University of Ghana  http://ugspace.ug.edu.gh
ESTIMATION OF FOOD PORTION SIZES: A COMPARISON BETWEEN THE 
USE OF HOUSEHOLD MEASURES AND A PHOTOGRAPHIC FOOD ATLAS IN A 
RURAL POPULATION IN GHANA 
 
 
BY 
 
KATE OPOKU 
(10247782) 
 
 
THIS DISSERTATION IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON 
IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF THE 
MSc DIETETICS DEGREE 
 
 
 
JULY 2017
 
 
University of Ghana  http://ugspace.ug.edu.gh
DECLARATION 
I, Kate Opoku, author of this dissertation, do hereby declare that it was done by me under the 
supervision of Dr. Gladys Peprah-Boateng. All references cited in this work have been duly 
acknowledged. 
 
 
 
 
Sign……………………………………….  Date………………………………….. 
Kate Opoku 
(Student) 
 
 
Sign:  Date:   February 5TH, 2018 
Dr. Gladys Peprah-Boateng 
(Supervisor) 
 
 
i 
 
University of Ghana  http://ugspace.ug.edu.gh
ABSTRACT 
Background: Dietary assessment can be defined as “a comprehensive evaluation of a person's 
food intake”. It serves as the basis for the diagnosis and management of chronic diseases. A 
number of dietary assessment techniques exist but the appropriateness of use depends on the 
purpose and food eating habits. Some examples of dietary methods are weighed food records, 
estimated food records and household measures, to name but a few. Household measures such as 
cups and spoons are most commonly used in Ghana. A photographic atlas of commonly 
consumed carbohydrate Ghanaian foods was developed by Peprah Boateng to aid in dietary 
assessment and counselling. The tool was validated among a cross section of participants in 
Accra and this study is focused on testing this tool in the rural area to confirm its use across the 
nation.  
Aim: To validate a photographic food atlas of commonly consumed carbohydrates based foods 
with estimated food portion sizes using household measures in Asesewa, a rural community in 
Ghana 
Method: A cross sectional study involving a three day dietary intake from individuals through 
face-to-face interviews using a structured pre-tested questionnaire to all males and females 
within the age group of eighteen (18) years and seventy (70) years who visited the Nutrition and 
Research center and the Asesewa Regional Hospital within the time period of collecting the data. 
Data collection was divided into three (3) main parts: socio-demographic and anthropometry, 24-
hour recall and portion size estimation. 
Results: A significantly greater proportion, 63.7% (121), of the participants underestimated the 
carbohydrate food (P <0.0001). Twenty-nine per cent (55) of participants overestimated whilst 
ii 
 
University of Ghana  http://ugspace.ug.edu.gh
only 5.3% correctly estimated the portion size of carbohydrate food. There was no significant 
difference in gender estimations of carbohydrate foods (P= 0.295). However, a higher proportion 
of female participants were able to estimate correctly (7.9 %) compared to males (3.0 %). No 
significant differences (P>0.05) were found between the number of participants who 
underestimated or overestimated carbohydrate among gender. Among the age range of 
participants, higher proportion (69.0%) within the age range of 39 to 49 years underestimated 
portion size of fufu followed by those within 50 to 60 years and 17 to 27 years indicated by 
62.5% and 61.7% respectively. Similarly, higher proportion (75.0%) within the age range of 39 
to 49 years underestimated portion size of banku compared to the other age ranges. Boiled yam 
on the other hand was estimated correctly by higher proportion (66.7%) of participants within the 
age range of 39-49 years followed by those within 17 to 27 years and 28 to 38 years as indicated 
by 47.1% and 42.9% respectively. Boiled rice was highly overestimated by the participants with 
the 61 to 70-year range and 17 to 27-year range (65.2%). Underestimation of sugar was seen 
among a statistically significant percentage (P=0.041; 93.8%) of participants within the 39 to 49-
year range. 
Conclusion: Portion size estimation of these carbohydrate foods using the photographic food 
atlas and the household measures showed an overall correct estimation of 5.3%, under and 
overestimation of 63.7% and 28.9% respectively. Gender showed no significant effect on portion 
size estimation although more females estimated carbohydrate food correctly compared to the 
males. The effects of the different BMI categories on participants’ ability to correctly estimate or 
overestimate were not significant. However, significant effect of BMI on underestimation of 
portion size in food was observed.  
 
