First Sex of Adolescents by Current Level of Education, Cape Coast and Mankrong 4.5 Mean Age at First Sex o f Adolescents by Religious Affiliation,Cape Coast and Mankrong 4.6a Percentage Distribution o f Sexually Active Adolescents byFrequency o f Sexual Activity During the Past Month, Cape Coast and Mankrong 4.6b Percentage Distribution o f Sexually Active Adolescents by Age and Frequency of Sexual Activity During the Past Month, Cape Coast and Mankrong 4.7 Percentage Distribution o f Sexually Active Female Adolescents With or Without Regular Sexual Partners by Age at First Sex,Cape Coast and Mankrong 4.8 Mean Age o f Regular Sexual Partner by Current Age o f Female Adolescent and Marital Status, Cape Coast and Mankrong 4.9 Mean Age o f Sexual Partner by Age o f Adolescent at First Sex, Cape Coast, Mankrong 4.10 Percentage Distribution o f Sexual Partners o f Female Adolescents by Occupation, Cape Coast and Mankrong 4.11 Percentage Distribution of Female Adolescents by Reason for First Sexual Encounter and Age at First Sex, Cape Coast and Mankrong 4.12 Percentage Distribution o f Female Adolescents by Current Age and Perception o f Pre-Marital Sex 65 72 78 81 82 84 85 86 86 88 91 94 97 99 102 xiii 4.13a Percentage Distribution o f Female Adolescents Who PerceivePre-Marital Sex to be Acceptable by Reason, Cape Coast and Mankrong 4.13b Percentage Distribution o f Female Adolescents Who do notPerceive Pre-Marital Sex to be Acceptable by Reason, Cape Coast and Mankrong 4.14 Percentage Distribution o f Female Adolescents by Current Age and Source o f Knowledge About Sexuality, Cape Coast and Mankrong 4.15 Percentage Distribution o f Female Adolescents by Current Age and Knowledge as to Whether a Girl can be Pregnant at First Sexual Intercourse, Cape Coast and Mankrong 4.16 Mean Age at First Marriage by Current Age o f Female Adolescents,Cape Coast and Mankrong 4.17 Mean Age o f Sexual Partner by Age at First Marriage o f Female Adolescent, Cape Coast and mankrong 4.18 Percentage o f Ever Married Female Adolescents by Person Who Influenced them to marry for the First Time, Cape Coastand Mankrong 4.19 Mean Age at First Marriage o f Female Adolescents by Person Influencing Her to marry 4.20 Percentage o f Female Adolescents in Respect o f Whether their Marriages were Pregnancy-Induced by Age at First Marriage,Cape Coast and Mankrong 4.21 Percentage o f Adolescent Females by Reasons Given as to Whether or Not they would Advise their Daughters to marry at the Same Age they got married for the First Time, Cape Coast andMankrong 5.1a Percentage o f Female Adolescents by Current Age and Number of Times Pregnant, Cape Coast and Mankrong 5.1b Percentage o f Female Adolescents by Age Group, Marital Status and Number of Times Pregnant, Cape Coast and Mankrong 5.2 Mean Age at First Pregnancy o f Female Adolescents by Number of Times Pregnant, Cape Coast and Mankrong 5.3 Percentage o f Female Adolescents by Number of Times Pregnant and Level o f Education, Cape Coast and Mankrong ... 104 105 112 113 117 118 119 120 122 123 129 130 130 132 xiv 5.4 Mean Age at First Birth Among Female Adolescents and the Youthby Current Age, Cape Coast and Mankrong ... 136 5.5 Percentage o f Female Adolescents Having Births in the Past Twelve Months Before the Survey by Current Age, Cape Coastand Mankrong ... ... ••• 137 5.6 Percentage o f Adolescents by Number of Children Ever Had andCurrent Age, Cape Coast and Mankrong ... ... 138 5.7 Percentage o f Female Adolescents by Current Age and Numbero f Times Terminating a Pregnancy, Cape Coast and Mankrong ... 142 5.8a Percentage Distribution o f Female Adolescents (15-24 Years) byNumber o f Births and Pregnancies Ever Had, Cape Coast ... 143 5.8b Percentage o f Female Adolescents (15-24 Years) by Numberof Births and Pregnancies Ever Had, Mankrong ... 144 5.9 Percentage o f Female Adolescents by Reason for TerminatingPregnancy, Cape Coast and Mankrong ... ... 145 5.10 Percentage Distribution o f Female Adolescents Ever Terminatinga Pregnancy by Person Providing the Service, Cape Coast and Mankrong 146 5.11 Percentage o f Female Adolescents Ever Becoming Pregnant by Current Age and whether First Pregnancy Took Place While inSchool, Cape Coast and Mankrong ... ... ... 150 5.12 Percentage o f Female Adolescents Ever Becoming Pregnant by Age at First Pregnancy and whether First Pregnancy Took PlaceWhile in School, Cape Coast and Mankrong ... ... 151 5.13a Percentage o f Female Adolescents by Current Age and Numberof Children Desired, Cape Coast ... ... ... 154 5.13b Percentage o f Female Adolescents by Current Age and Numberof Children Desired, Mankrong ... ... ... 155 5.14a Percentage o f Female Adolescents by Number of Children Desiredand Level o f Education, Cape Coast ... ... ... 156 5.14b Percentage o f Female Adolescents by Number o f Children Desiredand Level of Education, Mankrong ... ... ... 157 5.15 Percentage o f Female Adolescents by Knowledge ofthe Timing o f Pregnancy in the Ovulatory Cycle and CurrentAge, Cape Coast and Mankrong ... ... ... 160 xv 5.16 Percentage o f Female Adolescents by Knowledge o f theTiming o f Pregnancy in the Ovulatory Cycle and Level o f Education, Cape Coast and Mankrong 6.1 Percentage o f Female Adolescents by Contraceptive Knowledge and Current Age, Cape Coast and Mankrong 6.2 Percentage o f Female Adolescents by Contraceptive Knowledge and Level of Education, Cape Coast and Mankrong 6.3 Percentage o f Female Adolescents by Contraceptive Knowledge and Current Marital Status 6.4a Percentage o f Female Adolescents by Knowledge o f Family Planning Method and Current Age, Cape Coast 6.4b Percentage o f Female Adolescents by Knowledge o f Family Planning Method and Current Age, Mankrong 6.5a Percentage o f Female Adolescents by Source o f Knowledge o f Family Planning Method and Current Age, Cape Coast 6.5b Percentage o f Female Adolescents by Source o f Knowledge o f Family Planning Method and Current Age, Mankrong 6.6a Percentage o f Female Adolescents by Source o f Knowledgeof Family Planning Method and Level o f Education, Cape Coast 6.6b Percentage o f Female Adolescents by Source o f Knowledgeof Family Planning Method and Current Education, Mankrong 6.7 Percentage o f Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse and Age at First Sex, Cape Coast and Mankrong 6.8 Percentage of Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse and Current Age, Cape Coastand Mankrong 6.9 Percentage o f Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse and Level o f Education, Cape Coast and Mankrong 6.10 Percentage o f Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse and Knowledge o f Pregnancy Occurring at First Sexual Intercourse, Cape Coast and Mankrong .. 161 163 165 165 166 167 169 169 170 171 173 174 175 176 xvi 6.11 Percentage o f Female Adolescents by Reason for Non-Use o f Contraception at First Sexual Intercourse, Cape Coastand Mankrong ... ... ... ... 177 6.12 Percentage o f Sexually Active Female Adolescents by Current Contraceptive Use and Age, Cape Coast and Mankrong ... 179 6.13 Percentage o f Sexually Active Female Adolescents by Current Contraceptive Use and Level o f Education, Cape Coast and Mankrong 181 6.14 Percentage o f Female Adolescents by the Need to Seek Parental Consent Before Using Contraception and Current Age, Cape Coastand mankrong ... ... ... ... 184 6.15a Percentage o f Female Adolescents Who Consider it Necessaryto Obtain Parental Consent Before Using Contraception by Reason,Cape Coast and Mankrong ... ... ... 185 6.15b Major Reasons Given by Female Adolescents in Cape Coast and Mankrong Who Consider it Unnecessary to Obtain Parents’Consent Before Contraception ... ... ... 186 6.16 Percentage o f Female Adolescents by the Need to Seek Parental Consent before Using Contraception and Level o f Education,Cape Coast ... ... ... ... 187 6.17 Percentage o f Female Adolescents by Views on Whetherit is Right for Unmarried Sexually Active Persons to Practise Contraception and Current Age, Cape Coast and Mankrong ... 189 6.18 Percentage o f Female Adolescents by Views on Whetherit is Right for Unmarried Sexually Active Persons to Practise Contraception and Level o f Education, Cape Coast and Mankrong ... 191 6.19 Percentage o f Female Adolescents by Views on Whether Family Life Education Promotes Sexual Promiscuity and Current Age,Cape Coast and Mankrong ... ... ... 195 6.20 Percentage o f Female Adolescents by Views on Whether Family Life Education Promotes Sexual Promiscuity andLevel o f Education, Cape Coast and Mankrong ... ... 198 6.21 Percentage o f Female Adolescents by the Need for Health and Family Planning Counselling Centres and Current Age,Cape Coast and Mankrong ... ... 200 xvii 6.22 6.23 7.1 7.2 7.3 7.4 7.5 8.1 8.2 8.3 8.4 8.5 8.6 Percentage o f Female Adolescents by Views on Whether there is the Need for Health and Family Planning Counselling Centres and Level o f Education, Cape Coast and Mankrong Percentage o f Female Adolescents Who Accept Family Planning Counselling by Person Considered Best Positioned to Provide it, Cape Coast and Mankrong Percentage o f Female Adolescents by Knowledge as to Whether or Not Infections can be Contracted Through Sex and Current Age, Cape Coast and Mankrong Percentage o f Female Adolescents by Knowledge as to Whether or Not Infections can be Contracted Through Sex and Education,Cape Coast and Mankrong Percentage o f Female Adolescents by Type o f Infection that can be Contracted Through Sex, Cape Coast and Mankrong Percentage of Female Adolescents by Knowledge o f Modes of HIV/AID Infection, Cape Coast and Mankrong Percentage o f Female Adolescents by Knowledge About how to Avoid HIV/AIDS Infection, Cape Coast and Mankrong Percentage o f Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse, Current Age and Age at First Sex, Cape Coast and Mankrong Percentage o f Sexually Active Female Adolescents by Contraceptive Use at First Sexual Intercourse, Level o f Education and Age at First Sex, Cape Coast and Mankrong Percentage o f Sexually Active Female Adolescents by Contraceptive Use, Current Age and Number o f Times Having Sexual Intercourse During Past Month, Cape Coast and Mankrong Percentage o f Female Adolescents by Contraceptive Use, Level of Education and Frequency o f Sexual Intercourse During Past Month, Cape Coast and Mankrong Mean Age at First Pregnancy Among Female Adolescents by Current Age and Age at First Sexual Intercourse, Cape Coast and Mankrong Mean Age at First Pregnancy Among Female Adolescents by Level of Education and Age at First Sexual Intercourse,Cape Coast and Mankrong 202 205 207 208 216 219 223 225 227 228 230 232 201 xviii 8.7 Mean Number o f Children Ever Bom Among Female Adolescents by Current Age and Age at First Sexual Intercourse, Cape Coast and Mankrong 8.8 Mean Number o f Children Ever Bom Among Female Adolescents by Education and Age at First Sexual Intercourse, Cape Coast and Mankrong 8.9 Mean Age at First Pregnancy Among Female Adolescents by Current Age and Contraceptive Use at First Sexual Intercourse, Cape Coast and Mankrong 8.10 Mean Age at First Pregnancy Among Female Adolescents and Contraceptive Use at First Sexual Intercourse, Cape Coast and Mankrong 8.11 Mean Number o f Children Ever Bom Among Female Adolescents by Current Age and Contraceptive Use During First Sexual Intercourse, Cape Coast and Mankrong ... 8.12 Mean Number o f Children Ever Bom Among Adolescent Females by Level o f Education and Contraceptive Use at First Sexual Intercourse, Cape Coast and Mankrong 8.13 Mean Number o f Children Ever Bom Among Female Adolescents by Current Age and Current Contraceptive Use, Cape Coast and Mankrong 8.14 Mean Number o f Children Ever Bom Among Adolescent Females by Level o f Education and Current Contraceptive Use, Cape Coast and Mankrong 8.15 Results o f Multiple Regression Analysis on Age at First Sexual Intercourse by Selected Background Variables 8.16 Results o f Multiple Regression Analysis on Number o f Pregnancies Ever Had by Selected Variables in Respect of Adolescent Females in Cape Coast and Mankrong ... 234 235 236 238 239 240 241 241 241 250 xix LIST OF FIGURES FIGURE 1.1 Conceptual Framework Showing the Interrelationships between Adolescent Sexuality, Contraceptive Use and Childbearing 1.2 Map Showing the Location o f the Study Areas, Cape Coast and Mankrong 1.3 Map Showing the Residential Clusters in Cape Coast 3.1 Distribution o f Female Adolescents by Age Group, Cape Coast and Mankrong 3.2 Distribution o f Female Adolescents by Migration Status, Cape Coast and Mankrong 3.3 Distribution o f Female Adolescents by Religious Affiliation,Cape Coast and Mankrong 3.4 Distribution o f Female Adolescents by Ethnicity, Cape Coast and Mankrong 3.5 Distribution o f Female Adolescents by Type o f Occupation,Cape Coast and Mankrong 3.6 Distribution o f Female Adolescents by Living Arrangements,Cape Coast and Mankrong 4.1 Percentage Distribution o f Adolescents Ever Having Sex by Age at First Sex, Cape Coast and Mankrong 4.2 Percentage Distribution o f Sexual Partners o f Adolescents by Age, Cape Coast and Mankrong 4.3 Percentage Distribution o f Sexual Partners o f Female Adolescents by Level o f Education, Cape Coast and Mankrong ... 5.1 Percentage o f Female Adolescents and the Youth Ever Becoming Pregnant in the Past Twelve Months by Current Age, Cape Coast and Mankrong 5.2 Percentage o f Female Adolescents Ever Terminating a Pregnancy by Current Age, Cape Coast and Mankrong 6.1 Percentage o f Sexually Active Female Adolescents by Current Contraceptive Use and Source of Supply o f Method, Cape Coast and Mankrong PAGE 34 36 38 59 63 68 70 71 74 79 90 96 133 140 183 xx APPENDICES APPENDIX PAGE A Questionnaire Instrument for Data Collection During the AdolescentSurvey on Adolescent Sexuality, Contraceptive Use and Reproductive Health, 1997 ... ... ... ... 281 B Questionnaire Instrument for Focus Group Discussions During theAdolescent Survey on Adolescent Sexuality, Contraceptive Use and Reproductive Health, 1997 ... ... ... 291 xxi ABSTRACT The study on Adolescent Sexuality, Contraceptive Use and Reproductive Health was undertaken as part o f the ongoing effort at understanding some o f the problems that confront adolescents in Ghana, With a general objective o f examining the magnitude of reproductive health-related problems pertaining to adolescent sexuality and contraceptive use, the study used a sample o f 1,828 female adolescents aged 12-24 years (1,503 from Cape Coast and 325 from Mankrong) in the Central Region o f Ghana as a case study. This was with the primary purpose o f comparing the situation in an urban vis-a-vis a rural area. The two study areas are far apart: Cape Coast at the coast while Mankrong is located in the interior o f the Central Region. It uses simple techniques including cross-tabulations as well as multiple regression analysis to examine quantitative and qualitative data collected from structured questionnaire and focus group discussions held separately among male and female adolescents and adults in Cape Coast and Mankrong. Among other things, the study found that the proportion o f adolescents ever having sex increased with higher age o f the adolescent with higher proportion having sex in Mankrong compared to Cape Coast. The mean age at first sex was, however, almost the same at 16.9 years in Cape Coast and 17 years in Mankrong. Mean age at first sex also increased with higher age o f adolescents, implying plausibly o f a declining age at first sex at the two study areas. Furthermore, there were suggestions to indicate that females become sexually active earlier than their male counterparts on account o f the fact that at first sex most females were much younger than their male partners at the time. However, it was deduced that due to the fact that first sex usually may occur with persons who may be sexually more xxii experienced and older, it may not always be easy to determine whether or not it is the female who enters into sex earlier if one depended on responses exclusively from the male or female. Generally, the adolescents did not approve o f pre-marital sex but gave the indication o f the possibility to indulge in it for financial reasons. What people approve of may thus, be different from what they may practise depending on the issues at stake. Peer pressure and lack o f knowledge were the main reasons that provided grounds for a large proportion o f the adolescents entering into sex for the first time. Again, although most o f the adolescents did not approve o f abortion, the report on abortion among them showed a possibility o f an underestimation, suggesting a situation o f a much higher abortion rate among the sexually active adolescents than was reported. Abortion rates were found to be higher in Cape Coast relative to Mankrong. Also important was the finding that adolescents who used contraception at first sex had a higher age at pregnancy. Similarly, adolescents who practised contraception had a relatively lower number o f children ever bom in contrast with those who did not practise contraception at first sex. There was also a lower number o f children ever bom with higher education o f the woman. Results from the multiple regression analysis confirmed most o f the observations made in the cross-tabulations. These included a declining age at first sex at 100 per cent level of confidence. Besides, financial considerations were found to contribute more to the reduction o f an adolescent’s age at first sexual activity. The study confirmed at 95 per cent level o f confidence, four o f the five hypotheses, namely, that: i) There is an inverse relationship between age at first sexual intercourse and childbearing among adolescents; xxiii ii) There is an inverse relationship between adolescent childbearing and contraceptive use at first sex; iii) There is an inverse relationship between education and childbearing among adolescent females; and iv) There is a direct relationship between contraceptive use at first sex and age at first sexual activity among adolescent females. However, the hypothesis that contraceptive use is directly related to the level of education o f the adolescent could not be analysed due to the very small sample of adolescents who were reported to be contracepting at the time o f the survey. It is therefore, recommended that future research should consider a much larger sample o f adolescents to facilitate such an analysis. In addition, future research should probe into the educational level of adolescents as at the time of their first sexual act for incorporation into the analysis. Similarly, future research should consider collecting information on the educational aspirations o f adolescents as a proxy in analysing the timing o f first sex among adolescents. The study further recommends intensive and sustained public education against criminal abortion, highlighting the short and long-term health implications to the adolescents while underscoring the fact that abortion is not a family planning method. It also calls for a review o f the educational policy in Ghana with a possibility o f making Family Life Education (FLE) a completely separate subject from Social Studies and be taught at earlier stages o f the primary school education possibly at Primary Class Four. Besides, to ensure that teachers who are well trained in FLE teach pupils, FLE should be a compulsory component o f the curricula o f teacher training colleges in the country. xxiv A call is also made for the establishment o f District Youth Centres throughout the country to offer education and counselling on all issues pertaining to the adolescents, especially their sexual and reproductive health. The study concludes by calling on government and civil society to put adolescent sexuality and reproductive health as a national development priority and to show the maximum commitment in addressing the various dimensions o f the problem principally as critical components o f the efforts at addressing the HIV/AIDS epidemic in Ghana. This is borne out o f the realisation that the adolescents offer a window o f opportunity and hope in the fight against HIV/AIDS among the entire population o f Ghana. xxv CHAPTER ONE INTRODUCTION 1.1 Background to the Study Adolescent sexuality and reproduction are issues that are attracting worldwide recognition as social problems o f development. In many sub-Saharan African countries, teenage/adolescent pregnancy has been relatively high although declines in overall fertility are being recorded in most developing countries. For example, between 1988 and 1993, overall fertility in Ghana is reported to have declined by almost one child per woman from 6.4 to 5.5 according to the 1993 Ghana Demographic and Health Survey (Ghana Statistical Service, 1994) and a further decline to 4.6 in 1998 as per the 1998 Ghana Demographic and Health Survey (GDHS). In most developing countries, it appears that reduction in fertility among younger women would be achieved mainly through the postponement o f marriage. With age at marriage in Ghana steadily rising, the risk o f out-of-wedlock pregnancy among individual adolescent women is likely to increase as pre-marital sex continues with little or no contraceptive use. Rapid urbanization in Ghana has been the result o f natural population increase as well as inter- and intra-regional movements, which in recent times, appear to be dominated by adolescents and the youth in search o f jobs particularly in the cities and large towns. Separation o f adolescents from their parents obviously is likely to result in the steady breakdown o f traditional and parental controls on the lifestyles o f the adolescent migrants. Because o f difficulty in getting jobs in their places of destination, many female adolescents may be tempted to take to commercial sex as a matter of survival and, in the process may begin childbearing earlier than anticipated. In many parts o f Ghana, especially in the cities, defilement of female adolescents is increasing and has attracted much concern. Apart from its criminal nature, adolescent female 1 defilement or rape carries with it long term reproductive, psychological and health implications for the female adolescent victim o f sexual abuse. The issues o f adolescent sexual activity and reproduction are likely to become increasingly more important in Africa in the near future due to urbanization and the dramatic growth in secondary schooling (Cherlin and Riley, 1986). This observation is equally valid for Ghana and a source of worry particularly because contraceptive use among adolescents in Ghana is not acceptable by some sections o f the adult population. A situation o f this nature requires that a research is carried out to find out what really exists on the ground. To achieve this purpose, the Central Region is chosen as the study area with a focus on Cape Coast, the regional capital located at the coast and Mankrong, a rural settlement in the interior of the region. It is hoped that results from the study covering the two settlements will provide plausible bases for making conclusions to guide adolescent policy programmes aimed at reducing the magnitude of adolescent reproductive health problems in the Central Region in particular and Ghana as a whole. 