iii 
 
University of Ghana  http://ugspace.ug.edu.gh
DEDICATION 
I thank God for providing me the strength and resources to undertake this project. I also wish to 
show appreciation to my husband, Frank Kommey, and my lovely daughters, Earlene and Kayla, 
who have been a backbone to me. I also wish to express my gratitude to my parents, Mr. and 
Mrs. Opoku, and my brothers, Michael, Derek and Daniel, for their affection and support at 
every step in my life. I cannot forget the prayers and support of my favourite aunt, Ps. Evelyn 
Amihere, who has been a pillar to the family for all these years and my mother-in-law, Janet 
Arthur. 
 
 
 
 
 
 
 
 
 
 
 
iv 
 
University of Ghana  http://ugspace.ug.edu.gh
ACKNOWLEDGMENT 
To my heavenly father for the grace, favour and strength throughout the study course. 
To my supervisor Dr. Gladys Peprah Boateng for her counsel and guidance during the course of 
the work. 
To my head of department Dr. Matilda Asante for her great help. 
I would like to show my gratitude to Prof. Matilda Steiner-Asiedu and Prof. Saalia for taking 
time out of their busy schedule to assist me in my data analysis and encouraging me throughout 
the presentation of my thesis. 
I am immensely grateful to Janet Carboo for the great help in my data collection and Richard 
Ansong with the data analysis. 
Special thanks to my family and friends especially Belinda Damali, Naa Nyarkoa Gyan-Mantey, 
Lydia Anang, Zahari Abu-Naa Aminbo and Chris Borto Afful for their encouragement and 
support. 
 
  
 
 
 
 
v 
 
University of Ghana  http://ugspace.ug.edu.gh
Table of contents 
DECLARATION ........................................................................................................................................... i 
ABSTRACT .................................................................................................................................................. ii 
DEDICATION ............................................................................................................................................. iv 
ACKNOWLEDGMENT ............................................................................................................................... v 
APPENDICES ............................................................................................................................................. ix 
LIST OF ABBREVIATIONS AND ACRONYMS .................................................................................... xii 
CHAPTER 1 ................................................................................................................................................. 1 
1.0 INTRODUCTION .............................................................................................................................. 1 
1.1 Problem Statement .............................................................................................................................. 2 
1.2 Justification ......................................................................................................................................... 4 
1.3 Aim ..................................................................................................................................................... 4 
2.0 LITERATURE REVIEW ....................................................................................................................... 6 
2.1 Introduction ......................................................................................................................................... 6 
2.2 Prospective Dietary Methods .............................................................................................................. 7 
2.2.1 Portion Size Measurement Aids ................................................................................................... 7 
2.2.2 Food Records ............................................................................................................................... 8 
2.3 Retrospective Dietary Methods ......................................................................................................... 10 
2.3.1 24 hour Recall ............................................................................................................................ 10 
2.3.2 Diet Histories ............................................................................................................................. 11 
2.3.3 Food Frequency Questionnaire ................................................................................................. 11 
2.4 Portion sizes and dietary assessment ................................................................................................ 12 
vi 
 
University of Ghana  http://ugspace.ug.edu.gh
2.4.1 Portion size ................................................................................................................................ 12 
2.4.2 Portion Distortion ...................................................................................................................... 13 
2.4.3 Portion Size Measurement Aids ................................................................................................. 15 
2.5 Household Measures and dietary assessment ................................................................................... 17 
2.6 Photographic Atlas ............................................................................................................................ 18 
3.0 METHODOLOGY ............................................................................................................................... 20 
3.1 Study Design ..................................................................................................................................... 20 
3.2 Study Site .......................................................................................................................................... 20 
3.3 Study Population ............................................................................................................................... 20 
3.3.1 Inclusion and Exclusion criteria ................................................................................................ 21 
3.3.2 Sample size determination ......................................................................................................... 21 
3.4 Pre-testing of questionnaire .............................................................................................................. 22 
3.5 Data collection .................................................................................................................................. 22 
3.5.1 Subject recruitment .................................................................................................................... 23 
3.5.2 Data collection material ............................................................................................................ 23 
3.5.3 Anthropometric measurements .................................................................................................. 23 
3.5.4 Dietary Assessment .................................................................................................................... 24 
3.6 Data Analyses ................................................................................................................................... 25 
3.7 Ethical approval ................................................................................................................................ 25 
4.0 RESULTS ............................................................................................................................................. 27 
4.1 Background profile of participants ................................................................................................... 27 
4.2 Anthropometric assessment of Respondents ..................................................................................... 30 
4.3 Commonly consumed carbohydrate foods ........................................................................................ 32 
4.4 Participants’ estimation of common carbohydrate foods .................................................................. 34 
vii 
 
University of Ghana  http://ugspace.ug.edu.gh
4.4.1 Overall carbohydrate estimations by gender .............................................................................. 34 
4.4.2 Carbohydrate foods estimation by BMI categories .................................................................... 35 
4.4.3 Age assessment of carbohydrate foods estimation ..................................................................... 36 
4.4.4 Estimation of the types of carbohydrate foods identified .......................................................... 37 
4.4.5 Gender related estimation of the different types of carbohydrate foods .................................... 38 
4.4.6 BMI related estimation of the different types of carbohydrate foods ........................................ 39 
4.4.7 BMI and gender evaluation of the different types of carbohydrate foods.................................. 41 
4.4.8 Evaluation of the different carbohydrate foods by age ranges ................................................... 45 
4.4.9 Evaluation of the different carbohydrate foods by age ranges and gender ................................ 46 
5.0 DISCUSSION ....................................................................................................................................... 50 
5.1 General profile of the participants .................................................................................................... 50 
5.1.1 Demographics ............................................................................................................................ 50 
5.1.2 Anthropometric measurement .................................................................................................... 50 
5.2 Common carbohydrates foods consumed by the participants ........................................................... 51 
5.3 Overall estimation of carbohydrate food .......................................................................................... 51 
5.3.1 Gender and carbohydrate food estimations ................................................................................ 53 
5.3.2 Body Size and estimation of portion size ................................................................................... 54 
5.3.3 Age and food portion size estimation......................................................................................... 56 
6.1 Conclusion ........................................................................................................................................ 57 
6.2 Limitations ........................................................................................................................................ 58 
6.3 Recommendations ............................................................................................................................. 58 
REFERENCES ........................................................................................................................................... 59 
APPENDIX 1 .......................................................................................................................................... 66 
APPENDIX II ......................................................................................................................................... 68 
APPENDIX III ........................................................................................................................................ 69 
APPENDIX IV........................................................................................................................................ 70 
viii 
 
University of Ghana  http://ugspace.ug.edu.gh
APPENDICES 
Appendix I – Participants Information Form 
Appendix II – Participants Informed Consent Form 
Appendix III – Socio-Demographic Information  
Appendix IV – 3-Day Diet History Form 
 
 
 
 
 
 
 
 
 
 
 
 
 
ix 
 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF TABLES 
Table 4.1: Socio demographic characteristics of respondents 
Table 4.2: Distribution of mean height, weight and BMI among respondents 
Table 4.3: Commonly consumed carbohydrate-based foods 
Table 4.4: BMI-specific estimations of the different types of carbohydrate foods 
Table 4.5: BMI and Gender-specific estimations of the different types of carbohydrate foods 
Table 4.6: Gender and BMI association with carbohydrate food estimations 
Table 4.7: Age-specific estimations of the different kinds of carbohydrate foods 
Table 4.8: Age and gender-specific estimations of the different types of carbohydrate foods 
 
 
 
 
 