1.2 Statement of the Problem Rapid population growth has been one of the thorny problems most African countries have had to contend with in their socio-economic development efforts. Ghana is no exception and, as far back as 1969 a comprehensive National Population Policy was adopted. Over 25 years have passed but the indications are that the nation’s efforts at reducing population growth have achieved modest results. In recent times, the population problem has been assuming different dimensions, which should attract research. This is in the area of adolescent fertility, which is the result o f early age at sexual experience especially for females in Ghana. This constitutes a serious problem considering 2 that most o f these sexual activities are taking place outside wedlock and without conti aceptive use. As a result, teenage pregnancies and their accompanying child delinquency and poor development of the female adolescent victim are unfolding problems that Ghana will have to contend with. The foregoing situation is made worse by the cultural environment within which adolescent sexuality is taking place. In Ghana, it is felt in many circles that it is culturally unwelcome to discuss sexual issues with the adolescent. Again, sections o f the society and even some family planning service providers frown upon the provision o f family planning services to these youngsters. For example, according to the 1994 Situation Analysis o f Family Planning Service Delivery Points in Ghana, 40 per cent o f service providers expressed their unwillingness to provide family planning services such as intra-uterine device (RID) and injectables to unmarried adolescents and 25 per cent o f them would not administer the pill under the same conditions (GSS, 1994). Contraceptive advertisements on the television that use adolescents to present the messages have been criticised by a section o f the populace to constitute an introduction o f the adolescents to sexual promiscuity. Thus, although it is clear that adolescent sexuality is a common phenomenon of Ghana’s social life, adolescents are not encouraged to find out what methods there are for their use to avoid any untimely pregnancies should they choose to have sex. Not quite surprisingly, some girls continue to have their future development permanently impaired through early practice of sex and subsequent pregnancy. The conditions are again made fertile for criminal abortion to thri ve in Ghana. Data are not readily available on abortion in Ghana, but in a society where adolescents are increasingly indulging in sex with little or no family planning practices, it is obvious that abortion will be resorted to by many o f the girls who become pregnant. The possibility o f increasing unsafe abortion contributing to high maternal mortality ratios cannot be underestimated. Furthermore, adolescent sexuality, apart from having the tendency to cause unplanned pre­ marital pregnancies, carries another more serious problem o f the infection and spread o f sexually 3 transmitted infections (STIs), including HIV/ADDS. The extent o f AIDS prevalence in Ghana is often shrouded in mystery and people do not even want to hear about it. Yet, data obtained from the National AIDS Control Programme for the period 1986-2000 suggest that AIDS prevalence is increasing in Ghana and that a higher proportion o f its contraction takes place during adolescent ages of 15-19 years. This is because available data on reported AIDS cases by age indicate that for both males and females, almost 75 per cent o f AIDS cases were reported among persons aged 20- 39 years and 4 per cent among adolescents aged 15-19 years. The implication is that considering that the incubation period o f the disease ranges between 5 and 12 years or more, it is possible that many of the reported cases in the age group 20-39 must have been contracted within adolescent ages. Adolescent childbearing rates in Ghana have been found to be highest in the Central Region, at least in 1993 (Ghana Statistical Service, 1994). In the 1993 Ghana Demographic and Health Survey (GDHS), 33.3 per cent o f female adolescents aged 15-19 years in the Central Region had begun childbearing as against 21.6 per cent for the entire country. In the 1998 GDHS however, the region recorded a relatively lower percentage o f adolescents having begun childbearing (18.7 per cent) to place third after Eastern Region (21.2 per cent) and Ashanti (19.6 per cent) compared to a national average o f 14.1 per cent. The Central Region also has many fishing communities, which tend to have relatively low levels o f education. This situation suggests that adolescent reproduction within the region may be perpetuated for a long time to come if concrete efforts are not made to combat it. In the 1998 GDHS, it was found that 13.8 per cent and 40.1 per cent o f women aged 15-19 and 20-24 years respectively were sexually active in the last four weeks preceding the survey. However, overall, 37.8 per cent and 91.4 per cent had ever had a sexual activity respectively among women o f 15-19 and 20-24 years. Meanwhile, only 6.5 per cent o f women 15-19 years and 42.4 per cent o f women aged 20-24 years were enumerated as married (i.e., in a formal union). 4 This suggests a relatively high sexual activity among young women including the never married in Ghana. Linked to the high sexual activity among young women is their low contraceptive use. For example, current contraceptive use for modem methods in 1993 was 5.0 per cent and 8.3 per cent among women o f 15-19 and 20-24 years respectively compared to 9.3 per cent among all women in Ghana (Ghana Statistical Service, 1994). The corresponding figures for 1998 were 4.8 per cent and 10.4 per cent among women o f 15-19 and 20-24 years respectively as against 10.7 per cent among all women in Ghana. The implication is that low contraceptive use among adolescent and young women may foreclose education and employment opportunities to many o f them as they expose themselves to risks o f unplanned pregnancies. In addition, social and health problems that arise from early motherhood for both mother and child are not in the interest o f the nation’s manpower development. For example, teenage pregnancies are contributing to school dropouts, a phenomenon which is impeding the nation's efforts at empowering women through increasing female education. In view o f the foregoing social, health and economic problems that are the result o f adolescent sexual behaviour, the relevant questions that this study seeks to answer include the following: (i) At what age does sexual activity begin among female adolescents in the Central Region in particular and Ghana as a whole? (ii) What peculiar factors are contributory to the adolescents' decision to enter into sexual activity? (iii) What are the fundamental barriers that prevent most sexually active female adolescents from practising contraception considering the high level o f unmet needs o f 48 per cent for family planning among adolescents in Ghana? 5 (iv) In what ways could adolescents be well educated and better informed about negative reproductive health consequences that result from their early entry into sexual activity to ensure either safer sex practices or outright sexual abstinence considering th e Ghanaian cultural perception that sexual issues need not be discussed openly especially with children or adolescents? It is clear then that the issue o f adolescent sexuality as it relates to contraceptive use and reproductive health is a complex one and will require a comprehensive research such as the current study in order to resolve the intriguing questions that surround it. 1.3 Rationale of the Study The study is considered both important and timely for a number o f reasons. The first consideration is the acknowledgment that adolescent reproductive health issues have become problems of national interest and concern. In line with this reasoning, the National Population Policy (Revised Edition) of Ghana (National Population Council, 1994), unlike the 1969 Ghana Population Policy, highlights adolescent reproductive health as one key area that requires critical attention in the nation's population programmes. The Revised Population Policy therefore calls for increased efforts at empowering women through increasing female education. A research o f this nature is thus relevant in bringing to the fore the various dimensions o f adolescent sexual and reproductive health problems to attract increased government attention and funding. The importance o f a study o f this magnitude can therefore not be underestimated. The Adolescent Needs Assessment Study carried out in Ghana by Nabila et. al., (1997) recommended among other things, that research facilities should be established to study the extent to which early childbearing increases health risks and also, to educate the people on the consequences of early marriage and childbearing. The current study is in response to this need to provide a 6 relatively more detailed analysis o f the dimensions o f adolescent sexual and reproductive health problems with respect to the Central Region o f Ghana. The Central Region which is the study area is also appropriate in the sense that it benefited from the pilot survey conducted by the Ghana Population Agenda Project, a World Bank/UNFPA sponsored project which sought among other things, to integrate family planning education and awareness including adolescent reproductive health, with development projects a t th e community level through community mobilisation. Cape Coast, the regional capital o f the Central Region, benefited from the pilot survey. The current study therefore offers an opportunity to examine adolescent sexual and reproductive health situation in at least one o f the towns (Cape Coast), which benefited from the activities o f the Ghana Population Agenda Project to guide future programmes that are focused on adolescents. The choice of the Central Region is also relevant as it has one of the highest adolescent childbearing rates in Ghana (it had the highest adolescent childbearing rate in the 1993 GDHS and the third according to the 1998 GDHS). The study is also timely on account of efforts by the National AIDS Control Programme (and in recent times, the Ghana AIDS Commission) and other related organisations in forestalling an AIDS epidemic in the country. This is because, since STIs and AIDS infection appears to be occurring mostly among adolescents, a research of this nature could assist a great deal in offering concrete recommendations for the implementation of family planning programmes with special reference to adolescents as well as AIDS control in Ghana. It is also important to indicate that the 1993 GDHS did not collect detailed information on adolescent reproductive health to permit an in-depth study of adolescent sexuality and reproductive health. On the other hand, although the 1998 GDHS collected some data on adolescent reproductive health, the information was not detailed enough, on account o f its national character, to permit a thorough analysis of the problem particularly with reference to Cape Coast and Mankrong, the areas 7 which, this study focuses attention on. The GDHS studies covered only adolescents o f 15-19 years and not those below 15 years, a research gap which, the current study attempts to fill. The relationship that exists between adolescent sexuality, contraceptive use and reproductive health is not o nly c omplex b ut h as r eceived 1 imited r igorous r esearch a ttention i n Ghana. This is another important research gap, which this study attempts to fill and hence, makes the study not only timely but also very much justified. The study concentrates on female adolescents as against males because even in situations where an adolescent female is impregnated by an adolescent male, it is the female that usually drops out o f school. Any complications that occur during childbearing also affect women and not men. Biologically too, reproductive health problems are more grave when women are involved compared to men. The bias o f the study towards women is also justified within the context o f on­ going efforts at improving the socio-economic plight o f women especially adolescents in Ghana. 1.4 Objectives The general objective o f the study is to examine the magnitude o f reproductive health- related problems associated with adolescent sexuality and contraceptive use for the puipose of evolving effective strategies and programmes to combat unplanned childbearing among adolescents and its negative implications for development in the Central Region in particular and Ghana as a whole. The specific objectives are to: i) Estimate the age at first sexual activity among female adolescents in the Central Region o f Ghana, ii) Investigate the factors that influence female adolescents’ early entry into sexual intercourse, iii) Discuss the variation o f childbearing among the female adolescents in respect of differences in their socio-economic characteristics, iv) Estimate th e level o f knowledge and use o f contraception among sexually active female adolescents and their implications for fertility in the two study areas, v) Find out the means by which the sexually active female adolescents who practise family planning obtain information and services, vi) Analyse the barriers that inhibit sexually active adolescents from practising contraception, vii) Examine the reproductive health consequences o f early entry into sexual activity, viii) Evolve appropriate modes o f informing adolescents about sexually related issues as a way o f breaking the cultural myth that has surrounded the discussion o f sexual matters with adolescents in Ghana, ix) Make recommendations to guide policies that are geared towards addressing sexuality and adolescent reproductive health problems in Ghana. 1.5 Literature Review Adolescent reproduction has attracted much research throughout the world in recent times. This stems largely from the increasing incidence o f adolescent pregnancies, unrecorded criminal abortions and their effects on the socio-economic development o f the individual adolescent and the wider s ociety. T he 1 iterature r eview s eeks t o b ring t o t he f ore r elevant i ssues t hat a re r elated t o adolescent reproduction that have been raised in studies on the magnitude o f adolescent fertility, its determinants, contraceptive knowledge and use, the health implications and consequences of early age at sexual intercourse and reproduction. 9 1.5.1 Adolescent Fertility The magnitude o f adolescent fertility has varied widely within and between developing countries. The Pan American Health Organisation (PAHO, 1988) estimates that 83 per cent of the world's adolescent population is found in developing countries. According to the PAHO study, adolescent fertility rates in developing countries are for the most part, higher than those o f developed countries. For example, age specific fertility rates for women aged 15-19 recorded by the PAHO study varied from as low as four births per 1,000 women in Japan in 1981 to as high as 302 births per 1,000 women in Mauritania in 1981. The study observes further that although the proportion of adolescent births to total births is declining with time, the proportion o f out-of-wedlock births has increased. In Ghana, the three demographic and health surveys report that the contribution of adolescents to overall fertility has been around 10 per cent between 1988 and 1998. Deductions from the GDHS reports on the age specific fertility rates among women 15-49 years showed that women within 15-19 years contributed 9.7 per cent of total fertility in 1988, 10.8 per cent in 1993 and 9.9 per cent in 1998. On the other hand, the contribution of women aged 20-24 years to total fertility could be computed as 20.2 per cent in 1 988,21.0 per cent in 1993 and 21.1 per cent in 1 998 (Ghana Statistical Service, 1989, 1994 and 1999). It is clear that while the contribution o f adolescents 15-19 years has remained almost constant around 10 per cent, that o f their counterparts o f 20-24 years has shown only one percentage point increase during the ten-year period o f 1988-1998. In spite o f the observed decline in the contribution o f adolescents to total fertility, out-of-wedlock births may be quite common in the country considering that a large proportion o f the sexually active adolescent women are not married. Proceedings from the International Forum on Adolescent Fertility indicate that over 15 million adolescents of 15-19 years give birth every year. Most o f these births are said to be to women i n d eveloping countries and many are not desired (Centre for Population Options, 1992). 10 The proceedings o f the forum further estimate that between one million and 4.4 million abortions take place among adolescent women each year in the developing countries. What makes the problem more serious, according to the Centre for Population Options, is that adolescents make up a high percentage of abortion-related complications and a significant percentage o f deaths due to botched abortions. In Ghana, the magnitude o f abortion among adolescents is not known. However, occasionally, cases o f complications having developed from abortions undertaken under unsafe conditions including self-inducement, seeking help from traditional healers, chemists, shopkeepers and other non-medical personnel are reported. In Ghana, abortion can be procured under certain conditions including situations o f threat of life of the mother and pregnancy being the result o f incest or rape. Analysis of the Adolescent Fertility Survey data for Kampala, Uganda, also suggests high rates of sexual activity among adolescents (Adjei and Epema, 1990). The study reveals that 85 per cent of both male and female adolescents were sexually active with mean age at first coitus being 15.7 years for males and 15.3 years for females. It was also found that by age 17 years, 30 per cent o f the women had one or more pregnancies. Similar findings were recorded in Ghana from the 1993 GDHS, which showed that by the age of 20 years, more than 85 per cent o f adolescent females had ever had sex and that 45 per cent had already begun childbearing. Moreover, the 1993 GDHS reports that by age 19 years, five per cent of the adolescent females had already had two or more children. These show that in many sub-Saharan African countries, sexual activity and childbearing are taking place among a reasonable proportion of the adolescent population and are thus, exposed to various health risks and consequences. Furthermore, in a study on adolescent fertility and reproductive behaviour in Ghana, with special reference to Accra and Kumasi, Nabila and Fayorsey (1996) examined the ages and conditions which affected the first sexual exposure among adolescent boys and girls, the socio­ economic and other determinants o f adolescent fertility, the health and other consequences of 11 adolescent fertility and the social policies and programmes dealing with adolescent sexuality in Ghana. This study showed that an unspecified proportion o f adolescents start practising sexual activity very early in life. For example, the researchers reported that some o f the adolescents (number not specified) began sexual activity as early as 10 years and although the mean age at first sexual experience was 16 years, by age 15 years, about 58 per cent o f the adolescents had already experimented with sex for the first time. Again, 11.2 per cent o f all female adolescents were found to have been pregnant once while 4.4 per cent had been pregnant two times. A significant finding, however, was that, while about 17 per cent o f the female adolescents had ever been pregnant, only 8.3 per cent had ever had children. This implies that about half o f the pregnancies among the female adolescents were lost through either miscarriages or induced abortion, a situation, which could have lasting negative impacts on the reproductive health o f those involved. Singh (1998), in a global review o f adolescent childbearing in developing countries examines current levels and recent trends in adolescent childbearing rates, the timing o f first births as well as births to unmarried women for 43 developing countries. This study which is based on nationally representative fertility surveys, demonstrates among other things, that substantial declines in adolescent fertility have been recorded in countries o f North Africa and Asia although levels are still considered high in some countries. According to this study, declines are beginning to occur in sub-Saharan Africa. In spite o f this, the proportion o f births to unmarried adolescents is increasing in some countries especially in Niger and Mali where recorded age specific fertility rate during the 1980-90s was recorded as 215 (in 1992) and 206 (in 1987) per 1,000 adolescent women aged 15-19 years respectively compared to a low o f 51 in Burundi in 1987. It was further reported that a higher proportion o f rural and less educated women had a child before age 20 years compared to urban and higher educated women. 12 1.5.2 Determinants of Adolescent Sexuality and Reproduction Many studies have focused on the determinants o f adolescent reproduction. One such study examines the problem from the biological perspective (Becker, 1993). The biological factors that are highlighted to affect adolescent fertility include the probability o f non-ovulatory cycles and foetal loss which tend to be more common among adolescents than women in their twenties. This suggests that most adolescents may be under the conviction that on account o f their relatively young ages, it is unlikely that they would be pregnant. Most of them therefore resort to unprotected sex and eventually may become victims o f unplanned births. On the other hand, in a study in Kenya, Kiragu (1991), reports that substance use could predispose an adolescent to be sexually active. For example, he shows that males and females reporting substance use are four and two times respectively, more likely to be sexually active and consequently, be at risk o f unplanned pregnancy. This view is supported by Moore (1984), in a study showing that substance use may ensue as a consequence of early sexual or pregnancy experience. He adds that the possibility exists that early sexual activity, pregnancy, abortion, adoption or parenthood, all increase the risk o f becoming a drug user. The study further posits that studies on adolescent fertility would be enhanced by the inclusion o f drug use as a possible causal variable. Kiragu's 1991 study also acknowledges the important role o f peer pressure in introducing adolescents into sexual activity. His finding, which is also supported by the findings o f Nabila and Fayorsey (1996) in Ghana, suggests that in Kenya socialising with sexually active peers results in males and females being seven and three times respectively more prone to being sexually active. Other factors, which are cited by Kiragu to predispose an adolescent to sexual activity include residence in rural areas, attainment o f puberty, unstable family environment and boarding school attendance particularly for males. 