 
 
 
x 
 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF FIGURES 
Fig 4.1: Educational level of participants 
Fig. 4.2: Average monthly income of participants 
Fig 4.3: BMI distribution of participants 
Fig 4.4: BMI distribution based on gender 
Fig 4.5: BMI distribution of participants based on age groups 
Fig 4.6: Overall carbohydrate estimation by gender 
Fig 4.7: Participants’ BMI and overall portion size estimation 
Fig 4.8: Age range of participants and portion size estimation 
Fig 4.9: Estimation of common carbohydrate foods 
Fig 4.10: Gender-specific estimations of the different types of carbohydrate foods 
Fig 4.11: Age and gender-specific estimations of the different types of carbohydrate foods 
 
 
 
 
 
xi 
 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF ABBREVIATIONS AND ACRONYMS 
 
NCDs: Non Communicable Diseases 
WHO: World Health Organization 
FFQ: Food Frequency Questionnaire 
GSS: Ghana Statistical Service 
PSMAs: Portion Size Measuring Aids 
SSA: Sub-Saharan Africa 
2D PSMAs: Two dimensional Portion Size Measuring Aids 
3D PSMAs: Three dimensional Portion Size Measuring Aids 
USDA: United States Department of Agriculture 
UMKD: Upper Manya Krobo District 
BMI: Body Mass Index 
CE: Correct estimation 
UE: Under estimation 
OE: Over estimation 
FAO: Food and Agriculture Organization 
DES: Daily Energy Supply
xii 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 1 
1.0 INTRODUCTION 
Dietary assessment comprises of food production and supply at the national level, food purchases 
at the household level and food consumption at the individual level (Thompson & Subar, 2013). 
Dietary assessment can also be defined as a complete evaluation of a person's food intake 
reported either subjectively or objectively (Shim, Oh, & Chang Kim, 2014). Subjective 
assessment is done by use of open-ended surveys such as dietary recalls or records or by use of 
close-ended surveys such as food frequency questionnaire; whereas objective assessment is done 
by weighed food records (Shim, Oh, & Chang Kim, 2014). 
 
According to Boushey et al., 2009 chronic diseases can be prevented and managed when 
adequate information is available on dietary intake. Non communicable diseases (NCDs) such as 
hypertension, diabetes, cardiovascular diseases, cancers and obesity contribute highly to illness, 
disability and deaths in Ghana (Ofori-Asenso & Garcia, 2016; Bosu, 2013; Ministry of Health, 
2011). These have been linked to unhealthy dietary patterns and physical inactivity of an 
individual (WHO, 2014). According to (Wojtusiak et al., 2011) adequate dietary intake is the 
basis of good health  and measuring dietary intake and nutritional status is one of the most 
efficient and informative means of understanding the health status of a community. Peprah-
Boateng (2014) observed that household measures served as the most widely used portion size 
estimation tools by dietitian and nutritionists in Ghana. These measures however do not usually 
provide uniform and comparable results, thus hampering the development of dietary policies and 
1 
 
University of Ghana  http://ugspace.ug.edu.gh
guidelines. However, a photographic food atlas is a useful tool for field estimation of dietary 
intake with minimal  average errors in portion size estimation (Harris-Fry et al., 2015). 
 
Several methods have been developed for the assessment of dietary intake, but the appropriate 
method to use depends on the purpose for the assessment, which include measuring nutrients, 
foods or eating habits (Wrieden et al.,2003). The dietary methods are weighed food records, 
estimated food records, 24-hour recall, multiple pass recall, food frequency questionnaire (FFQ) 
and household surveys (Wrieden et al.,2003). All of these methods have their advantages and 
disadvantages and are affected by portion size estimation in one form or the other.  
 
Different jurisdictions employ different portion size estimation tools to assess dietary intake. 
Some of the tools employed are food atlas, household measures and the weighing of food. The 
use of any one measure is usually influenced by convenience and habit.  The use of food weights 
generally provides more uniform results than other portion estimation tools. It is commonly used 
in developed countries such as the United Kingdom. However it is not widely used in Africa 
because of its cost implications, low levels of literacy and the fact that it is not a traditional way 
of estimating portion sizes.  
1.1 Problem Statement 
Increased rate of diet related non communicable diseases have been greatly linked to dietary 
habits and patterns of individuals. The World Health Organization (WHO) estimated that in 2008  
majority (60%) of all deaths were as a result of NCDs and over 80% of these deaths were 
reported in low and middle income countries (World Health Organization, 2010). 
 
2 
 
University of Ghana  http://ugspace.ug.edu.gh
Developing countries including Ghana are currently recording diet-related diseases at a rather 
fast pace (Bosu, 2010). In 2007 urban Accra reported 35% obesity and 28% overweight among 
adult women (Hill, 2007).  The Ghana Statistical Service (GSS) also stated in 2008 that one-third 
of Ghanaian women were overweight or obese. Additionally, high rates of hypertension ranging 
between 20% and 50% has been documented (Bosu, 2010).  
 
To help curb the high rates of NCDs, good nutrition throughout life should be prioritized by all 
given the correlation between diet and health. Nonetheless, most people choose foods for reasons 
other than their nourishing values (Anti, 2008). In addressing dietary behaviours both quantity 
and quality should be addressed. Quality can be ensured easily, but the amount eaten requires 
estimation and this continues to pose a challenge to dietitians and nutritionist. It must be 
emphasied that accurate and consistent measurement of dietary intake and patterns of eating 
behaviour is important when evaluating the effectiveness of public health interventions to 
improve diet and reduce obesity and other non communicable diet related diseases.  
 
There has been an increase in the portion sizes of foods sold commercially,with  evidence from a 
limited number of studies suggesting that the availability of larger portions is correlated  with an 
increase in total caloric intake, which could lead to weight gain  and an increased risk of non 
communicable diseases (World Health Organization, 2010). The increase in obesity and NCDs in 
Ghana calls for an effective portion size measurement aid (PSMA) that would be understood by 
all irrespective of gender to help control quantities of food consumed by the general public and 
inform dietary guidelines. A photographic atlas of commonly consumed carbohydrate Ghanaian 
foods was developed by Peprah Boateng to aid in dietary assessment and counseling. Although it 
3 
 
University of Ghana  http://ugspace.ug.edu.gh
does not cover all the food groups it may be a good tool that targeted what Ghanaians eat in large 
quantites to start with. This tool was validated among a cross-section of Ghanaians  in Accra and 
it was reported in the study that more than half (52.45%) of the participants were able to make 
correct estimations based on the use of the household measures and food photographic atlas. This 
current study sought to  validate  the carbohydrate photo atlas tool by Peprah Boateng (2014) in a 
rural setting in Ghana to confirm its use across the nation. It is expected that the findings will 
make it possible for Ghana to have  empirical data on carbohydrate eating habits nation wide to 
inform policy.   
 
1.2 Justification 
The outcomes of this study will provide information on the use of the food photographic atlas as 
a tool in assessing portion sizes in a rural setting in Ghana. This study will also provide 
information that will help determine whether household measures or the photographic food atlas 
will be a better portion size measurement aid (PSMA) in a rural area. 
 
Furthermore the study will seek to provide further empirical evidence and guidance in the 
development of a national dietary assessment framework which will help curb diet-related 
diseases. 
1.3 Aim 
The main aim of this study was to validate a photographic food atlas of commonly consumed 
carbohydrates based foods with estimated food portion sizes using household measures in 
Asesewa, a rural community in Ghana 
4 
 
University of Ghana  http://ugspace.ug.edu.gh
1.3.1 Specific Objectives 
1. To determine the background characteristics of the study participants. 
2. To determine participant’s portion sizes of commonly consumed carbohydrate foods 
from dietary recall using some common Ghanaian household measures. 
3. To assess participants’ ability to accurately pick out recalled portion sizes of 
carbohydrate foods from portion sizes depicted in the photographic food atlas. 
4. To determine if  age, gender and BMI significantly affected participants’ ability to 
accurately select portion sizes from the photographic food atlas. 
 