13 Nabila and Fayorsey’s (1996) study on Accra and Kumasi using focus group discussions and structured interviews also emphasised poverty, peer pressure, lack o f parental control as the main factors determining adolescent sexuality and their reproductive behaviour in both Accra and Kumasi. These findings are also supported by a Population Impact Project (PIP, 1995) study on Adolescent Fertility and Reproductive Health in Ghana which further underscores the fact o f an unmet need of about 48 per cent o f family planning among adolescents. The PIP (1995) study examined the problem of adolescent fertility in the country using 1993 GDHS data. The PIP study listed the determinants of adolescent fertility to include early age at menarche and marriage, premarital sexual experience, lack o f economic incentives, lack of knowledge of reproduction and contraception and low use o f modem family planning methods. These, according to the study, operate in various ways to affect the sexual and reproductive health of adolescents in Ghana. The problem o f poor knowledge o f a woman’s reproductive cycle and its relationship with pregnancy occurrence among adolescents has been cited as contributing to increasing adolescent fertility in Africa. Ajayi et. al. (1991) record that in Kenya, one in 10 adolescents recognised the fertile or ovulatory cycle and only 50 per cent o f them knew that pregnancy could occur at first sexual intercourse. This is an important consideration for, if adolescents are unaware that a first sexual intercourse could produce a pregnancy, they may not even consider the use o f contraception if even they have knowledge o f any. Other studies have focused attention on the limited access o f most adolescents to contraception and abortion services as a principal cause for high incidence o f adolescent fertility in most countries. For example, McGrath (1979) found that among the factors in the etiology of unintended early childbearing, the single most important factor amenable to policy intervention is lack of teenage access to contraception and abortion services. This observation is as valid today in many countries in Africa as it was almost two decades ago. In a study in Kenya and Nigeria, Barker 14 and Rich (1990) state that in Kenya for example, there is much reluctance by church and government officials t o p rovide family p lanning services to adolescents. Similarly, the Centre for Population Options urges that adult discomfort with adolescent sexuality is a major obstacle to providing the youth with needed family planning information and services. There has been one comprehensive study in Ghana by Nabila et. al. (1996) to assess the reproductive and health needs o f adolescents. It also reviewed relevant legislations and policies that affected the reproductive health o f adolescents. In addition, the study assessed the roles and capabilities of major government and non-governmental agencies in the area o f adolescent reproductive health in the country. It also provided an inventory o f all non-governmental organisations and other institutions that were involved in reproductive health activities with the youth and identified potential settings where youth services can be integrated with current activities. In addition, the obstacles inhibiting the provision o f adolescent reproductive health services throughout Ghana were examined. The obstacles include socio-cultural, legislative, medical and economic b arriers. F inally, t he r esearchers r ecommended p rogramme i nterventions for advocacy, policy formulation and research on issues pertaining to the reproductive health of adolescents. In another study on Adolescent Sexuality in Southern Africa, Meekers and Ahmed (1997) examine patterns of male and female adolescent sexual behaviour among the Tswana in Botswana. The researchers use data obtained from the 1995 Botswana Adolescent Reproductive Health Survey in addition to focus group interviews among University students. One striking finding of this study was that at younger ages (15-16 years), a higher percentage o f males (41 per cent) as against females (15 per cent) were sexually experienced. However, at later ages (17-18 years) the differences between the sexes disappear. Although level o f education was found not to affect the likelihood o f females being sexually active, the proportion o f sexually active females was lower 15 among females enrolled in school. On the other hand, males with higher levels o f education were found to be more likely to be sexually experienced than their counterparts with lower levels of education. Furthermore, they observed that incidence o f casual sex was quite common among the adolescents interviewed because both males and females were likely to have one or more casual sexual partners. This was in spite of the fact that both males and females were aware o f the potential negative consequences o f sexual activity including stigma, pregnancy or STIs. At the same time, while 33 per cent o f the females saw sexual activity as having no advantages, information from the focus group interviews showed that female sexual activity is often motivated by economic gain. Finally, the study showed that adolescents do not consider parents as ideal sources of information about sex and related matters although they were included among peejS;2ndJ siblings as having the strongest influences on female reproductive health attitudes. Another study in Nigeria by Olawoye (1995) on adolescent sexuality indicates that parents and the media are blamed for the change in sex behaviour. However, adolescents studied pointed to a situation of overcrowded living conditions that allow children to observe parental sexual intimacy as an influencing condition for early involvement o f adolescents in sexual activity. Respondents were also reported to having multiple sexual partners although many o f them were aware of the dangers o f sexually transmitted infections. A high incidence o f unreported rape and of unplanned pregnancy was noted with adolescents rarely seeking modem medical advice. Gage (1998), in a study that focuses mainly on literature from sub-Saharan Africa, discusses non-marital sexual activity, contraceptive use including condom use. This study explores adolescents’ perceptions of the costs and benefits o f engaging in these bahaviours, their assessment of their susceptibility to the potential consequences o f their actions and the role o f family, peer and dynamic factors in shaping their reproductive decisions. Among other things, 16 Gage shows that cultural values regarding sexuality and gender roles, the power dimensions o f adolescents’ lives, and economic disadvantage exert powerful influences on the decision making process. Besides, decisions to engage in unprotected sex may also be based on insufficient knowledge and distorted judgements o f the risks o f becoming pregnant and acquiring sexually transmitted infections. The study also finds that perceptions about what adolescent peers are doing and what is accepted in their peer groups may be more strongly related to their motivations to engage in sexual activity or risk-taking than perceptions about the opinions o f parents and other family members. Again, motivations for engaging in certain types o f sexual behaviour such as offering sex for money or having intercourse as a result o f force or coercion appear to be more common among teenagers than among adults. Zelaya et. al. (1997), sought to examine gender and social differences in adolescent sexuality and reproduction in Nicaragua through an investigation o f age at first coitus and pregnancy. The study uses data from a 1993 cross-sectional, community-based survey o f a representative sample o f 7,789 households in the municipality o f Leon. In all, 10,867 women o f 15-49 years were studied while more detailed analysis was done on a random sub-sample o f 388 men and 413 women. Among other things, Zelaya et. al. found that although the median age at first coitus was higher for females (17.8 years) than males (16.2 years), women had a lower median age (19.6 years) at delivery o f first child than males (21.2 years). Another significant finding was that one in four persons engaged in coitus before age 15 years, a situation suggesting quite an early sexual activity among the Nicaraguan population. Zelaya et. al. also found that lack of formal education increased the risk o f earlier pregnancy for adolescents by 2.5 times but no increased risk was found to occur between persons o f rural as against urban residence. Furthermore, the study underscores the fact that earlier pregnancy occurred among women who did not live with their biological fathers during their childhood and period of 17 adolescence. On the other hand, living with a stepfather was found to increase the risk o f early coitus and delivery even more. 1.5.3 Adolescent’s Reproductive Health Knowledge, Attitudes and Practices A study on attitudes and practices regarding pre-marital sex among adolescents has been carried out by Adomako-Ampofo (1991). This study was based on a sample o f adolescents drawn from Sunyani, Kumasi, Hohoe, Cape Coast, Aburi and Accra. It examined issues including the age at which young people begin to have a clear understanding o f reproductive related issues, their main sources o f reproductive information, the age at which they become sexually active, the duration of adolescent sexual relationships and the frequency o f sexual intercourse. The study also investigated the knowledge and use o f contraceptives as well as the incidence o f pregnancy and abortion among the adolescents. In the analysis, Adomako-Ampofo found 13.5 years as the average age at which adolescents begin to have a clear understanding of reproductive related issues. It was also shown that only 7 per cent o f the boys and 22 per cent o f the girls that were studied demonstrated that they had an accurate knowledge o f when in the reproductive cycle it is most likely that a woman will become pregnant if she indulged in sex, i.e., the woman's fertile period. Another fundamental finding from Adomako-Ampofo’s study was that 32 per cent o f the adolescents had ever had a sexual intercourse, the youngest age at sexual intercourse being eight years. For some female adolescents therefore, sexual activity could start very early in Ghana. On the use o f birth control, half o f the sexually active adolescents reported to have ever used some form of contraception. Furthermore, among the sexually active adolescent girls who had ever been pregnant, 10 per cent were reported to have ever had an abortion. Adomako-Ampofo further found that among 61 per cent o f adolescent boys who had ever made a girl pregnant, the girl concerned was reported to have had an abortion. This suggests that abortion, although officially 18 outlawed in Ghana, appears to be a practice among some sexually active adolescents in the country. In their study o f adolescent reproductive behaviour in Ghana, Ampofo and Gyepi-Garbrah (1986), explained among other things that the pronatalist beliefs o f the Ghanaian society, often result in early marriage o f females, coupled with a desire to prove their fertility during the first year of marriage. In addition, improvements in health and nutrition conditions, have resulted in the early physical development o f adolescents who are easily drawn into sex and in the process become pregnant. T he study further examined extracts o f antenatal attendance at the Korle-bu Teaching Hospital and reported that 18.4 per cent and 19.3 per cent o f the women who reported to the clinic for services were adolescents o f 15-19 years in 1983 and 1984 respectively. Although the sample used in this study is not representative o f the situation in Ghana, it may suggest quite fairly that the contribution o f adolescents to childbearing in Ghana could be quite substantial. Another study by Agyei and Hill (1997), entitled “Sexual Behaviour, Reproductive Health and Contraceptive Use among Adolescents and Young Adults in Ghana”, found that one in three sexually active but unmarried adolescents had been pregnant at least once, with some variation between the rural and urban areas. For example, the study showed that 47.1 per cent o f adolescents aged 15-19 years in the rural areas had ever been pregnant compared to 23.6 per cent and 36.7 per cent of their counterparts in Accra and peri-urban areas respectively. According to the study therefore, adolescents in rural areas are also twice as likely to become pregnant relative to their friends in Accra. It must however, be noted that the sample used in the study was quite small and hence it would be unrealistic to make valid conclusions based on this study especially o f rural and urban areas in Ghana in addition to making realistic comparisons between rural and urban areas in Ghana as the study sought to do. In another study covering a sample o f 1,356 adolescent residents in high population density areas in Accra, including Chorkor, James Town, Nima, Madina, Maamobi, La, Korle-Gonno and 19 Teshie-Nungua, Kumekpor (1997) observed that almost one in four o f the partners o f teen mothers was a teenager himself, who depended on his parents for his welfare. According to Kumekpor, this is the main reason why most o f them deny responsibility for the pregnancies they have caused. The study however, also shows that adults rather than adolescents were responsible for a relatively higher proportion o f the pregnancies among adolescents. What was clear in the study was that as high as 90 per cent of the adolescent females who had ever become pregnant did not intend or had no plans to become pregnant at the time they engaged in sexual intercourse. While such a situation results in untold hardships in respect o f the maintenance o f children o f these teenage mothers, it suggests that most o f them could be helped to avoid the pregnancy in the first place through education on either outright abstinence or use o f pregnancy preventive methods if they cannot abstain. In a study on knowledge and attitude of adolescents towards reproductive health, carried out at Okponglo, near Legon, Peprah (1998) observed that 80 per cent o f the sexually active adolescents were not using contraceptives. Another finding from this study was that 44 per cent of the adolescents responded that they knew o f a friend who had ever had an abortion. This is interesting, since they themselves scarcely admit openly to have undertaken an abortion but are happy to report on friends they know have had an abortion. On his part, Anarfi (1997) studied 1,147 street children aged between 8 and 19 years who usually congregated or slept around four major markets in Accra. In this study, Anarfi found that 53 per cent o f them had ever had a sexual experience. It was also found that sexually active adolescents had had two or more sexual partners, a situation which obviously has many implications for their reproductive health as far as sexually transmitted infections are concerned. Thus, a little over 7 per cent reported that they had contracted one kind o f STI or another while only 18 per cent of those who had ever contracted an STI went to a hospital for treatment. Self- medication was therefore, found to be quite common among the adolescents with respect to STIs. 20 It was also established that majority of the street children are aware o f the risk o f contracting HIV/AIDS, but they are not doing much to protect themselves from HIV infection. This situation, according to Anarfi, is partly explained by the kind o f misconceptions they have about the disease. Ankomah and Ford (1993), also interviewed 400 never married females aged 18-25 years who live in Cape Coast and found a high level o f premarital sexual experience amongst them. The researchers also found that 86 per cent o f the respondents had engaged in sexual intercourse at least once, the median age at first sexual intercourse being 16 years which is about one year below the average of 16.9 years among women 20-24 years according to the 1993 GDHS. A number o f studies have been done on adolescents elsewhere in other countries in the world. One of these studies done in the United States of America was by Johnson et. al. (1999). This study examines adolescent attitudes, knowledge and values on sexuality and sexually transmitted infections. It covered a sample of 170 students from one rural high school. Hypotheses that formed the basis o f the analysis included the following: a higher value o f an exciting life has a higher attitude toward sexual intercourse scores; a higher knowledge score on sexual intercourse will have a lower attitude toward sexual intercourse scores and a higher knowledge score about AIDS/STIs will have a higher attitude toward condom use scores. Using the Rotter's Social Learning Theory, the study by Johnson et. al., made significant findings. The adolescents were found to have higher accurate knowledge on sexual intercourse and STIs. They also found among the adolescents a positive attitude toward condom use and sexual intercourse with the girl/boy friend. They also found that health values have a positive correlation between an exciting life and the knowledge and attitude about sex and condom use, and a negative correlation towards sex. The findings of minimal impact o f knowledge o f sexual intercourse and STIs on sexual behaviour and attitudes among adolescents, were considered consistent with earlier studies. The limitation o f the study however, was its small sample size, which may not be reliable in making generalizations. 21 1.5.4 Knowledge of HIV/AIDS and STIs Among Adolescents In a survey among 1,553 adolescents, McCombie and Anarfi (1991), sought to collect baseline data to evaluate the impact o f mass media campaign on AIDS and assist the Ministry of Health in planning future AIDS prevention efforts. From the analysis, it was shown that the most important source of information on AIDS so far has been the radio (88.7 per cent), friends (78.7 per cent) and television (75.3 per cent). The researchers also revealed in this study that misconceptions and negative attitudes as far as the disease i s concerned were quite high. One finding in this study was that less than 40 per cent o f the respondents in the study mentioned AIDS as one of the three most dangerous diseases in Ghana while a quarter o f them thought that most young people did not need to worry about the disease. Interestingly too, 46 per cent o f the people interviewed believed that AIDS could be cured. Such thinking creates problems for the nation in its fight against HIV/AIDS since it gives some false hope that may, however, not be realized in the foreseeable future. In another study carried out among adolescent students o f second cycle institutions in the Greater Accra Region, Asamoah-Odei et. al. (1992) probed into the knowledge o f students about HIV/AIDS as a disease. The findings o f this study included a high knowledge o f HIV/AIDS among all segments o f the sample interviewed irrespective o f their sex, educational level and religious affiliation. Knowledge about the modes o f HIV transmission was equally found to be high with 90 per cent o f the students having knowledge about the three main modes o f transmission, namely, sexual contact, intra-venous drug abuse and blood transfusion. Misconceptions about some aspects of the disease were, however, found as in the case of McCombie and Anarfi’s study o f 1991. For example, 20.5 per cent o f the students reported that, AIDS virus could be transmitted by living together with an AIDS patient; 33.1 per cent stated that one could be infected with the virus through an insect bite and 26 per cent were o f the opinion that 22 one could contract the disease by merely kissing someone. These are misconceptions, which HIV/AIDS educational programmes should aim at correcting. On the whole, adolescents in Ghana face a lot o f reproductive health problems most of which stem from their early entry into sexual activity without adequate knowledge about the dangers or risks they subject themselves to. Contraceptive use has been generally low among them and although many of them are aware o f the transmission o f many STIs including HIV/AIDS through sexual activity, many o f them do not abstain from sex or have protected sex. The Health Systems Development Unit (1998) in a study on Adolescent Sexuality and Reproductive Health presents a summary o f research findings o f the Adolescent Sexual and Reproductive Health Programme in addressing adolescent sexual and reproductive health issues and needs in the Northern Province o f the Republic o f South Africa. The study uses interviews and focus group discussions among school-going adolescents in respect o f their knowledge, attitudes, and behaviour on sexual and reproductive health. The views o f teachers and health workers were also sought. Major findings from the South African study included the observation that majority o f the adolescents were sexually active by age 15 years with only a few using preventive methods against sexually transmitted infections and pregnancy. Many o f those who were infected with STIs failed to consult a health worker, a problem that the study attributes to the lack o f reproductive health knowledge, inaccurate or unreliable information, inaccessible reproductive health services and negative perception o f health services. It therefore recommends an adolescent health programme that places emphasis on the significant roles o f the youth, health care workers, teachers and the community in addressing the observed problems. 