 
 
 
 
 
 
 
 
 
5 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 2 
2.0 LITERATURE REVIEW 
2.1 Introduction               
The complexity of food choices and understanding the reasons why people choose and eat the 
foods they do is very important in dietary assessment because diet has been shown to be a major 
lifestyle-related risk factor for various chronic diseases (Shim, Oh, & Chang Kim, 2014). Dietary 
intake can be assessed by subjective report and objective observation. Subjective assessment of 
dietary intake is possible through the use of open-ended surveys such as dietary recalls or 
records, or using closed-ended surveys including food frequency questionnaires. Each method 
has inherent strengths and limitations (Shim, Oh, & Chang Kim, 2014). Dietary assessment is 
needed to evaluate the nutritional status of a population so as to provide appropriate nutrition 
education and intervention tailored to improve dietary habits and food choices. This is to help 
curb under nutrition and improve nutritional status (Wunderlich, 2013). Another importance of 
dietary assessment is that it is helpful in the management of diet related diseases especially in 
older adults who are often dealing with chronic medical conditions that require multiple 
pharmaceuticals for therapy and as such drug-nutrient interactions are a prime concern among 
this population group (Wunderlich, 2013). Sub-Saharan Africa (SSA) for instance, is one of three 
regions with the highest risk of NCD deaths between ages 30 and 70 years. The World Health 
Organization’s surveillance on NCD risk factors suggests that SSA is being affected at a rather 
fast pace (World Health Organization, 2010). It is reported that the proportion of adults with 
elevated blood pressure (46%) is greater in Africa than any other region. Also, almost 30% of 
people in SSA do not achieve sufficient physical activity. Overweight prevalence has been 
6 
 
University of Ghana  http://ugspace.ug.edu.gh
reported to be rising rather rapidly in SSA (Ziraba et al., 2009). Among preschool children, 
Africa had the highest incidence of overweight between 1990 and 2010 (de Onis, Blössner, & 
Borghi, 2010). 
Measuring dietary intake enables the assessment of nutritional adequacy of individuals and 
groups and can provide information about nutrients, including energy, food, and eating habits 
(Burrows, Martin , & Collins, 2010). One way of classifying dietary measurement methods is to 
group them into prospective and retrospective methods. (Johansson, 2006).  
2.2 Prospective Dietary Methods 
In the prospective methods foods and beverages are recorded at the time of consumption 
(Johansson, 2006). Examples of prospective methods are the use of Portion Size Measurement 
Aids (PSMAs) such as household measures and food atlases, and food records. One major 
advantage with the prospective methods is that they are not affected by memory since the foods 
and beverages are recorded at the time of consumption. Another advantage is that it provides a 
reasonable estimate and a detailed diet and nutrient intake of an individual (Collins, Watson , & 
Burrows, 2009). The disadvantage is that they often affect the quantity of foods consumed. 
Several studies have demonstrated that the total intake will be low and that there is a selective 
underreporting (Johansson, 2006).  
2.2.1 Portion Size Measurement Aids 
Portion size measurement aids (PSMAs) refer to tools that facilitate the estimation of food 
portions or servings (Ball, 2014). Household measures such as measuring cups and spoons, 
bowls, glasses and plates which are commonly used to quantify food portions are classified as 
7 
 