23 1.5.5 Contraceptive Education and Practice Ruusuvaara (1997) provides a review of family planning literature and concludes that little concern is shown in most o f the articles reviewed about promoting the idea o f more communicative, pleasurable, and egalitarian sexual relations among teenagers because the focus is on avoiding behaviours that are defined as “high risk” . The review also notes that in the United States of America, there is little evidence that contraceptive-based sex education programmes have resulted in reduced sexual activity, diminished teenage-pregnancy rates, or increased effective contraceptive use. Similarly, in Canada, Ruusuvaara notes that the clinical evidence is that existing prevention strategies are not working. However, it is acknowledged that countries where sex education has been accepted, combined with widely spread family planning services and abortion on demand, have the lowest pregnancy and abortion rates in the world. The Centre Francais sur la Population de le Development (CEPED) has carried out a study of adolescent sexuality in five Sahel countries, namely Burkina Faso, The Gambia, Mali, Niger and Senegal (CEPED, 1997). According to this study, strong taboos on discussing sexuality have led to most studies on reproductive health in the Sahel Region in Africa to pay little attention to adolescents, who constitute over one-fifth o f women in the reproductive ages. Consequently, awareness of problems in adolescent reproductive health is limited. The study also found that knowledge of sexually transmitted infections is limited among girls and most o f them do not know that seemingly healthy persons can be HIV seropositive. While friends and the media were identified as the most common sources o f information about sex, health agents, family members and teachers were among the least frequent sources. According to the Sahel Adolescent Sexuality Study, much older respondents agreed that premarital sexual activity has increased in the recent past. From the point o f view o f the adolescents themselves, the disapproving attitude of health workers prevents them from seeking contraception and other needed reproductive health services. This suggests that attitude of service 24 jroviders is quite important in encouraging adolescents to patronize reproductive health services to brestall most of the reproductive health problems that most o f them are faced with, Lunin et. al. (1995) conducted a study on adolescent sexuality in Saint Petersburg, Russia vith the objective o f describing adolescent knowledge, attitudes and behaviour relevant to sexuality and the prevention o f AIDS. The study focused on 185 female and 185 male students vho completed a self-administered questionnaire with a response rate o f 94 per cent. The study ;omes out with interesting findings. For example, only 25 per cent o f females and 34 per cent of nales believed that condoms should be used just once, 38 per cent o f each sex believed that if washed, condoms could be used multiple times. Clearly, the study reported gross misinformation ind unwarranted attitudes towards condom use and hence the need for rigorous sex education orogrammes for Russian youth who, the study found to have a strong support for sex education especially among females, with most respondents seeing sex education as improving sexual pleasure. Accordingly, the study recommends that sex education should be introduced at an early ige so that children could be taught how to reduce the risks o f sexual abuse, HIV infection, and other STIs as well as improve their sexual experiences as responsible adults. Goddard (1995) sought to present the facts in respect o f adolescent sexuality in Nigeria. In bds account, 22 per cent of Nigeria’s 1991 population o f 96 million is aged 10-19 years. According to this study, adolescent women experience first sexual intercourse at the median age o f 16 years with 80 per cent o f them having sexual intercourse by age 20 years. In spite o f this high level of sexual activity among the adolescents in Nigeria, Goddard’s (1995) study shows that 60 per cent of the adolescents that were surveyed were unaware that pregnancy can result from first sexual intercourse. It was also found that nationwide, contraceptive use among women 15-19 years was low with less than 1 per cent using condoms, foaming tablets, or inter-uterine device (IUD). The study also acknowledges that early, unprotected sexual intercourse, leads to many unintended 25 •egnancies, unsafe abortions, and abortion-related complications. It also cites the practice of male genital mutilation (FGM), vesico-vaginal fistula, early female marriage, and low lucational status among women as relevant factors that complicate adolescent development in igeria. Kilboume-Book (1998) reports that worldwide, about 15 million people o f 15-19 years ive birth, as m any a s 4 m illion obtain abortions, and up to 100 million become infected with urable STIs. He further reports that 40 per cent o f all new HIV infections occur among people ged 15-24 years old. The study underscores the fact that adolescents encounter many obstacles to ecuring the reproductive health care services they need. However, the study declares that it takes >rogrammes that satisfy the information and service needs o f adolescents to make a real difference owards solving adolescents reproductive health problems in most countries o f the world. Blanc and Way (1998) provide an overview o f sexual behaviour and contraceptive Knowledge and use among adolescent women across a large number o f developing countries. The results of this study indicate that in almost all o f sub-Saharan Africa and in majority o f countries in other regions, there has been an increase between age at first sexual intercourse and age at first marriage. This suggests that between age at first sexual intercourse and marriage, the adolescent is exposed to so many risks that are associated with non-stable sexual relationships. The study also found that in sub-Saharan Africa, current contraceptive use among sexually active but unmarried teens is higher than among their married counterparts. On the other hand, the reverse is true in Latin America and the Caribbean. Overall, however, adolescents were found to have a high probability of not practising contraception the first time they have sexual intercourse. 26 1.5.6 Adolescent Pregnancy Risks and Consequences Adjei and Ampofo (1986) carried ou t a study to assess some o f th e social and medical problems associated with adolescent pregnancy in Ghana. Among other things, the researchers concluded that more than 50 per cent o f pregnancies to girls under 20 years were reportedly unwanted or unplanned. Similarly, the study found up to 50 per cent o f their partners not wanting the pregnancy. In a survey o f teenage pregnancy and substance abuse in Ablekuma Sub-District o f Accra, carried out by the University o f Ghana Medical School (UGMS) in 1990, the findings were that, 9 per cent of the teenagers were pregnant at the time of the survey and that among them, 27 per cent had complications such as general weakness, bleeding and anaemia. A further 21.5 per cent said that they would opt for termination o f the pregnancy if they found out that they were pregnant. Besides, even though 89 per cent o f the teenagers stated that abortion was bad, 20 per cent were reported to have actually visited a health institution within the sub-district to terminate their pregnancy just prior to the survey. Liskin et. al. (1993) also concluded that young women are especially ill-prepared to respond to reproductive health risks including HIV/AIDS, STIs and unwanted pregnancies. The study showed that in spite o f the fact that adolescent females under 18 years had the least knowledge about family planning and AIDS from such sources as friends or peer groups, partners, health care providers, parents as well as the media, a large proportion of them were sexually active and only a few used contraceptives. The implication is that many o f them could either become pregnant and or contract sexually transmitted infections (STIs) including HIV/AIDS. The dimensions and implications of adolescent fertility in sub-Saharan Africa have been studied in Nigeria, Kenya, Liberia, Sierra Leone and Botswana (Gyepi-Garbrah, 1985). In all these studies, i t i s e mphasised t hat w ith i ncreasing a ge a t m arriage a s a r esult o f increasing education, births to adolescents, which hitherto have been mainly within marriage, are likely to be conceived 27 out-of-wedlock. The findings also point to a situation o f increasing use of illegal abortion in cases of unplanned pregnancies and a growing incidence of sexually transmitted infections due mainly to unprotected sex. Adolescent childbearing therefore, is gradually becoming a major problem in the sub-Saharan African Region. One other focus o f the 1995 PIP study was the reproductive health consequences o f adolescent fertility among adolescents themselves. The major reproductive health consequences o f adolescent fertility were identified in this study to include sexually transmitted infections, health risks of early pregnancy and abortion. There was however, lack o f reliable data on abortion rates in Ghana to make relevant conclusions. Finally, the study recommended certain policy intervention measures that could be taken to curb the incidence o f adolescent sexuality. These include family life education, the setting up o f youth counselling centres, encouraging girls to pursue educational opportunities to the highest levels by substantially subsidising female education and making available family planning services to all adolescents regardless o f marital status. The other policy intervention strategies suggested by the study were improving reproductive health care, involving adolescents in the design o f programmes that seek their welfare and evolving policies to protect young girls from sexual harassment and rape. The study by Nabila and Fayorsey (1996) also found that the high knowledge that adolescents showed about sexually transmitted infections (STIs), HIV/AIDS and possible incidence of pregnancies through sexual activity was not translated into practically wanting to use any contraceptives at their first sexual intercourse. The finding was that 51.2 per cent and 70.1 per cent of the adolescents did not practise any form of contraception during their first sexual intercourse in Kumasi and Accra respectively. It was also found that unplanned teenage pregnancies were largely responsible for most cases o f females that dropped out o f school especially in the peri-urban and rural areas that were covered in the study. This suggests that if adolescents are able to avoid unplanned pregnancy, most of them can complete their education to a 28 large extent. Caution should however, be exercised in the interpretation o f the results from such studies since their representativeness is not easy to determine. According to Kumekpor’s (1997) study, early pregnancy has either interrupted or led to outright cessation o f all socio-economic activities that the girls were engaged in before getting pregnant. For example, 39.3 per cent o f school girls were reported to be in school or awaiting their examination results before they got pregnant and obviously had to abandon their education in order to become mothers. Buvinic (1998) has reviewed results from studies done in Chile, Barbados, Guatemala and Mexico. The review looks at the costs o f adolescent childbearing in the four countries. It addresses questions including whether early childbearing contributes to perpetuating poverty that has been found to plague the Latin American and Caribbean countries. It also answers questions regarding the social and economic impacts o f adolescent childbearing on mothers and children. One important observation in Buvinic’s review is that the disadvantages o f early parenthood may be only transitory and that adolescent mothers may overcome these disadvantages over time. It is argued that the negative outcomes o f early childbearing may be a result of associated conditions, such as women’s poverty and that these mothers may fare poorly even if they delay childbearing. For example, Buvinic shows that little evidence exists in the four countries to demonstrate that early childbearing has negative consequences on the marriage options o f young women. Adolescents who bore children were no more or less likely than adult mothers to be married in the subsequent years according to the Barbados, Chile and Guatemala studies. This was in sharp contrast to the finding in the United States of America that adolescent childbearers spend nearly five times more o f their young adult years as single parents than do later childbearers (Marynard, 29 1996). It was however, illustrated that a high proportion o f the adolescent mothers are more likely than adult mothers to pass on their adolescent motherhood and poor life prospects to their daughters in a hereditary fashion. Furthermore, women who gave birth at early ages were found to be more likely than their counterparts who gave birth at adult ages to earn incomes, hence emphasising the poor economic opportunities that may be the lot o f adolescent mothers compared to adult mothers. Early childbearing among women is therefore likely to contribute to the creation and perpetuation of poverty among women. The Family Care International, a Safe Motherhood Inter-Agency Group (1998) presents findings from the analysis o f data on adolescent sexuality and childbearing. This study focuses on adolescents of 10-19 years and finds that adolescent pregnancy is common in many countries in the world. It also reports that adolescents aged 10-19 years give birth to 15 million children every year worldwide and consequently, subject themselves to considerable health risks during pregnancy and childbirth. Specifically, the study shows, among other things, that pregnant adolescents are two times as likely to die from childbirth than women in their twenties. Again, pregnancy-related complications were identified as the main causes o f death among adolescent girls of 15-19 years worldwide due to the prevalence of early childbearing. Recommendations made in this study include the institution o f programmes for the removal o f legal, regulatory and cultural barriers towards family planning service provision to adolescents in addition to providing appropriate, accurate sexual and reproductive health education; and the development o f sensitive and confidential reproductive health services to protect the sexual and reproductive health of adolescents. In Yaunde, Cameroon, Kamtchouing et. al. (1997), in their adolescent sexuality study in a school environment, covering 574 students made up o f 233 males and 341 females aged 12-19 years in five randomly selected secondary schools, reported undesired pregnancy as a major cause 30 for dropping out o f school among adolescent females in Cameroon. Other findings o f this study include the observation that 41 per cent o f the 53 per cent who reported to be sexually active used a contraceptive method during the most recent sexual encounter. While most o f these contraceptive users (54 per cent) used condoms, close to a third o f them relied on periodic abstinence. Consequently, 24 per cent o f them reported a pregnancy ending in either delivery or abortion. The major source o f information on STIs and contraception was also reported to be the mass media for 54 per cent o f the adolescents, school for 21 per cent, friends for 16 per cent while parents constituted just 9 per cent. The study recommends that sex and family planning education should be offered through youth organizations and schools while teaching parents to discuss sexual topics with their children in order to prevent STIs, unwanted pregnancies, illegal abortions and their medical and psychological consequences. From the literature review, a few research gaps have come out. First, many o f the studies on adolescent sexual and reproductive health used data whose representativeness is quite suspect. Example is the 1998 GDHS, which used data on women o f 15-19 years to describe adolescents who, by definition, include persons of 10-14 years as well. Secondly, each o f the studies focused on certain aspects o f the subject matter o f adolescent sexual and reproductive health leaving out the comprehensive analysis o f the subject. Besides, many o f the studies used actual birth performance i.e., children ever bom to measure adolescent childbearing, thus, clouding the role of abortion in understanding adolescent reproductive health. This is a gap that the current study attempts to fill by focusing on the number o f pregnancies ever had by the adolescents. Finally, the complex relationship that exists between adolescent sexuality, contraceptive use and reproductive health has not been comprehensively studied, a gap the current study attempts to fill in its combination o f both quantitative and qualitative methods o f analysis. 31 1.6 Conceptual Framework Several relationships have been revealed in the available literature on adolescent sexuality and reproductive health. While many studies focus on the determinants o f the phenomenon, others draw attention to the impact o f the problem on the individual adolescent. It is clear however, that one cannot fully discuss the phenomenon o f adolescent sexuality and reproduction without considering some intermediate variables. Yet, on account o f the small age group (12-24), which is the target in this study, the relationships may not be as straight forward as the case may be in a situation where one is considering all women o f reproductive ages. Variables have therefore been selected to suit the purpose and interests of the study. Figure 1.1 illustrates the general interrelationships that are conceived in the conceptual framework with respect to adolescent sexuality, contraceptive use and reproductive health. It shows the main factors that have come up in the literature to influence adolescent childbearing all over the world. In the framework shown in Figure 1.1, the age at first sexual activity and knowledge and use of contraception are presented as the main intermediate variables, which directly influence adolescent childbearing defined as the number of pregnancies ever had by adolescents. Number o f pregnancies is used in order to measure plausibly the proportion o f pregnancy wastage as a result of either abortion or miscarriage. Age at first sexual intercourse is conceived to be influenced by peer pressure, parental control, poverty and migration status of the adolescent. Parental control is defined in terms of whether the adolescent is staying with both parents, one of the parents or none of them. Poverty is also included to find out adolescents who indulge in sex for monetary considerations. On the other hand, migration status (i.e., whether the adolescent is bom in the study area or has moved to stay there from another area outside the district in which the study area is located) is considered in view of the current wave o f migration o f adolescents into the cities and large towns in Ghana. This is 32 conceived to be particularly important in the case of Cape Coast, the regional capital. The study is also interested in finding out how much knowledge adolescents have about the ovulatory cycle as an explanatory factor for the incidence o f adolescent childbearing. Age is an important demographic variable that is put in the model to examine the variation of the adolescents in their contraceptive knowledge and use, sexual behaviour patterns and childbearing experience. Education is also included as a major social variable that influences the fertility behaviour of people. Similarly, religion is u sed t o e xamine t he e xtent t o w hich i t i nfluences a n adolescent’s knowledge and use o f contraception and hence influences her reproductive behaviour. Furthermore, the conceptual framework seeks to explore the consequences o f adolescent sexuality and childbearing. Issues like the contraction of sexually transmitted infections, abortion, sterility and death as well as other fertility associated health concerns are qualitatively considered. These are, however, explored from the point o f view o f the adolescent's knowledge o f these as consequences of adolescent reproduction. The perception of the adolescents (male and female) and adults on the consequences o f adolescent sexuality, contraceptive use and childbearing are also analysed from the focus group discussions that were conducted. The reproductive health issues are thus, more descriptive and exploratory. 33 INDEPENDENT INTERMEDIATE DEPENDENTVARIABLES VARIABLES VARIABLE Figure 1.1. Conceptual Framework Showing the Interrelationships betweenAdolescent Sexuality, Contraceptive Use and Childbearing Current AgeCurrent EducationParental Control (HouseholdLiving Arrangements)Migration StatusMarital StatusPeer PressurePovertyReligion Age at First Sexual ActivityContraceptive Knowledge AdolescentKnowledge of Ovulatory Cycle ChildbearingContraceptive Use at First Sex (Number ofCurrent Contraceptive Use PregnanciesHad') Reproductive Health Concerns including STIs, HIV/AIDS, Number o f abortions ever had Source: Adapted from Bongaarts' (1978) idea o f intermediate variables in understanding fertility. The study acknowledges that adolescent childbearing affects their reproductive health. On account of data limitations however, the reproductive health concerns are not statistically analysed but discussed from the point o f view of the respondents who participated in the focus group discussions. The direct reproductive health effects of adolescent childbearing pertaining to the data collected from adolescents in this study are therefore not analysed quantitatively. 1.7 Hypotheses The following hypotheses have been examined in the study. i) There is an inverse relationship between age at first sexual intercourse and childbearing among adolescents; ii) Contraceptive use is directly related to the level o f education o f the female adolescent; 34 iii) There is an inverse relationship between female adolescent childbearing and contraceptive use at first sex; iv) There is an inverse relationship between education and childbearing among adolescent females; v) There is direct relationship between contraceptive use at first sex and age at first sexual activity among adolescent females. 1.8 Sources of Data and Methodology 1.8.1 Sources of Data The main source of data for the study is primary. The data have been collected in a survey undertaken in Cape Coast, the regional capital, and Mankrong, a rural area in the Central Region. The two study areas were chosen purposely for comparing adolescent sexual and reproductive health situation in an urban vis-a-vis a rural area. Figure 1.2 shows the geographical location of the two study areas. 35 Figure 1.2. Map Showing the Location of the Study Areas, Cape Coast and Mankrong The main instrument for data collection was a structured questionnaire. Respondents were females aged 12-24 years in the two study areas. Secondary sources including the Ghana Demographic and Health Surveys as well as information gathered from focus group discussions were also used. As earlier explained, the study excludes male adolescents because the magnitude of adolescent s exuality and reproductive health problems especially with respect to school drop out, affects female adolescents more than their male counterparts. The focus on adolescent females therefore, is for the sake o f drawing more attention to the critical problems female adolescents face regarding their sexuality and reproductive health. However, it does not limit adolescent sexual and reproductive health problems to females only. 36 1.8.2 Sample Design The selection of the sample for the study in Cape Coast was carried out in two stages. First, Cape Coast was divided into 34 clusters made up o f 15 low-income, 11 middle-income and 8 high- income residential areas to form the sampling units. These were the existing clusters made by the Town and Country Planning Department (TCPD) in the town. The selection o f the clusters for the study was done by simple random sampling, taking into account their geographical location such that clusters that were eventually selected were spread across the Cape Coast township as well as cutting across all socio-economic groups. With this at the background, three clusters were each selected from the high and middle-income residential areas and four from the low-income residential areas. In all, 10 residential clusters were purposively selected for enumeration in Cape Coast. Figure 1.3 shows the residential clusters in Cape Coast from which the sample was generated. At the second stage, the survey aimed at a sample o f 150 female adolescents aged 12-24 years from each of the 10 randomly selected residential clusters such that in Cape Coast 1,500 adolescent respondents were targeted based on the assumption that it will constitute a relatively large sample to facilitate the analyses that were envisaged in the study. The survey was therefore targeted purposively at households that had females of 12-24 years who were subsequently selected for interview with the individual questionnaire. On the contrary, considering the small size o f Mankrong (made up of Mankrong and Mankrong Junction) and to avoid suspicions o f selectivity, total enumeration of all households was undertaken and females aged 12-24 years selected for the individual questionnaire. Overall, the survey achieved a sample o f 1,828 female adolescents (i.e., 1,503 in Cape Coast and 325 at Mankrong) of 12-24 years in both Cape Coast and Mankrong instead o f2,000 as was initially envisaged. 37 Figure 1.3. Map Showing the Residential Clusters in Cape Coast Two sets o f questionnaire were therefore administered: a household questionnaire and an individual questionnaire for eligible females aged 12-24 years. Collecting information from females aged 12-24 years facilitated a comparison between females aged 15-19 years and 20-24 years as well as adolescents below 15 years o f age. Apart from the structured questionnaire, the study conducted focus group discussions among some selected male and female adolescents on one hand, and some male and female adults on the other in respect of their views on the magnitude, causes and impact of adolescent sexuality on the reproductive health o f female adolescents as pertains to each o f the study areas. These sources of data enabled us answer some fundamental sociological questions about adolescent reproduction in these study areas. The focus group discussions were conducted in Fanti, the main local language of the Central Region. The qualitative data collected from the focus group discussions were transcribed 38 and translated from Fanti into English. These data were analysed and used to beef up the explanation for the results obtained from the quantitative analysis as they were considered appropriate. The qualitative data thus, provided additional information to strengthen the results o f the quantitative analysis throughout the thesis. 1.8.3 Techniques of Analysis Simple techniques o f analysis, including cross tabulations, rates, ratios and proportions were employed. Path analysis was intended to be used to examine the direct and indirect effects o f the socio-economic and demographic variables on adolescent sexuality and childbearing. However, due to the small size o f the sample o f female adolescents who had ever become pregnant at the time of the survey, this technique could not be employed. A multiple regression analysis was however, done to attempt a statistical explanation for some of the relationships. The validity o f the hypotheses was determined at 95 per cent level o f confidence. The analysis was also intended to be done separately for each study area to enable comparisons to be made before adding them together to find out the situation for all the study areas in the Central Region. However, this was not possible due to the small sample size for the women who were eligible for this kind o f analysis particularly in Mankrong. The Statistical Package f or S ocial S ciences ( SPSS) w as u sed f or d ata p recessing a nd analysis. 1.9 Definition of Concepts Adolescent: The World Health Organisation defines adolescents as persons aged 10-19 years using chronological age. In this study however, an adolescent is defined to include all persons aged 12-19 years. This is because persons of 10 and 11 years are considered too young to have much knowledge about sexual matters let alone to have a sexual experience. Persons aged 10-11 years who have ever had a sexual experience would be very few and hence would not facilitate any statistically 39 meaningful analysis. The study therefore did not collect any information on persons o f 10 and 11 years old. Persons of 20-24 years are however, included in the analysis not necessarily as adolescents but to allow a comparison between them and the adolescents (12-19 years). The emphasis therefore is more on the adolescents o f 12-19 years although throughout, the study examined all categories of respondents i.e., 12-24 years. In the study, teenagers are used interchangeably with adolescents. On account of the small sample size however, while in most cases adolescents less than 20 years have been compared with the youth o f 20-24 years, in some cases, the analysis puts all the respondents (12-24 years) together. Reproductive Health: The study adopts the International Conference on Population and Development (ICPD) broad definition, which has been adopted by the Ministry o f Health (MOH) o f Ghana together with its implications as spelt out in the Ministry o f Health’s National Reproductive Health ServicePolicy and Standards (Ministry of Health, 2000). Reproductive health is therefore defined as “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity in all matters related to the reproductive system and its functions and processes” (MOH, 2000). The implication here according to the MOH (and which is adopted in this study) is that people are able to have a satisfying and safe sexual life and that they have the capability to reproduce and the freedom to decide if, when and how often to do so. Adolescent Reproductive Health: All aspects of reproductive health that pertain to adolescents who are the subject of investigation in this study. Adolescent Sexuality: According to the Sex Information and Education Council o f the United States (SIECUS) whose main goal is to promote education about sexuality, “the concept of sexuality refers to the totality o f being a person” and includes “ all the aspects o f the human being that relate specifically to being a boy or girl, woman or man, and is an entity subject to lifelong dynamic change’’ (Greenberg et. al., 1 992). For the purpose o f this study however, adolescent sexuality is defined as the knowledge, perception, beliefs and behaviour that are related to sexual relationships 40 between the opposite sexes, at least one o f whom should be an adolescent. In other words, it is used to include any human behaviour that has the potential to affect the reproductive health of the adolescent. Sexually Active Adolescents: This study defines sexually active adolescents broadly to include all adolescents who had ever had sex prior to the survey. These adolescents are therefore conceived in this study to have the potential to engage in sex at some regular interval and hence are identified as being sexually active. Socio-Economic Status: This is used to describe the average standard o f living o f the respondents. The selection of the sample o f adolescents in Cape Coast was based on three socio-economic residential groups defined by the Department o f Town and Country Planning in Cape Coast as High Class, Middle Class and Low Class residential areas. Persons living in these residential areas are therefore classified respectively as high, medium and low socio-economic status groups. High socio­ economic status is used to describe persons in households with ownership o f basic facilities including means of transport, access to potable supply, electricity, radio, television and, in addition, are above the upper poverty line o f 0900,000.00 per capita as defined in the Ghana Living Standards Survey (GLSS) o f 1992. Medium socio-economic status persons are those from households who are above the upper poverty line but do not own all the basic household facilities. The low socio­ economic status persons however, refer to persons conceived to be below the upper poverty line. The limitation of these definitions based on the determination of the poverty line, which is in turn dependent on the non-stable nature of the local currency is however acknowledged by the study. 1-10 Limitations of the Study The first limitation o f this study is its restriction to two settlements in the Central Region i.e., Cape Coast and Mankrong although the goal was to study adolescents in the Central Region. The sample is to this extent, limited in that it is not representative of the population o f adolescents in the 41 entire Central Region. This however, does not seriously affect the findings in the sense that the study seeks to principally compare adolescents in one rural community in the hinterland o f the region to their counterparts in the capital town, Cape Coast. This must be borne in mind in the interpretation of the analysis throughout the study in order to be cautious in making generalizations based on the analysis in the study for the whole o f the Central Region. The second limitation, which has also been emphasised in the analysis, is with respect to the effect of education on the sexual practices o f the adolescents. In the analysis, current education of the adolescents is used instead o f their level o f education at the time o f first sexual activity which could be of more relevance in understanding the variation o f adolescents in terms o f the timing of their first sexual intercourse, their contraceptive use at their first sexual activity, etc., as far as their education is concerned. This is because the survey failed to collect information on the level of education of the respondents at the time of their first sexual experience. Finally, not much rigorous statistical analysis could be done on account o f the small proportion of adolescents who had ever had a sexual activity among the total sample. In view o f this, results of focus group discussions held separately among adolescents and adults at the study areas have been used to give further explanation to some of the key findings. The study in this respect, introduces qualitative dimensions to the analysis thus, providing readers the opportunity to assess the relationships that have been investigated in the study from two perspectives i.e., the use of quantitative and qualitative data collection processes. 1-11 Organisation of the Study The study is organized into nine chapters. The first chapter, which is the introduction, provides the background to the study by defining the research problem, the rationale for the study and the key objectives. In addition, it presents the conceptual framework against the background of 42 available literature on the topic, while describing the sources o f data and the methodology as well as the techniques used in the analysis. Chapter two examines the household characteristics of the study areas emphasising the differences that exist between them. This is to assist in understanding possible sexual behaviour differences between adolescents in the two study areas. On the other hand, Chapter three discusses the socio-economic characteristics of the adolescents in the two study areas with respect to their age, education, marital status, migration status, religious affiliation, ethnicity, occupation and living arrangements. Chapter four also examines adolescent sexuality and marriage with respect to their age at first sexual activity and marriage as well as the reasons behind the timing o f both first sexual activity and marriage. In Chapter five, analysis is done about adolescent childbearing and their reproductive knowledge under which issues including age at first pregnancy and birth were 1 ooked at. 0 ther issues examined in this chapter include adolescent fertility, pregnancy wastage and the timing o f first pregnancy and as to whether it led to the interruption o f their formal education. Adolescents’ knowledge and use of contraception are the subjects o f investigation in Chapter six. Efforts are made to analyse the adolescents’ use of contraceptives at first sexual activity, current contraceptive use as well as the perception of the adolescents regarding the use o f contraceptives by unmarried adolescents and whether the provision of family life education among adolescents could lead to sexual promiscuity among them. Chapter seven focuses attention on the knowledge of adolescents about reproductive health problems as well as the incidence and prevention o f sexually transmitted infections including HIV/ADDS. The study also attempts to examine the possible relationships between adolescent sexuality, contraceptive use and reproductive health in Chapter eight using statistical techniques involving multiple regression analysis. Finally, Chapter nine provides a summary o f the key findings of the study and makes relevant recommendations and conclusions. 43 CHAPTER TWO HOUSEHOLD CHARACTERISTICS OF THE STUDY AREAS 2.1 Introduction The behaviour o f p eople i s c onditioned b y a n umber o f c ircumstances t hat s urround their socialisation process from infancy to adulthood. Sociologically, the household constitutes the primary institution within which the child begins his/her process o f socialisation. The household characteristics and the socio-demographic and economic environment within which the child is brought up is therefore crucial in patterning the behaviour o f the child including her sexuality. This chapter seeks to examine the demographic as well as socio-economic background characteristics o f the households within which the female adolescents have, and are being socialized. This is to find out to what extent the socio-economic conditions in the households have influenced the sexual and reproductive behaviour o f the adolescents. 2.2 Age and Sex Distribution of Household Members A total o f 5,209 persons were listed as household members from 759 and 215 households from Cape Coast and Mankrong respectively. It could be computed from Table 2.1 that the distribution o f household m embers w as 4,083 o r 7 8.4 p er c ent fo r C ape C oast and 1,126 or 21.6 per cent for Mankrong. 44 Table 2.1. Age and Sex Characteristics o f Household Members, Cape Coast and Mankrong Age Cape Coast Mankrong Male Female Male Female Number % Number % Number % Number % 0-4 84 2.1 80 2.0 37 3.3 47 4.2 5-9 128 3.1 123 3.0 85 7.6 57 5.1 10-14 193 4.7 479 11.7 54 4.8 130 11.5 15-19 197 4.8 721 11.7 41 3.6 110 9.820-24 165 4.0 472 11.6 34 3.0 115 10.2 25-29 80 2.0 97 2.4 24 2.1 27 2.4 30-34 41 1.0 94 2.3 17 1.5 40 3.6 35-39 38 0.9 145 3.6 23 2.0 36 3.2 40-44 88 2.2 132 3.2 20 1.8 37 3.3 45-49 100 2.4 141 3.5 23 2.0 23 2.0 50-54 99 2.4 99 2.4 13 1.2 22 1.9 55-59 58 1.4 39 0.9 9 0.8 18 1.6 60-64 46 1.1 39 0.9 8 0.7 11 1.0 65-69 13 0.3 23 0.6 11 1.0 9 0.8 70+ 30 0.7 39 0.9 21 1.9 24 2.1 Total 1,360 33.3 2,723 66.7 420 37.3 706 62.7 Source: Compiled from Cape Coast-Mankrong field survey, Aug-Sept. 1997 For both Cape Coast and Mankrong, 65.9 per cent o f the household members were females compared to 34.1 per cent males (figures not shown in table). This shows a big bias in favour o f females. T his i s also true when we compare Cape Coast to Mankrong. In Cape Coast, 66.7 per cent o f the household members were females as against 33.3 per cent males. Similarly, there were 62.7 per cent females and 37.3 per cent males in Mankrong. This huge bias could be due to the high interest the survey placed on female adolescents. For example, the interviewers were under instruction to interview only households which had at least one female between the ages o f 12 and 24 years. This could probably have been misinterpreted leading to possible non-inclusion o f some male members o f some o f the households. This is more pronounced in the ages 10-24 (Table 2.1). The sample bias for females, however, did not affect the results o f the analysis since the main focus o f the study is on female adolescents of between 12 and 24 years. 45 Again, the to tal household members could be broken down as 26 per cent male and 52.3 per cent female from Cape Coast as against 8.1 per cent male and 13.6 per cent female from Mankrong. It is also clear that the age distribution in both Cape Coast and Mankrong is significantly skewed in favour o f females within the ages 10 to 24. This is clearly in support of the earlier explanation that both enumerators and household heads were perhaps more interested in listing their female members within the ages o f 12 to 24 years upon prior information that they formed the primary basis o f the study. The analysis further shows that more than two-thirds and four-fifth o f the household members in Cape Coast and Mankrong respectively were bom within the Central Region (Table 2.2). It confirms results o f the 1960, 1970 and 1984 Population censuses that the Central Region is not home to many migrants from other regions o f Ghana. The sample is thus, more of a homogeneous group by virtue of where they were bom. Table 2.2. Distribution o f Household Members by Region o f Birth, Cape Coast and Mankrong Region of Birth Study Area ' Cape Coast Mankrong Number % Number % Western 178 4.4 19 1.7 Central 2,866 70.2 993 88.2 Greater Accra 232 5.7 29 2.6 Eastern 173 4.2 34 3.0 Ashanti 233 5.7 16 1.4 Brong Ahafo 81 2.0 1 0.0 Volta 130 3.2 19 1.7 Northern 61 1.5 11 1.0 Upper West 17 0.4 2 0.2 Upper East 14 03 0 0.0 Outside Ghana 96 2.3 2 0.2Not Stated 2 0.1 0 0.0Total 4,083 100.0 1,126 100.0Source: Compiled from Cape Coast-Mankrong field survey, August/September, 1997. 2.3 Household Headship There were in all 974 households comprising 759 or 77.9 per cent in Cape Coast and 215 or 22.1 per cent in Mankrong. The distribution o f the households is presented in Table 2.3. Table 2.3. Distribution o f Households by Sample Area Sample Area Number % 1. Cape Coast University Community 81 8.3 Eyifuwa 85 8.7 Third Ridge Residential Area 53 5.5 North Ola/Ola Estate 75 7.7 1 Aboom 115 11.8 Pedu Estate/Medical Village 85 8.7 Kotokroba Zongo 90 9.2 Brofoyedru/Amoakofuwa 55 5.7 , Idan/Amissa Ekyir/Anaafo 35 3.6 Adisadel Village 85 8.7 Total 759 77.9 2. Mankrong 215 22.1 Source: Compiled from Cape Coast-Mankrong field survey, August/September, 1997. The socialization process o f household members would vary in accordance with whether the household head is a male or female. In most cases, where the head o f household is a female, it may be that she is living as a single parent as a result o f the death or out-migration of her husband. In that case, children in a female-headed household may not b enefit from training from both parents, a s ituation w hich d oes n ot r egularly o ccur i n m ost m ale-headed households. Overall, 62.1 per cent o f the total number o f households that were captured in the survey were headed by males as against 37.9 per cent female-headed households. The female 47 headship rate is slightly higher in Mankrong, almost 42 per cent as against 37 per cent in Cape Coast as shown in Table 2.4. Table 2.4. Sex Composition o f Household Heads in Cape Coast and Mankrong Study Area Male Female TotalN % N % N % Cape Coast 480 63.2 279 36.8 759 77.9Mankrong 125 58.1 90 41.9 215 22.1 Total 605 62.1 369 37.9 974 100Source: Compiled from Cape Coast-Mankrong fie ld survey, August/September, 1997. The relatively higher proportion o f female-headed households in Mankrong may be attributed largely to male out-migration, which came out strongly during the focus group discussions with a section o f the adult members o f the community. This could have negative implications for the upbringing o f children in general and their sexual behaviour in particular. The age distribution o f the heads of household for the two study areas has also been examined. Overall, it shows that about 10 per cent o f all household heads were less than 35 years and about 59 per cent were within the middle ages o f 35-54 years. However, there is some significant variation between Cape Coast and Mankrong. For example, from Table 2.5, it can be observed that Mankrong has almost two times the proportion o f heads o f household who are less than 35 years as in Cape Coast (15.9 per cent versus 8.2 per cent). Thus, while almost 62 per cent o f the heads o f household in Cape Coast are between 35 and 54 years, in Mankrong, the proportion for the same age group is 46 per cent. Conversely, Mankrong has a higher proportion o f older heads o f household (i.e., 55 and over) compared to Cape Coast. In sum, Mankrong has higher proportions of younger and older heads o f household compared to Cape Coast. The reason could be that Mankrong being a rural area which experiences high out-migration tends to lose most o f the young and economically active men to the bigger towns perhaps including Cape Coast, leaving their equally young spouses as heads of household back at home. The larger proportion o f household heads, who are more than 55 48 years in Mankrong in comparison with Cape Coast may be plausibly explained by a relatively higher out-migration o f the economically active young persons in Mankrong for economic opportunities i n t owns c lose b y. T his i s h owever, n ot t o discount the practice in most rural households in Ghana where the oldest person in the household is often regarded as the head of household irrespective o f his or her economic ability to take care o f the household. Table 2.5. Age Distribution o f Heads o f Household in Cape Coast and Mankrong Age Cape Coast Mankrong Total Number % Number % Number % 20-24 11 1.4 10 4.7 21 2.2 25-29 21 2.8 8 3.7 29 3.0 30-34 30 4.0 16 7.5 46 4.7 35-39 75 9.9 25 11.6 100 10.3 40-44 124 16.3 22 10.2 146 15.0 45-49 136 17.9 31 14.4 167 17.0 50-54 135 17.8 22 10.2 157 16.1 55-59 70 9.2 14 6.5 84 8.660-64 71 9.4 15 7.0 86 oo oo65-69 29 3.8 18 8.4 47 4.870+ 54 7.1 32 14.9 86 oo boNot stated 3 0.4 2 0.9 5 0.5Total 759 100 215 100 974 100Source: Compiled from Cape Coast-Mankrong field survey, August/September, 1997. 