University of Ghana  http://ugspace.ug.edu.gh
portion size measurement aids (PSMAs) (Weber et al., 1997). There are two (2) main types of 
PSMAs and these are the two dimensional portion size measurement aids (2D PSMAs) and the 
three dimensional portion size measurement aids (3D PSMAs). A Two Dimensional portion size 
measurement aid usually refers to drawings of real foods, abstract shapes and house hold 
measures estimating portion sizes. A three Dimensional portion size measurement aid refers to 
models which come in the form of actual objects such as a golf ball, tennis ball, a deck of cards, 
a sardine tin, just to mention a few which are used to represent portions of foods (Owusu et al., 
1995). In the United States of America there is a more advanced production of food models 
made from either plastics or Styrofoam that resemble actual foods, and these food models are 
classified as 3D PSMAs. Real food samples have also been categorized as 3 dimensional models 
in portion size estimation. Some 3 dimensional food models used in Ghana include an orange 
representing a ball of Ga kenkey and an egg representing corn dough (Peprah Boateng, 2014). 
2.2.2 Food Records  
The most common dietary assessment method for national surveys, and popular with health-
service and diet-related professionals, is the use of food records. They involve the individual 
recording, either manually (with a pen and paper) or electronically), all food consumed in real 
time (i.e. at the time of consumption) over a defined number of days. (Gibson et al., 2016).   
2.2.2.1 Weighed Food Records 
Weighed food records involve an individual or a researcher weighing every food item prior to 
consumption and the leftovers (Wrieden et al., 2003). This method is considered to be precise, 
however the tiresome nature of weighed food records can alter one’s intake and weighing foods 
is not always achievable (Gibson et al., 2016). It includes a written record of actual intake of 
8 
 
University of Ghana  http://ugspace.ug.edu.gh
foods and beverages consumed at the time of consumption for a specified period (usually 3, 5, 7 
days). Food can be measured using a food scale for weights. Multiple days should be randomized 
to include a weekend to measure the actual food intake of an individual. A 7-day weighed record 
has been identified as a ‘gold standard’ against which other methods can be compared. Weighed 
records have the advantage of reporting precision in portion sizes and providing an accurate 
dietary intake of an individual (Wrieden et al., 2003). However, there are some disadvantages to 
this method which includes a high respondent burden for both researchers and participants and 
requires a high motivation of respondents. Habitual eating patterns may be influenced due to 
inconvenience of recording, choice of foods which are easy to record, beliefs about which foods 
are healthy or unhealthy and requires participants to be literate and cooperative (Collins, Watson 
, & Burrows, 2009). Weighed foods over a long period results in respondent fatigue thereby 
affecting food records.  
2.2.2.2 Estimated Food Record 
Estimated food record is a prospective dietary method which provides detailed information on 
food intake. This is quite similar to the weighed food record except that the quantification of 
foods and drinks are estimated rather than weighed. The estimation of the foods in this method is 
done with the help of portion size measurement aids such as household measures with examples 
including cups or spoons, food photographs and food models. The portion size estimation is then 
converted into weights which can be used to calculate an individual’s food and nutrient intake. 
This method of dietary assessment is widely used and has a lower respondent burden compared 
to the weighed food records where an individual is not required to carry a food scale to measure 
all food intakes. However, the only disadvantage of estimation portion sizes is that the possibility 
of mis-reporting is very high in terms of over estimation or under estimation of foods (Wrieden 
9 
 
University of Ghana  http://ugspace.ug.edu.gh
et al.,2003). This can negatively affect appropriate dietary counseling and intervention provided 
by nutrition professionals.  
2.3 Retrospective Dietary Methods 
The commonly deployed retrospective methods are food recalls, dietary histories and food 
frequency questionnaires. They usually make use of food aids such as food models, rulers, and 
household measures. These groups of methods have the advantage of being simple and cheap. 
Also they are suitable for a large sample size. The main demerits of this group of methods are 
their reliance on the memory of people (which is fallible) and the ability to estimate quantities 
varies (Johansson, 2006). 
2.3.1 24 hour Recall 
A 24-hour dietary recall is a retrospective method of dietary assessment in which an interview is 
conducted to obtain detailed information on all foods, beverages and possibly dietary 
supplements consumed by a respondent in the past 24 hours. In a 24-hour recall, subjects are 
required to recall the exact food intake during the previous 24 hours or the preceding day.  It 
includes the time of day at which the food was consumed as well as the portion size of each food 
and beverage recorded. A single 24 hour recall is not sufficient enough to provide an accurate 
estimate of long-term energy intake as diets vary considerable on a daily basis. It is therefore 
suggested by some researchers that multiple 24 hour recalls such as three (3), four (4) five (5) or 
seven (7) days are necessary to provide adequate estimate of energy intake (Yunsheng et al., 
2009). A 24-hour recall has a low respondent burden, requires no literacy and is suitable for a 
large scale survey (Wrieden et al., 2003) and can be used to calculate the average dietary intake 
of a population (Raina, 2013). There are some limitations in the use of the 24-hour recall; one of 
10 
 