2.4 Household Size According to Nabila and Fayorsey (1996), most female adolescents in Accra and Kumasi explained, during focus group discussions that they indulge in sexual activities for economic rewards. This is usually the case when parents are either irresponsible or are economically not in a position to take adequate care o f their children. It should however, be noted that reward is an intrinsic part of the traditional mpena (consensual sexual union) relationship in Ghana. In Ghana, where incomes are generally low, one’s economic capability in providing adequately for the needs o f his/her children or dependents largely depends on the size of his/her household. Thus, granted that economic incentives are crucial in understanding 49 an adolescent's sexual behaviour, the size of the household is important in investigating into the sexual practices o f female adolescents. Table 2.6 shows the average household sizes for the study areas. The results in the table are quite contrary to expectation. Overall, the average household size for both Cape Coast and Mankrong is 5.3. However, while Mankrong, which is a rural area has average household size o f 5.2, Cape Coast, an urban settlement has a relatively higher household size of 5.6. The variation could be explained with reference to high rates o f out-migration from Mankrong as compared to Cape Coast. This view was supported by participants in the focus group discussions in Mankrong who attributed this to the limited number o f economic opportunities in that area apart from farming activities, which the youth are no t very much interested in. On the other hand, it could be due to differences arising from the sample selection. It is also significant to note that within Cape Coast itself the variation in household size is such that areas defined as first class residential areas and o f high socio-economic status tend to have a higher household size than the low socio-economic status areas. For example, the average household size for the University Community, Eyifuwa and Third Ridge Residential areas supposed to be first class, high status areas is 6.8 on the average compared to 4.6 for the middle class made up o f North Ola and Ola Estates, Aboom and Pedu Estate/Medical Village. On the other hand, the very low socio-economic status areas made up o f Kotokoraba Zongo, Brofoyedru and Amoakofuwa, Idun/Amissa Ekyir/Anaafo and Adisadel Village have an average household size o f 5.4. 50 Table 2.6. Average Household Size by Study Area Area Average Household sizeUniversity Community 6.7 Eyifuwa 6.9 jThird Ridge Residential Area 6.9North Ola/Ola Estate 5.7 Aboom 4.8Pedu Estate/Medical Village 3.4 Kotokoraba Zongo 4.2Brofoyedru/Amoakofuwa 6.4Idun/Amissa Ekyir/Anaafo 6.9 Adisadel Village 4.1Average for Cape Coast 5.6Mankrong 5.2Total 5.3Source: Compiled from Cape Coast-Mankrong Field Survey, Aug-Sept., 1997. The pattern o f variation in household size among the study areas is probably indicative of the more attraction o f the high-class areas to migrants from family relations in the rural areas and not necessarily due to natural population increase. In any case, the pattern o f household size variation is more progressive in as much as the relatively higher income areas have more household members to cater for compared to areas with relatively lower incomes. In line with this reasoning, it may not be very common to find most adolescents citing monetary gains as the major reason for indulging in sexual activities in Cape Coast and Mankrong. 2.5 Household Conditions Household conditions include the type o f material used in constructing outer walls and roofing, availability o f electricity and piped water in household for its use. Analysis o f these characteristics is in conformity with the status o f Cape Coast as an urban settlement and Mankrong as a rural community. With respect to the type of material used in constructing the outer walls o f household premises, almost 85 per cent o f the household structures were o f cement blocks in Cape Coast 51 while at Mankrong, 76 per cent were o f mud (Table 2 .7a). This is to be expected in most rural settlements in Ghana although in recent times the situation is changing. Table 2.7a. Type o f Main Material Used in Constructing Outer Walls o f Household Premises, Cape Coast and Mankrong Type of Material Cape Coast Mankrong Number % Number % Cement blocks 642 84.6 33 15.3 Mud 111 14.6 163 75.8 Other* 6 0.8 19 8.9 Total 759 100.0 215 100.0 * Includes wood and bricksSource: Compiled from Cape Coast-Mankrong field survey, Aug-Sep., 1997 On the other hand, as shown in Table 2 .7b, while almost 44 per cent o f the household buildings in Cape Coast were roofed with iron sheets, about 17 per cent with concrete and 37 per cent with slate, almost every house in Mankrong was roofed with iron sheets. Thus, although a rural area, Mankrong does not appear to be too poor to use modem roofing materials. Table 2.7b. Type o f Material Used in Roofing Household Premises, Cape Coast and Mankrong Type of Material Cape Coast MankrongNumber % Number %Iron Sheets 332 43.7 198 92.1Concrete 126 16.6 0 0.9Asbestos Sheets 283 37.3 0 0.0Roofing Tiles 7 0.9 0 0.0Other* 11 1.5 17 7.9Total 759 100.0 215 100.0* Includes wood and bricksSource: Compiled from Cape Coast-Mankrong field survey, Aug-Sept., 1997. Investigation was also made into the availability o f electricity and piped water within the household premises. For electricity, only 3 per cent o f the households in Cape Coast did 52 not have supply unlike Mankrong where although the settlement is connected to the national electricity grid, just about half o f the households have electricity connected to their houses for their use (Table 2 .7c). In a similar vein, while 79.1 per cent o f the households in Cape Coast reported that they had pipe borne water within their premises, none o f the households in Mankrong has that facility because the village has not been provided with pipe borne water facility. Mankrong therefore depends mainly on the Ayensu River, which lies close by and a bore hole for its household water requirements. Table 2.7c. Availability o f Electricity Within Household Premises, Cape Coast and Mankrong Study Available Not Available Area Number % Number % Cape Coast 736 97.0 23 3.0 Mankrong 107 49.8 108 50.2 Total 843 86.6 131 13.4Source: compiled from Cape Coast and Mankrong field survey, Aug-Sept. 1997. 2.6 Socio-Economic Status of Households The heads o f household were asked about their ownership o f the structure in which they were housed and certain household facilities which are often considered to be associated with socio-economic status. These included the possession o f radio, television, means o f transport (i.e., vehicle, bicycle, motor cycle, etc.,) and refrigerator. In addition, they were also asked about their own perception o f their socio-economic status and their responses are analysed in this section. Analysis o f the household data suggests that in Cape Coast a large number o f the household premises (54 per cent) were rented while at Mankrong, the large proportion of the premises (43 per cent), were rather owned by the household members themselves. Table 2.8a shows that the proportion o f households living in family houses varied from 21 per cent in Cape Coast to 27 per cent in Mankrong. This is to be expected because in a regional capital 53 such as Cape Coast it is common to find many households in rented premises than in Mankrong, a rural area where a relatively large proportion o f the households are likely to be in their own houses or family houses. It is however, difficult to compare the two study areas in respect of their socio-economic status based on this evidence, since the value o f the structures varies considerably between the two areas. Table 2.8a. Ownership o f Household Premises, Cape Coast and Mankrong OwnershipStatus Cape Coast Mankrong Number % Number % Owned 191 25.2 92 42.8 Rented 410 54.0 62 28.8 Family House 157 20.7 57 26.5Not stated 1 0.1 4 1.9 Total 759 100.0 215 100.0Source: compiled from Cape Coast and Mankrong field survey, Aug-Sept. 1997. Availability o f r adio and t elevision facilities o n the o ther h and, s uggests that female adolescents in Cape Coast may have more access to information including that on their reproductive health than their counterparts at Mankrong. The evidence is that almost 90 per cent and 82 per cent o f households in Cape Coast responded to having a radio and a television respectively. On the contrary, one in three and one in five households respectively owned a radio and television at Mankrong (Table 2.8b). It is however, possible that households which have no radio and, or television at Mankrong could have access to them from other households which have, considering the practice o f sharing that is a common feature in most rural areas in Ghana. 54 Table 2.8b. Availability o f Radio and Television to Households in Cape Coast and Mankrong Study Radio Television Area Available Not Available Available Not Available N % N % N % N % Cape Coast 673 88.7 86 11.3 626 82.5 133 17.5 Mankrong 71 33.0 144 67.0 43 20.0 172 80.0 Total 744 76.4 230 23.6 669 68.7 305 31.3 Source: compiledfrom Cape Coast and Mankrong fie ld survey, Aug-Sept. 1997. Ownership of refrigerating facility and means of transport also shows a similar pattern between Cape Coast and Mankrong. While 71 per cent o f the households in Cape Coast indicated their ownership o f a refrigerator, less than one in 10 households in Mankrong owned one. This is to be expected considering the non-universality o f electricity to all households at Mankrong. Similarly, about 30 per cent of households in Cape Coast reported to own a means of transport (i.e. a vehicle) compared to less than 2 per cent at Mankrong. This suggests that adolescents in Cape Coast generally are from households that are o f a relatively higher socio­ economic status than their counterparts at Mankrong. This is consistent with the Ghana Living Standards Survey (GLSS 4) finding o f a relatively higher rural poverty than the urban in Ghana (Ghana Statistical Service, 2000). Finally, the household heads were asked how they perceived their households with respect to socio-economic status. The results presented in Table 2.8c clearly indicate that in Cape Coast most o f the households perceived themselves as belonging to a medium socio­ economic status (89.3 per cent) while at Mankrong, majority considered themselves to be in the low socio-economic status. It is revealing to observe that none o f the households at Mankrong reported to be in the high socio-economic class. 55 Table 2.8c. Perception o f Household About their Socio-Economic Status, Cape Coast and Mankrong. Study Area Socio-Economic Status High Medium Low Not Stated N % N % N % N % Cape Coast 31 4.1 678 89.3 43 5.7 7 0.9 Mankrong 0 0.0 46 21.4 169 78.6 0 0.0 Total 31 3.2 724 74.3 212 21.8 7 0.7Source: compiled from Cape Coast and Mankrong fie ld survey, Aug-Sept. 1997. The results in Table 2.8c should however, be interpreted with caution since many people particularly in the rural areas associate anything such as incomes that are related to their socio-economic status with a possibility o f payment o f high rates o f taxes. Many people would therefore, wish to be associated with either low or medium status groups in order not to attract high taxation. On the whole, however, one can fairly conclude that the households in Cape Coast are of a relatively higher socio-economic status than those at Mankrong. Access to information on sexuality and reproductive health may equally be more available to adolescents in Cape Coast than at Mankrong considering the more availability o f radio, television and electricity to respondents in Cape Coast, which has the advantage o f being the regional capital as against Mankrong. 56 CHAPTER THREE DEMOGRAPHIC AND SOCIO-ECONOMIC CHARACTERISTICS OF FEMALE ADOLESCENTS 3.1 Introduction The sexual and reproductive health behaviour o f female adolescents may depend on their demographic and socio-economic characteristics. This chapter therefore devotes attention to examining the characteristics o f the adolescents who were interviewed to facilitate a better understanding o f their sexuality and reproductive health behaviour and practices. Comparisons are made in the analysis between Cape Coast as an urban area and Mankrong as a rural settlement. 3.2 Age Distribution of Adolescents The eligible respondents in the individual questionnaire numbered 1,828 in all: 1,503 (82.2 per cent) in Cape Coast and 325 (17.8 per cent) at Mankrong. Although the distribution is biased in terms o f the numbers for Cape Coast, proportionally, considering the population of the two study areas, that is to be expected. The distribution o f the adolescents by age does not show too much variation between adolescents in Cape Coast and their counterparts at Mankrong. The two most remarkable variations, however, can be observed at ages 12 and 20 years where Mankrong recorded a much higher proportion o f respondents than in Cape Coast. It is also clear that the proportion of respondents for ages 21 and above is smaller than that for younger ages especially between 15 and 20 years for either Cape Coast or Mankrong. As is shown in Table 3.1, the drop in proportion from age 20 to 21 is very remarkable especially for Mankrong. This could be a result of most of the respondents who were older than 20 years preferring to report their ages at 20, i.e. at digit zero. It also suggests a situation o f age misreporting among the respondents. 57 At the same time, the explanation that both interviewers and respondents were under the impression that the survey was mainly focusing attention on females aged 15-19 years appears to be valid here. Thus, overall, in Figure 3.1 when ages are grouped, the concentration is clearly within the age group 15-19 years. Table 3.1. Distribution o f Female Adolescents by Single Years o f Age, Cape Coast and Mankrong. Age Cape Coast M ankrong TotalNumber % Number % Number % 12 115 7.7 38 11.7 153 8.413 125 8.3 31 9.5 156 8.514 149 9.9 31 9.5 180 9.815 145 9.6 28 8.6 173 9.516 154 10.3 29 8.9 183 10.017 131 8.7 14 4.3 145 7.918 144 9.6 22 6.8 166 9.119 113 7.5 17 5.2 130 7.120 129 8.6 43 13.2 172 9.421 83 5.5 22 6.8 105 5.722 81 5.4 19 5.9 100 5.523 60 5.0 18 5.6 78 4.324 74 4.9 13 4.0 87 4.8Total 1,503 82.2 325 17.8 1,828 100.0Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. 58 Figure 3.1, Distribution o f Female Adolescents by Age Group, Cape Coast and Mankrong □ Less than 15 M 15-19 □ 20-24 3.3 Education The survey asked questions about the level o f education o f the adolescents in order to investigate into how it varies in respect of their sexual behaviour and the role education could play. Table 3.2 shows the proportion of the adolescents by age and current level o f education for Cape Coast and Mankrong. Overall, adolescents from Cape Coast tend to have relatively higher levels o f education than those at Mankrong. This is supported by the evidence that while just about 6 per cent o f the female adolescents in Cape Coast had no formal education, the corresponding proportion at Mankrong was 16 per cent. Similarly, almost 26 per cent of the respondents in Cape Coast had Secondary/Senior Secondary School (SSS) education compared to about 5 per cent at Mankrong. Besides, at least 5 per cent o f the respondents in Cape Coast had some Post-Secondary or higher education but no adolescent at Mankrong had had any Post-Secondary or higher form o f education. It is however, instructive to observe that most of the adolescent girls had attained a Middle or Junior Secondary School (JSS) level o f education 59 (45 per cent in Cape Coast and almost 50 per cent at Mankrong). The distribution across all the ages tends to follow this pattern. The implication o f the above pattern o f education among the adolescents at this stage is that with most o f the females in both study areas pursuing formal education a t least to the Middle or Junior Secondary School level, there is a relatively higher probability that their knowledge about reproductive health would be enhanced. This is in the light o f the generally observed relationship that exists between level o f education o f women and their fertility behaviour worldwide (Ghana Statistical Service, 1988, 1994 and 1999). Table 3.2. Percentage Distribution o f Female Adolescents by Age and Current Level o f Education, Cape Coast and Mankrong. Age NoEducation Primary Middle/JSS s s s /Secondary Post-Secon­dary *Other TotalNumber CC MK CC MK CC MK CC MK CC CC CC MK12 6.1 5.3 63.5 81.6 30.4 13.1 0.0 0.0 0.0 0.0 115 3813 6.4 9.7 39.2 38.7 53.6 51.6 0.8 0.0 0.0 0.0 125 3114 4.7 9.7 17.4 32.3 73.8 58.1 4.1 0.0 0.0 0.0 149 3115 10.3 3.6 13.1 25.0 60.7 67.8 15.9 3.6 0.0 0.0 145 2816 4.5 13.8 12.3 0.0 42.2 75.9 39.6 10.3 0.0 1.3 154 2917 3.1 14.3 17,6 14.3 42.0 64.3 37.4 7.1 0.0 0.0 131 1418 5.5 22.7 8.3 13.6 35.4 59.1 49.3 4.6 1.4 0.0 144 2219 3.5 47.0 7.1 29.4 40.7 11.8 42.5 11.8 5.3 0.9 113 1720 8.5 32.6 8.5 27.9 34.9 39.5 35.7 0.0 10.9 1.5 129 4321 2.4 13.6 3.6 22.7 30.1 54.6 36.2 9.1 27.7 0.0 83 2222 4.9 15.8 9.9 10.5 34.6 63.2 29.6 10.5 21.0 0.0 81 1923 6.7 16.7 15.0 11.1 46.7 55.5 16.7 16.7 13.3 1.6 60 1824 4.1 15.4 8.1 30.8 44,6 53.8 25.7 0.0 17.6 0.0 74 13<15 5.7 8.0 38.0 53.0 54.5 39.0 1.8 0.0 0.0 0.0 389 10015-19 5.5 18.2 11.8 15.4 44.4 59.1 36.7 7.3 1.2 0.4 687 11020-24 5.7 21.7 8.7 21.7 37.2 50.4 30.2 6.1 17.6 0.7 427 115Total 5.6 16.3 17.7 29.2 45.0 49.9 25.7 4.6 5.5 0.5 1,503 325Notes: C.C. refers to Cape Coast and MK represents Mankrong *Represents ArabicSource: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. 60 3.4 Marital Status An analysis o f sexual behaviour patterns among female adolescents requires some knowledge about their marital status at the time o f the survey considering the different reproductive health risks that each marital status presents. Table 3.3 presents the distribution o fthe adolescents according to age and three main marital status groups i.e. never married, currently married and formerly married comprising the divorced, separated and widowed. Table 3.3. Percentage Distribution of Female Adolescents by Age and Current Marital Status, Cape Coast and Mankrong. Age Never Married CurrentlyMarried Formerly Married Total Number CapeCoast Mankrong CapeCoast Mankrong CapeCoast Mankrong CapeCoast Mankrong 12 100.0 100.0 - - - - 115 38 13 100.0 100.0 - - - - 125 31 14 100.0 100.0 - - - - 149 3115 100.0 100.0 - - - - 145 2816 100.0 100.0 - - - - 154 2917 99.2 100.0 0.8 - - - 131 1418 99.3 95.5 0.7 4.5 - - 144 2219 92.9 47.7 7.1 47.0 5.9 113 1720 88.4 55.8 10.8 34.9 0.8 9.3 129 4321 94.0 68.2 6.0 27.3 - 4.5 83 2222 81.5 57.9 17.3 42.1 1.2 - 81 1923 88.3 38.9 10.0 58.8 1.7 60 1824 62.2 28.6 36.5 64.3 1.3 5.5 74 14Total N 1,423 260 76 58 4 7 1,503 325% 94.7 80.0 5.0 17.9 0.3 2.1 82.2 17.8Note: - means no one belongs to that cellSource: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept. 1997. From the table, it can be observed that a high proportion o f the respondents were never married at the time o f the survey, i.e., 94.7 per cent in Cape Coast as against 80 per cent at Mankrong. Thus, a small proportion o f adolescents (5.0 per cent) from Cape Coast reported to have ever married compared to 20 per cent at Mankrong. However, when the two sets of 61 sample are considered together, 8 per cent o f the female adolescents had ever been in a marital union as at the time o f the survey. The youngest age at which a female adolescent reported to be in a marital union was 18 years for Mankrong and 17 years for Cape Coast. The overall proportion o f adolescents reporting to have formerly been in marriage was also small: 0.3 per cent in Cape Coast and 2.1 per cent at Mankrong. It is clear that majority of the young girls particularly in Cape Coast are not married and hence are likely to have the opportunity to continue their education. However, contrary to the United Nations’ (1989) assertion that adolescent childbearing in developing countries takes place in marriage and hence teenage fertility is not a problem, sexuality and reproductive health may not necessarily depend on whether or not one is married but m ost importantly, the timing o f one's entry into sexual activity and the kinds o f sexual practices one may be involved in. 3.5 Migration Status One emerging problem especially in the cities and urban centres is the issue of streetism, particularly among the adolescent migrants from the rural areas. According to Nabila and Fayorsey (1996), migration o f many teenagers into cities in Ghana in recent times, has led to many young girls being drawn into sexual relations purely as a matter o f survival. Media reports have also indicated a number o f these female street adolescents having frequently fallen prey to sexual abuse by some men. Most often, the victims o f such sexual abuses are migrant adolescents who have no regular homes in their places o f destination and hence are forced to fend for themselves on the streets. On the basis o f the foregoing background, the survey asked questions to find out the migration status o f the respondents in order to assess their sexual behaviours. The analysis presented in Figure 3.2 shows that overall, 46 per cent o f the female adolescent respondents 62 were migrants bom outside their current place o f residence and enumeration. However, there is a variation between those in Cape Coast and Mankrong. For Cape Coast, it is almost 50 per cent migrants and 50 per cent non-migrants compared to only about 30 per cent migrants for Mankrong. This is expected because on account o f Cape Coast's position as the regional capital, an urban area as well as an educational centre, more persons are likely to be attracted there than rural Mankrong especially considering the adolescent population that i s i nvolved and their need for education. Figure 3.2. Distribution o f Female Adolescents by Migration Status, Cape Coast and Mankrong Cape Coast Mankrong Total □ Migrant B Non-Migrant It was also considered relevant to look at the distribution o f the migrants by their region of birth. This is shown in Table 3.4a. One observation that stands out clearly in the table at both Cape Coast and Mankrong is that a high proportion o f the adolescents classified as migrants were bom in the Central Region. This shows that most o f the migrants were short distance ones. It can thus, be inferred that the migrants were geographically homogeneous in that majority were from the Central Region. In other words, there is more intra-regional 63 mobility than inter-regional as far as adolescent migrants in Cape Coast and Mankrong are concerned. It is also observable that limiting the discussion to inter-regional migrants, 13 per cent of them, were from the Greater Accra Region with respect to adolescents in Cape Coast and 11.5 per cent from Eastern Region when one considers adolescents at Mankrong. While Cape Coast recorded a very small p roportion o f i ts a dolescent m igrants t o h ave b een b om i n t he Northern, Upper East and Upper West regions, at Mankrong, none o f the adolescent migrants was reportedly bom in any o f the three Northern regions in the country as is depicted in Table 3.4a. It is also significant to add that for all the respondents, while only 10 per cent o f those in Cape Coast said they were bom in a rural area, almost 87 per cent o f those at Mankrong reported to have been bom in a rural area. Table 3.4a. Distribution o f Adolescent Migrants by Region o f Birth, Cape Coast and Mankrong. Region of Birth Cape Coast Mankrong TotalNumber % Number % Number % Western 60 8.0 9 9.4 69 8.2Central 318 42.7 61 63.5 379 45.1Greater Accra 99 13.3 8 8.3 107 12.7Eastern 56 7.5 11 11.5 67 8.0Ashanti 73 9.8 2 2.1 75 8.9Brong Ahafo 31 4.2 0 0.0 31 3.7Volta 35 4.7 4 4.2 39 4.6Northern 14 1.8 0 0.0 14 1.7Upper West 5 0.7 0 0.0 5 0.7Upper East 5 0.7 0 0.0 5 0.6Outside Ghana 49 6.6 1 1.0 50 5.9Total 745 100.0 96 100.0 841 100.0Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug. - Sept., 1997. Considering the reasons given by the migrant adolescents for moving into their current place of residence, Table 3.4b shows that the main reason cited by majority o f the adolescent migrants for moving to stay at either Cape Coast or Mankrong was the decision o f their parents to move. In fact, in the case o f adolescents in Cape Coast in particular, many o f the 64 respondents indicated they had to move together with their parents who were on job transfer to the regional capital. Clearly, most o f them had no personal control over their decision to migrate. This is not strange considering their young ages and their dependence on their parents for their livelihood. In that case, the problem o f street living by many adolescents in the cities in Ghana appears to be non-existent in the two study areas. The evidence that only 1.3 per cent and 11.0 per cent o f the adolescent migrants in Cape Coast and Mankrong respectively reported that it was their personal decision to voluntarily move to settle where they were enumerated supports this reasoning. Table 3.4b. Distribution o f Adolescent Migrants by Reason for Migrating, Cape Coast and Mankrong. Reason Cape Coast Mankrong Total Number % Number % Number %Employment 37 5.1 9 11.1 46 5.7Move with parents 415 57.0 55 67.9 470 58.1Marital reasons 16 2.2 10 12.4 26 3.2Learn a trade 78 10.7 1 1.2 79 9.8Educational 128 17.5 4 5.0 132 16.3Involuntary 10 1.4 0 0.0 10 1.2Voluntary 10 1.4 1 1.2 11 1.4Work as house-help 18 2.5 0 0.0 18 2.2Other 16 2.2 1 1.2 17 2.1Total 728 100.0 81 100.0 809 100.0Note: 17 adolescents in Cape Coast and 15 in Mankrong did not give any response and were excluded from the analysis in the table.Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept.1997. It is again significant to note that the proportion o f adolescents citing educational reasons for migrating into Cape Coast is quite substantial (17.5 per cent). At the same time, about 11 per cent had gone to stay in Cape Coast in order to learn a trade. It is also important to observe that although not a high proportion (2.5 per cent), some o f the adolescents went to Cape Coast mainly to be employed as house help. It is also noted that 5 per cent o f adolescents in Cape Coast and 11 per cent at Mankrong migrated to seek jobs. Marital reasons including movement with spouse, joining spouse or getting married were relatively more important for 65 adolescent migrants at Mankrong than those in Cape Coast. This is to be expected considering that a relatively higher proportion o f adolescents at Mankrong were currently married compared to those in Cape Coast (See Table 3.3). During focus group discussions, the views o f participants regarding the migration o f adolescents into Accra and the major cities in Ghana in recent times were probed into. Those in Cape Coast felt that Cape Coast, unlike other cities like Accra-Tema and Kumasi, does not receive many adolescent migrants and hence, they did not consider it as a big problem in the regional capital and so did not comment on it. To them, Cape Coast does not have many economic opportunities for adolescents and hence many o f them are not attracted to the town. In Mankrong however, participants made it clear that, there are no jobs apart from subsistence farming especially in coconut cultivation in the village. Hence, after JSS, most adolescents move out o f the village to continue their education a t th e S SS level elsewhere. Those who fail to qualify to continue at the SSS level, on the other hand, jo in their relatives in towns and cities especially in Accra in an attempt to find a job to do. The small number of young persons o f 12-24 years in the village was confirmed during the survey carried out for this study, which had to be extended to Mankrong Nkwanta (Junction), due to the failure to attain quite a substantial number o f adolescents from Mankrong village for the study. The movement o f adolescents, particularly, young girls to the large cities, however, has been found to impact negatively on the reproductive health o f most adolescents who make such movements. For example, an article in the “Junior Graphic” edition o f August 22-28, 2001 recounts the experiences o f many teenage girls who migrate to Accra to become what has come to be known as kayayei i.e., female head porters. In an article authored by Nuhhu-Billa (2001), the “Junior Graphic” narrates the painful initiation rites that many o f the kayayei (who mainly hail from the Northern regions of Ghana) go through before they are given places to lay their heads. 66 According to Nuhhu-Billa (2001), most o f the girls who “are between 14 and 18 years old are forced to engage in commercial sex to make more money, in addition to what they earn from carrying loads during the day to be able to survive”. In the article, Nuhhu-Billa quotes one of the kayayei to have recounted her experience for example that “sleeping outside is not easy, but with time you become used to it since there are hundreds o f us sleeping at the same place. Though some men sometimes take advantage o f us by abusing us sexually, we cannot complain since no one sympathises with us”. Another young lady was also reported to have said that she had been raped three years ago by a driver’s mate and has since not set eyes on him. This lady, according to the “Junior Graphic” report, has a two-year-old daughter in respect o f whom she pays 0 1,000.00 (one thousand cedis) a d a y to a n a n n y to ta k e care of while she goes about her business o f carrying load for money. These kayayei sleep anywhere they find comfortable, notably “Accra Central, Kantamanto, Kinbu and Agbogloshie markets, verandas in front o f stores, kiosks and the mosque at Tudu”. These areas are all open to both human and natural disasters including sexual harassment, flooding and air-borne infections. Most o f the adolescent girls who are going through all forms o f harassment, i.e., both economic and sexual, are lured into the cities often with the collaboration o f their parents. For example, Nuhhu-Billa provides the account o f why one young lady had to come to Accra when she says “a relation returned from Accra to the village with lots o f items as a result o f her work in the city”. Without finding out how that relation managed to get all those items, her parents asked her to also come down to seek a fortune. It is thus clear that some parents themselves push their own children to find fortunes in the cities. Unfortunately, many o f them eventually suffer a lot of socio-economic hardships, get raped and are subjected to all sorts o f sexual relationships as survival strategies, which often lead many o f them into all kinds of reproductive health problems. 67 3.6 Religious Affiliation The survey asked about the religious affiliation o f the adolescents in both study areas in order to assess the possible impact o f their religious beliefs on their sexual behaviour and reproductive health. The breakdown o f their responses is presented in Figure 3.3. Figure 3.3. Distribution of Female Adolescents by Religious Affiliation, Cape Coast and Mankrong O'* . s / $ & & • & 8P O-Or □ Cape Coast 11 Mankrong □ Total Figure 3.3 shows that a large proportion o f the adolescents were Protestants. At least, one in every three adolescents in either study area was a protestant by religion. On the other hand, one in four persons, was a Pentecostal/Charismatic in either area. The major differences between the two study areas are in respect o f the Catholic and Islamic religions. For example, while Catholics were made up of at least 26 per cent of the adolescent respondents in Cape Coast, they were just about 9 per cent at Mankrong. On the contrary, there were almost 20 per 68 cent Muslims at Mankrong as against just about 8 per cent in Cape Coast. Similarly, there was a higher proportion o f adolescents who said they had no religion at Mankrong than their counterparts in Cape Coast. Traditional religion and other religions including the Jehovah's Witnesses and Hinduism were not found to show significant differences between the two study areas. In sum, with the exception o f the Catholic and Islamic religions, the respondents from the two study areas may not vary much in their sexual and reproductive health beliefs and practices as may be determined by their religious affiliation. 3.7 Ethnicity Respondents from the two study areas do not appear to vary remarkably from each other on the basis o f ethnicity. This is because 62 per cent o f all the respondents in the two study areas reported to be Fanti. This is to be expected since the Central Region is mainly peopled by the Fanti as the indigenous ethnic group. In the same vein, the proportion o f Ewe, Ga-Adangbe, Guan and Non-Ghanaian in each area is not widely different. For example, in Cape Coast, these ethnic groups were found to be 4.9 per cent (Ewe), 2.6 per cent (Ga- Adangbe) and only 0.5 per cent (Guan) with 0.9 per cent as non-Ghanaians (Figure 3.4). The corresponding figures for Mankrong were 6.1 per cent (Ewe), 2.8 per cent (Ga-Adangbe), 0.9 percent ( Guan) and 0.3 p e r cent non-Ghanaians. The only exceptions though are the Other Akan and Mole-Dagbani categories. Here, while the proportion o f Other Akan (made up of Asante, Kwahu, Akwapim, Akyem, Brong) in Cape Coast is nearly twice that in Mankrong, the reverse is the case regarding the Mole-Dagbani ethnic group. 69 Figure 3.4. Distribution of Female Adolescents by Ethnicity, Cape Coast and Mankrong 75 62.3 62.2 50 - 4 R * From Figure 3.6, it is observable that just about one in three adolescents in Cape Coast were living with both parents compared to two in five adolescents in Mankrong. This suggests that supervision o f a dolescents b y b oth p arents i n r espect o f t heir s exual b ehaviour m ay b e relatively higher at Mankrong than in Cape Coast. Thus, in Cape Coast almost one-third of the female adolescents said they were living with other relations and not necessarily their parents. In Mankrong, a little over one in four adolescents were living with their mothers only compared to about one in six in Cape Coast. While the relatively high proportion of female adolescents living with their mother only could be due to possible out-migration o f their fathers, one should not lose sight of the concern that this situation poses. This is because mothers may find it more difficult to supervise or guide their daughters than perhaps the case 74 may be with “fathers only” . This is especially problematic when mothers are unable to provide for the needs o f these young girls, a factor which was raised by many adolescents in the focus group discussions for taking to sexual activity. In general, over 60 per cent o f the female adolescents were living with either one of their parents or with persons who were not their direct parents. The sexual behaviour o f these young girls would, to some extent, depend on the kind o f relationship that may exist between them and the persons with whom they live, i.e., the concern the latter may show in the former in providing the right information and a right sense o f direction as well as supervision. 75 CHAPTER FOUR ADOLESCENT SEXUALITY AND MARRIAGE 4.1 Introduction From the literature, it has been demonstrated that sexual activity is not limited to marital unions but also among unmarried adolescents. In this chapter, the analysis looks at the variation among the female adolescents with respect to their sexual behaviour taking key background characteristics including current age and level o f education into consideration. The reasons provided by the adolescents for commencing sexual activity and deciding to marry at the time they did are also examined. 4.2 Age at First Sex The reproductive health status o f female adolescents anywhere in the world may be assessed and understood first and foremost on the basis o f how early in age they start practising sexual intercourse and circumstances within which sexual activity takes place. Against this background, this chapter examines female adolescents in Cape Coast and Mankrong as to what proportion o f the sample interviewed had ever had a sexual intercourse and the environment or conditions within which their first sexual activity may be understood. The survey asked the adolescents whether or not they had ever had a sexual intercourse. Table 4.1 shows that overall, the percentage o f adolescents who have ever had sex increases with age. This is very much clear when the results o f age groups less than 15, 15-19 and 20-24 years are considered. There is, however, a relatively higher percentage o f adolescents who had ever had sex in Mankrong than in Cape Coast, except at ages less than 15 years where the result is to the contrary. Table 4.1 also shows that for the entire sample, 37 per cent responded to have ever had a sexual intercourse. On the other hand, the 1993 GDHS recorded that by the age o f 15, only 12 per cent o f the respondents had ever had sexual relations. However, by age 20, more than 76 85 per cent o f women had had sexual intercourse at the time o f the survey for the whole of Ghana. The corresponding figure for women aged 20-24 years in the 1998 GDHS was 81.4 per cent. The 1998 GDHS also found that 62 per cent o f women o f 15-19 years had never had sex. This finding compares with that o f the current study where 69 per cent o f women o f age 15-19 years (not shown in table) had never had sex at the time o f the survey. When we compare Cape Coast to Mankrong, it is clear from Table 4.1 that the proportion having ever had sexual intercourse in Mankrong was higher than Cape Coast by almost 10 per cent. One other significant deduction that could be made from Table 4.1 is that by the age of 20 years, 70.3 per cent (i.e., 121 o f 172 adolescents) o f the adolescents had ever had sex for the entire sample compared to more than 85.0 per cent and 81.4 per cent reported in the 1993 and 1998 GDHS respectively for Ghana. When this is broken down by study area, we have 79.1 per cent (i.e., 34 o f 43 adolescents) in Mankrong as against 67.4 per cent (i.e., 87 o f 129 adolescents) in Cape Coast (See Table 3.1a for number o f adolescents aged 20 years). These figures suggest some improvement on the 1993 recorded figure for Ghana. It is also noted that the distribution o f the respondents who have ever had sexual intercourse is higher at age 20 compared to all other ages. This is valid for both Cape Coast and Mankrong although the figure for Mankrong is much higher (almost 24 per cent) compared to Cape Coast (16 per cent). This is difficult to explain as the proportions relate to the total number of persons who had ever had sexual intercourse for the entire sample as against the proportion within individual ages. In spite of this, caution should be exercised when interpreting the results o f the analysis considering the possibility o f some adolescents not providing accurate information concerning whether or not they have ever had sex. 77 Table 4.1. Percentage Distribution of Adolescents Who Have Ever Had Sex by Current Age, Cape Coast and Mankrong CurrentAge Cape Coast Mankrong Total N % N % N % 12 1 0.2 0 0.0 1 0.1 13 2 0.4 0 0.0 2 0.314 3 0.6 0 0.0 3 0.415 9 1.7 5 3.5 14 2.116 23 4.3 6 4.2 29 4.3 17 40 7.5 7 4.9 47 6.918 70 13.0 13 9.1 83 12.2 19 61 11.4 13 9.1 74 10.9 20 87 16.2 34 23.8 121 17.8 21 63 11.7 19 13.3 82 12.1 22 59 11.0 16 11.2 75 11.123 51 9.5 17 11.8 68 10.024 67 12.5 13 9.1 80 11.8 <15 6 6.0 0 0.0 6 1.2 15-19 203 29.5 44 40.0 247 31.0 20-24 327 76.6 99 86.1 426 78.6 Total 536 35.7 143 44.0 679 37.1Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. Again, it is o f interest to know the timing o f first sexual activity among the adolescents. This has been investigated and the results are presented in terms o f percentage distribution and mean age at first sexual intercourse in Figure 4.1 and Table 4.2 respectively. In Figure 4.1, the overall situation is that the peak period o f first sexual activity is within 15-19 years where as high as 82 per cent o f the respondents indicated having had sexual intercourse for the first time. This is true for both study areas although it is higher in Mankrong (91 per cent) as against Cape Coast (80.0 per cent). Conversely, a relatively higher proportion of the respondents in Cape Coast had their first sexual experience earlier (i.e. less than 15 years) than their counterparts in Mankrong: almost 10 per cent had sexual intercourse at ages less than 15 years in Cape Coast compared to 4 per cent in Mankrong. There is also a higher proportion o f respondents in Cape Coast initiating sexual intercourse at ages between 20 and 24 years (10 per cent) than in Mankrong (a little less than 5 per cent). This suggests that 78 although a relatively higher proportion o f adolescents start sexual intercourse earlier in Cape Coast than a t Mankrong, a t later ages (20-24 years) one is likely to find a relatively higher proportion o f them yet to be initiated into sex in Cape Coast compared to their counterparts in Mankrong. The foregoing finding may be due to the fact that in the rural areas, adolescents who attain age 20 years without continuing their education, are likely to have entered into sexual activity during their teen ages. Thus, during 20-24 years, a relatively fewer proportion of adolescents may be left yet to initiate sexual intercourse compared to adolescents in urban areas, most o f whom may continue their formal education. One could however, not rule out possible misstatement o f age at first sex by the adolescents in Mankrong. Figure 4.1 .Percentage Distribution of Female Adolescents Ever Having Sex by Age at First Sex, Cape Coast and Mankrong Cape Coast Mankrong Total □ Less than 15 H 15-19 □ 20-24 In terms o f the mean age at first sexual intercourse, adolescents from the two study areas are almost the same. While in Cape Coast, the mean age at first sexual intercourse was found to be 16.9 years it was 17.0 years in Mankrong. This finding is quite inconsistent with 79 the expected situation o f a much lower mean age at first sexual intercourse in a rural area than in an urban area in Ghana. However, as was observed earlier, the distribution o f the respondents who had ever had sex suggests that a relatively higher proportion o f the respondents were sexually active by age 19 years in Mankrong (rural area) compared to Cape Coast (urban area). On the other hand, the pattern o f distribution o f mean age at first sex (Table 4.2) suggests a possible decline in the age at first sexual intercourse for respondents in both study areas. This is because as current age increases, mean age at first sex also generally increases. So far, none o f th e respondents aged 12-14 years in Mankrong responded to have ever had sexual intercourse. In spite o f this, it was found that mean age at first sex for respondents aged 15-19 years was 16.1 as against 17.4 for those aged 20-24 in Mankrong. Similarly, it was 12.8, 15.9 and 17.5 respectively for respondents aged 12-14, 15-19 and 20-24 years in Cape Coast. It is to be noted that the proportion o f persons becoming sexually active by the age o f 20 years in Ghana declined from 59 per cent in 1993 to 38 per cent in 1998 according to the 1993 and 1998 GDHS results (Ghana Statistical Service, 1994 and 1999). It is also underscored that the incidence o f first sexual intercourse among adolescents who had ever had a sexual experience (measured by age at first sex) in the current study does not vary from the 1993 and 1998 GDHS records o f 16.9 years in 1993 and 17.5 years in 1998 among females aged 20-24 years (the GDHS did not compute age at first sex for adolescents o f 15-19 years due to small sample size). Early age at sex exposes adolescents to reproductive health risks. In Zimbabwe, it has been explained that, “stereotyped sexual norms and peer pressure encourage young males to prove their manhood and enhance their social status by having sex” (Kim et. al., 2001:11). On the other hand, according to Kim et. al., young women are socialized to be submissive and not to discuss sex, thereby leaving them to be unable to refuse sex or insist on condom use. 80 Table 4.2. Mean Age at First Sex o f Adolescents by Current Age, Cape Coastand Mankrong CurrentAge Mean age at first sex - Cape Coast Mean age at first sex - Mankrong 12 12.0 13 12.014 13.7 - 15 14.8 15.0 16 15.0 15.5 17 15.2 16.318 15.9 16.5 19 16.7 16.320 16.9 17.2 21 17.4 17.7 22 18.2 17.123 17.4 18.3 24 18.3 17.1 <15 12.815-19 15.9 16.1 20-24 17.5 17.4 Total 16.9 17.0Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. 4.2.1 Living Arrangements The study again sought to find out whether or not the persons the adolescent girls live with have any impact on the timing o f their first sexual intercourse. This was defined broadly as the living arrangements within which the female adolescent found herself. The results, which portray little variation among the adolescents from each study area, are presented in Table 4.3. 81 Table 4.3. Mean Age at First Sex of Adolescents by Living Arrangements, Cape Coastand Mankrong LivingArrangements Mean age (years) Total Cape Coast Mankrong Both parents 16.9 16.9 16.9Father only 17.2 17.4 17.3 Mother only 17.1 17.1 17.1Other relations 16.5 17.0 16.8 Husband 16.6 16.7 16.7 Friends 17.0 17.0 Unrelated person 17.2 17.8 17.5Alone 18.0* 16.8 17.4Note: * Represented by one person.Source: Computed from Adolescent Survey, Cape Coast and Mankrong,Aug.-Sept., 1997. Overall, there is some evidence to show that adolescents who were currently living with their husbands had the lowest mean age at first sexual intercourse. This is quite understandable since many o f them might have been married at very young ages, which m eant th a t sexual activity could start shortly after marriage even in situations where the young girl had never had sex before marriage. It is also possible that having indulged in early sex that might have resulted in a pregnancy, marriage became a matter o f course for most o f the young girls. It is plausible that while early marriage could lead to early age at a woman’s introduction to sexual intercourse, at the same time, early age at sexual intercourse could result in early marriage. This category o f adolescents (i.e. those living with their husbands) was followed, with respect to the timing o f first sexual intercourse, by those who were living with other relations apart from their own parents and then those who were living with both parents. What is quite unexpected is that adolescents who were living with both parents (where one expects to have better care and supervision) tended to show not much difference in respect o f their mean age at first sexual intercourse as their counterparts who lived with either parent or with unrelated persons. Accordingly, comparing Cape Coast to Mankrong, Table 4.3 does not reveal any 82 uniform pattern that could suggest any possible impact o f these living arrangements on the timing of first sexual intercourse o f the adolescent females. It should be noted that in rural areas in Ghana, communal type o f living is still common but was not investigated in this study and therefore, does not appear in the table. 4.2.2 Level of Education Another dimension o f one's initiation into sexual intercourse is the role o f education. In line with this, the respondents were asked about their level o f education at the time o f the survey. Table 4.4 presents the mean age at first sexual intercourse by the level o f education o f the adolescents. The results in the table show that overall, there is a general increasing age at first sexual intercourse with increasing level o f education. The only inconsistency is with respect to those with primary level education who unexpectedly showed a lower mean age at first sex compared to their counterparts with no education. This i s th e result o f th e general lack o f consistency in the pattern exhibited by respondents in Mankrong. It is also possible that adolescents with primary education might have dropped out o f school as a result of a pregnancy and hence could have a lower age at first sexual intercourse compared to their counterparts with no education. In Mankrong, results in Table 4.4 do not show much difference in mean age at first sexual intercourse by level o f education. In fact, the relatively highest mean age at first sexual intercourse (17.3 years) was recorded among adolescents with no education. This however, may have arisen due to possible misstatement o f ages at first sexual intercourse by the adolescents with no formal education. On the other hand, when the discussion is limited to respondents in Cape Coast, it is clear that mean age at first s exual intercourse increases with rising level o f education o f the female adolescent. 83 The foregoing finding appears to be inconsistent with Zabin and Kiragu s (1998) observation that seems to suggest that schooling may actually encourage sexual onset especially as it tends to remove young people from the supervision o f traditional caretakers. This explanation may tend to support the finding in this study o f a lower mean age at first sex among adolescents o f primary level education relative to others with no education. This could also be the result that adolescents with primary level education might have dropped out of school at the primary level due to a pregnancy. In spite o f the foregoing observation, no conclusion on the impact o f education on a female's entry into sexual intercourse can be made. This is because, the stated levels o f education relate to the time o f the survey and not the timing o f the first sexual experience when their level o f education might have been different and perhaps lower. This is a limitation the current study is unable to address. Table 4.4. Mean Age at First Sex o f Adolescents by Current Level o f Education,Cape Coast and Mankrong EducationLevel Mean age (years) Mean age (years) Total Cape Coast Mankrong No Education 16.0 17.3 16.7Primary 16.1 17.1 16.4 Arabic 17.0 17.0JSS/Middle 16.8 16.9 16.8 SSS/Secondary 17.1 17.1 17.1Post-Sec./Higher 18.0 18.0Not stated* 18.0 18.0Note: *Represented by one person.Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept.,1997. 4.2.3 Religion Religion has been found to have a big influence on family life and reproductive health and rights (Familusi, 1999). In the current study however, the effect o f religious differences among the respondents on their mean age at sexual intercourse does not show any marked 84 consistency between the two study areas (Table 4.5). There are no significant differences in the mean age at sexual intercourse between the various religious groups. With the exception of the traditional religious group, which has the lowest mean age at sex o f 14.5 years (possibly due to the very small number o f women involved), the rest fall between 16 and 17 years with the Muslim, Catholic and Pentecostal/Charismatic being slightly higher compared to the Protestant and No religion categories. Considering the two study areas, the highest mean age at first sex was recorded among the Muslim group at Mankrong but in Cape Coast, i t was among the Catholic and Pentecostal/Charismatic groups followed closely by the Muslim group. Table 4.5. Mean Age at First Sex o f Adolescents by Religious Affiliation, Cape Coast and Mankrong Religion Mean Age at First Sex Total Cape Coast Mankrong N Years N Years N Years Catholic 125 17.0 11 16.7 136 17.0 Protestant 196 16.8 53 16.8 249 16.8Pentecostal/Charismatic 142 17.0 39 17.1 181 17.0 Muslim 55 16.9 25 17.6 80 17.1 Traditional 2 14.5 - 2 14.5No religion 9 16.1 12 17.1 21 16.7 Other 7 15.9 3 15.3 10 15.7Total 536 16.9 143 17.0 679 16.9Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. 4.2.4 Frequency of Sexual Intercourse The study again looked at the frequency o f sexual intercourse during the month preceding the survey among respondents who had ever had sexual intercourse (Tables 4.6a and 85 Table 4.6a. Percentage Distribution o f Sexually Active Adolescents by Frequency o f Sexual Activity During the Past Month, Cape Coast and Mankrong Study Area Number of Times Notstated Total0 1 2 3 4 5+ Cape Coast N 300 79 81 23 18 6 29 536 % 59.2 15.6 16.0 4.5 3.5 1.2 5.4 100.0 Mankrong N 72 15 28 12 8 6 2 143 % 51.1 10.6 19.8 8.5 5.7 4.3 1.4 100.0Note: The category "Not Stated” was not included in the computation in the table. Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. Table 4.6b. Percentage Distribution of Sexually Active Adolescents by Age and Frequency of Sexual Activity During the Past Month, Cape Coast and Mankrong Age Number of Times Having Sex in the Past MonthCape Coast 0 1-2 3+ Not Stated TotalN %<15 25.0 50.0 25.0 33.3 6 100.015-19 64.8 29.5 5.7 4.9 203 100.020-24 56.1 32.6 11.3 5.2 327 100.0Total % N 59.2 31.5 9.3 5.4 100.0300 160 47 29 536 100.0 Mankrong<15 - -15-19 48.8 25.6 25.6 2.3 44 100.020-24 52.0 32.7 15.3 1.0 99 100.0Total % N 51.1 30.5 18.4 1.4 100.072 43 26 2 143 100.0Note: The category "Not Stated" was not included in the computation in the table. Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. Results from Tables 4.6a and b show that more than half o f them stated that they did not have any sexual intercourse during the month preceding the survey (59.2 per cent in Cape Coast and 51.1 per cent in Mankrong - Table 4.6a). The table should however be interpreted with caution. This is because, the question as to the number o f times a person (who is not 86 married) has had sexual intercourse is so sensitive that the tendency to understate the frequency may be quite high particularly among females in Ghana. Nonetheless, after controlling for current age o f adolescents (Table 4.6b), one key observation is that the frequency o f sexual intercourse during th e p ast month preceding the survey, appears to be relatively higher among younger adolescents compared to the older ones. For example, as is shown in the table, in Cape Coast, while the proportion having sexual intercourse three times or more is 25.0 per cent among adolescents o f less than 15 years, the corresponding figures for adolescents aged 15-19 years and 20-24 years are 5.7 per cent and 11.3 per cent respectively. Similarly, 25.6 per cent o f adolescents o f 15-19 years in Mankrong reported to have had three times or more o f sexual intercourse during the month preceding the survey compared to 15.3 per cent among their counterparts o f 20-24 years. This finding is not a positive one as far as the sexual and reproductive health o f these adolescents is concerned especially as majority o f them are not married and may be engaging in unprotected sex with its negative health consequences. It was also considered important in this study to examine the distribution o f the sexually active female adolescents in relation to the age at which they had their first sexual intercourse and whether or not they had identifiable sexual partners at the time o f the survey. A look at the results presented in Table 4.7 shows that, on the whole, about three-quarters of the respondents who had ever had sex had identifiable sexual partners at the time o f the survey in either Cape Coast (75 per cent) or Mankrong (74.1 per cent). In either case, there is a high concentration within ages 15-19 suggesting that most o f the sexually active females had their first sexual experience within ages 15-19 years irrespective of whether or not they had sexual partners at the time o f the survey. Overall, while less than 10 per cent o f the first sexual activities took place at either less than 15 years or 20 years and over, as high as 82 per cent o f first sex occurred within 15-19 years. The table also 87 suggests that the timing o f adolescents’ first sexual activity does not appear to influence their decision to have sexual partners. This is due to the observation in Table 4.7 that among either respondents with sexual partners or those without, the pattern o f distribution according to the timing of first sexual intercourse does not vary remarkably from each other. Table 4.7. Percentage Distribution o f Sexually Active Female Adolescents With or Without Regular Sexual Partners by Age at First Sex, Cape Coast and Mankrong Study Area Age at First Sex Total <15 15-19 20+ % Number Has Sexual Partner 9.5 79.7 10.8 75.4 404Cape Coast Mankrong 4.7 89.6 5.7 74.1 106 Has no Sexual PartnerCape Coast 11.4 80.9 7.6 24.6 132 Mankrong 2.7 94.5 2.7 25.9 37 Total 8.7 82.3 9.0 100.0 679 Source: Computed from Adolescent Survey, Cape Coast and Mankrong Aug.-Sept., 1997. In sum, the adolescents in Cape Coast do not vary substantially from their counterparts in Mankrong with respect to age at first sexual intercourse. It is also clear that most o f them tend to have their first sexual experience within the ages o f 15-19 years i.e., the peak o f their adolescence irrespective o f their place o f residence, whether rural or urban. 4.3 Characteristics of Sexual Partners of Adolescents Sexual activity always takes place between two persons. It is therefore not enough to limit the discussions to the characteristics o f the sexually active females without examining the demographic and socio-economic characteristics o f their sexual partners. This section accordingly examines possible age variations between the females and their s exual p artners both currently and at the time o f their first sexual encounter. Characteristics such as the 88 education and occupation o f male sexual partners are also looked at as a way o f finding out what category o f males are usually picked as sexual partners o f the female adolescents. The age distribution o f all male sexual partners o f the female adolescents is shown in Figure 4.2. The Figure shows that overall, a higher proportion o f adolescents in Mankrong (31.4 per cent) reported having sexual partners compared to Cape Coast (25.9 per cent). With respect to age, more than two-thirds o f the male sexual partners were reported by their female partners to be within 20-29 years. About 16 per cent o f them were 30 years and over while just about 10 per cent were less than 20 years. For the two study areas, the results do not vary much. However, it is observable that a slightly higher proportion o f the male sexual partners at Mankrong were older (21.6 per cent o f them aged 30 years and over) relative to their counterparts in Cape Coast (15.1 per cent). Correspondingly, a s lightly s mailer p roportion o f t he m ale p artners i n M ankrong were aged less than 20 years compared to that o f Cape Coast. What is quite conclusive on the pattern o f age distribution o f the male sexual partners is that for both Cape Coast and Mankrong, only a small proportion o f them could be said to be the age mates o f their female counterparts. This becomes clearer when the table is considered against the background o f the ages o f the female adolescents under study ranging between 12 and 24 years. 89 Figure 4.2. Percentage Distribution of Sexual Partners o f Female Adolescents by Age, Cape Coast and Mankrong 1 0 0 . 0 - i — •— -------------------------- Cape Coast Mankrong Total □ Less than 20 Hi 20-29 □ 30+ □ Total The mean age distribution o f the male sexual partners is again examined in comparison with the female to compute the actual variations at each o f the ages. These are presented in Table 4.8. It should be noted that the information in the table was collected from females who reported as having sexual partners at the time o f the survey. Again, the results pertain to whom the adolescent female considered as her sexual partner irrespective o f whether or not she had more than one sexual partner. As is shown in the table, about 96 per cent o f the female adolescents who had ever had a sexual experience stated that they had regular sexual partners at the time o f the survey. What is not clear is whether the remaining 4 per cent had sexual partners who were not considered to be their regular sexual partners or did not have a sexual partner at all at the time o f the survey. 90 Table 4.8. Mean Age o f Regular Sexual Partner by Current Age o f Female Adolescent and Marital Status Cape Coast, Mankrong Current Mean Age o f Sexual Partner Variation (Years)Age of Cape Coast Mankrong Cape Coast MankrongWoman Never Ever Never Ever Never Ever Never EverMarried Married Married Married Married Married Married Married 12 13.0 - - 1.0 -13 13.0 - - 0.0 -14 20.0 - 6.0 - - - 15 17.0 18.3 2.0 3.316 20.1 22.0 - 4.1 6.0 17 20.6 14.0 19.7 - 3.6 -3.0 2.7 -18 23.1 20.0 21.5 19.0 5.1 -2.0 3.5 1.019 22.9 24.3 23.0 25.0 3.9 5.3 4.0 6.0 20 24.4 26.3 27.8 26.3 4.4 6.3 7.8 6.321 24.6 28.0 22.1 24.4 3.6 7.0 1.1 3.4 22 26.1 26.8 26.0 27.3 4.1 4.8 4.0 5.3 23 27.3 28.6 26.0 31.2 4.3 5.6 3.0 8.2 24 28.1 31.2 25.0 31.3 4.1 7.2 1.0 7.3<20 21.8 22.7 20.7 24.320-24 25.7 28.8 25.3 28.2 -Note: Variation fo r the age groups was not computed due to their wide ranges Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. Information presented in Table 4.8 shows that contrary to observations that most sexual partners of adolescents are themselves adolescents, the age variations between the females and their male sexual partners clearly paint a different picture. It is observable that at a 11 a ges except age 13 years where the only female in Cape Coast had the same age as her sexual partner, the male sexual partners were on the average older than their female counterparts. Another exception is found at age 17 years where one married adolescent in Mankrong is reported to have a sexual partner three years younger than she was at the time o f the survey. This is not normal especially in a situation where the male sexual partner was reported to be as young as 14 years. This suggests a clear case o f misstatement o f age by the adolescent female. The results as presented here should therefore be interpreted with caution. This is because the analysis is based on information provided by female adolescents who may not be accurate on 91 the a ge o f t heir s exual p artners. M isstatement o f a ge o f s exual p artners m ay t herefore have been introduced into the analysis. Comparing ever married female adolescents to the never married, Table 4.8 shows that the sexual partners o f the never married are relatively younger than those o f their ever married counterparts except at ages 17 and 18 in Cape Coast and, 18 and 20 in Mankrong where the finding is to the contrary. This finding is clearer when the two age groups o f less than 20 years and 20-24 years are considered. Here, the mean age o f male sexual partners o f ever married female adolescents is higher than that o f the never married in either Cape Coast or Mankrong. There is however, no clear consistency when the situation in Cape Coast is compared to Mankrong. While for some o f the ages, the variation in age between the females and their male partners was much higher in Cape Coast than in Mankrong, a look at other ages reveals a contrary picture. This is the case for either the ever married or never married. At this stage, the study is equally interested in the difference in age between the sexual partners at the time o f the first sexual encounter o f the female adolescent who was interviewed in the survey. In the survey, the female respondents were asked about how old they were at the time of their first sexual encounter as well as the age o f the persons with whom they had their first sexual activity. In Table 4.9, the ages o f the females at the time of the first sexual activity are compared with those of their male partners at the time o f the act. The comparison shows that on the whole, the females were younger than their male partners at their first sexual activity. This suggests that the females have first sex earlier than the males. Agyei and Epema (1990) have documented a similar finding in Kampala, Uganda where mean age at first sexual intercourse was recorded as 15.7 years for adolescent males and 15.3 years for adolescent females. This is however, contrary to Nabila and Fayorsey’s (1996) finding in their study o f Accra and Kumasi, Ghana with mean age at first sexual intercourse recorded as 15.8 years among the adolescent 92 males and 16.2 years among the females. For all the ages reported for first sexual activity to have taken place, the variation clearly depicts the males to be older, the only exception being at 21 years where the females were older than their male counterparts by just 0.3 years on the average. From the foregoing analysis, it could be concluded quite plausibly that females become sexually active earlier than males contrary to findings recorded by researchers such as Nabila and Fayorsey (1996). However, it could be argued that when males are interviewed about differences in ages between them and the female counterparts, they might appear to have been younger than their female partners at the time o f the first sexual act. What perhaps is the situation is that anyone who engages in first sexual activity does so normally with a more sexually experienced partner who is more likely to be older. Therefore, the question as to which of the sexes enters into sexual activity earlier would depend on whether the responses are from males or females. What may bring out a more reliable result would be when males are interviewed separately from the females and their ages at first sex compared instead o f resorting to the differences in age between persons who are sexual partners both o f whom may not necessarily have their first sexual activity with each other at the same time. 93 Table 4.9. Mean Age o f Sexual Partner by Age o f Adolescent at First Sex, Cape Coast,MankrongAge at First Sex Mean Age of Sexual Partner Variation (Years) Cape Coast Mankrong Cape Coast Mankrong8 16.0 - 8.0 -11 12.0 - 1.0 -12 14.5 16.5 2.5 4.5 13 17.9 15.0 4.9 2.014 19.1 17.3 5.1 3.3 15 19.4 19.6 4.4 4.616 21.1 22.1 5.1 6.117 21.2 20.7 4.2 3.718 23.2 22.0 5.2 4.0 19 23.0 21.4 4.0 2.420 23.7 27.7 3.7 7.7 21 20.7 27.0 -0.3 6.0 22 26.6 4.623 25.7 - 2.7Source: Computed from Adolescent Survey, Cape Coast and Mankrong, Aug.-Sept., 1997. It came out in the focus group discussions that often it is wealthy adults who entice unsuspecting adolescents into sexual relationships. The adolescents were therefore asked in the focus group discussions as to their views about the appropriateness o f such sexual relationships between an adult and an adolescent. In the discussions they confirmed that this goes on i.e., “sugar daddy” (older man and adolescent girls) and “sugar mammy” (older ladies and adolescent boys). Some o f the adults in Mankrong were o f the opinion that the discussions should not focus on only perceived wealthy older men taking advantage o f young girls but also older rich women who engage young boys for sex. They consequently, condemned this kind of sexual relationship as, to them, it breeds disrespect between the young and the elderly. However, one male adult narrated a case where he said he overheard two young females having a chat, alluding to the fact that it is better to have a sexual relationship with an elderly person than with their age mates. According to him, the young women argued that “the young m en a re n ot r esponsible e nough a nd a re a lways r eady t o d eceive a nd d isappoint you 94 while the elderly people will not on account of their experience”. From his view, it therefore appears that it depends on the impression the individual has but it is not a good thing. The views expressed by most of the adolescents were interesting. Although they all invariably did not endorse such a relationship, it appeared that for financial reasons, they seemed to accept it if they were in a position to benefit from the relationship. For example, while some said “it is not good to have someone as old as your father as your sexual partner as it could lead to disrespect”, there were others who argued that it could be acceptable because “some of them take good care o f the girls” . For financial reasons therefore, some o f them felt it was good. As one girl remarked, “it is good to have a rich man who will take care o f the girl and her boyfriend as well” . What this intervention implied was that some o f the girls who befriend older men have boyfriends o f their age and hence, get funds from the older persons to take care of themselves and their young sexual partners as well. Clearly, this is a recipe for the spread o f sexually transmitted infections (STIs) including HIV/AIDS. The bottom line o f all this as further explained by one adolescent female in Mankrong is money. She argued that if the young girl is well catered for at home by her parents or is working and is able to cater for her basic needs, most young girls would not give in to sexual offers by older persons. However, some give in to sexual offers by older persons. Some o f them even were o f the opinion that some parents do encourage their young girls to go into such sexual relationships with rich but older men because o f financial gains they (the parents) hope to benefit from rather than seeking the interest o f their daughters. The female respondents were also asked about the educational levels o f their male sexual partners. Figure 4.3 shows that more than half o f the respondents in Mankrong indicated that their sexual partners had Middle/JSS level of education compared to about one-third in Cape Coast in this educational category. It is also observed that, close to a third o f the sexual 95 partners in Cape Coast were reported to have SSS/Secondary level o f education, while a quarter o f them had post-secondary and higher level o f education. On the other hand, comparatively smaller proportions o f the sexual partners in Mankrong had higher than Middle/JSS level o f education. This shows that the sexual partners in Cape Coast were o f a relatively higher level o f education than their counterparts at Mankrong. This is quite understandable considering the status o f Cape Coast as an urban area while Mankrong is a rural settlement. Another observation is that for both areas the proportion of the sexual partners who had no education was quite small indicating that most adolescent females tend to have sexual partners from among persons who have had some form o f formal education. 75.0 50.0