University of Ghana  http://ugspace.ug.edu.gh
such limitation is the fact that it relies on memory and as such foods consumed may be forgotten 
by respondents. Another disadvantage is that data generated through this method may not 
represent long-term dietary habits of individuals (Raina, 2013).  
2.3.2 Diet Histories 
Diet History is a detailed retrospective dietary assessment method which is more frequently used 
in clinical practice than in research studies. It is used to describe usual food and nutrient intake of 
an individual over a relatively long period, for instance, one (1) month, six (6) months or as long 
as one (1) year ( Fagúndez et al., 2015). Diet history is considered as a traditional method of 
analysis of dietary intake. It traditionally consists of three (3) components that make available 
detailed information on the usual food consumption pattern of individual and detailed 
information on certain foods. The first of the three (3) constituents include an interview about the 
usual food intake pattern of an individual and estimation of the quantity of food consumed by 
means of household measures. This is followed by a questionnaire consisting of a detailed list of 
foods to assess the overall food intake pattern and to verify the information obtained from the 
first part.  The final part is a 3-day food record with estimated portion sizes of the foods and 
beverages consumed (Fagúndez et al.,2015). The main advantage of the diet history is its ability 
to detect seasonal changes in dietary pattern. It is also efficient in obtaining data on all nutrients 
and to correlate well with biochemical measures. However, this method bears a high respondent 
burden and consequent loss of data quality (Naska, Lagiou, & Lagiou, 2017). 
2.3.3 Food Frequency Questionnaire 
Food Frequency Questionnaire (FFQ) refers to a limited checklist of foods and beverages 
common to a group of people with a frequency response section for subjects to report how often 
11 
 
University of Ghana  http://ugspace.ug.edu.gh
each item was consumed over a specified period of time. The frequency response includes the 
number of times per day, week, month a food item is consumed. It is designed to collect dietary 
information from large numbers of individuals such as a hundred (100) individuals or more and 
is normally self-administered (Wrieden et al., 2003). It can also be administered via a phone 
interview involving a researcher and a participant. There are two (2) of FFQs and they are semi-
quantitative and non-quantitative. Semi-quantitative FFQs are designed to collect portion size 
information as standardized portions or as a choice of portion sizes, whereas the non-quantitative 
FFQs do not require portion size information (Wrieden et al., 2003) It has the advantage of 
representing a habitual dietary intake of an individual or a group of individuals and their food 
patterns. It is also a preferable method of measuring intake of nutrients with very high day-to-day 
variability. Administration of questionnaire is significantly less expensive in comparison to food 
records or dietary recalls. It has a low respondent burden and as such suitable for very large scale 
surveys. The disadvantages of a food frequency questionnaire include a high dependence on 
respondents’ memory, requires that a participant be literate to understand the frequency of 
consumption and a high possibility of over-reporting of healthy foods (Collins, Watson , & 
Burrows, 2009). 
2.4 Portion sizes and dietary assessment  
2.4.1 Portion size 
A portion according to Nelson & Haraldsdottir (1998) is the quantity of food eaten at any one 
occasion; this selected portion may be bigger or smaller than the standardized serving of the 
food. A serving is a unit of measure used to describe the amount of food recommended from 
12 
 
University of Ghana  http://ugspace.ug.edu.gh
each food group (USDA, 2015). "We consume much more than we think we do," says Edgar 
Chambers IV, Ph.D., professor of human nutrition at Kansas State University.  
The rise in fast food joints, who serve cheaper foods in large portions, has been connected to the 
spate of obesity in the United States. The frequency of consumption of large portion of food, 
especially food that is high in energy density, may be a contributing factor to excess energy 
intake coupled with physical inactivity results in the development of obesity (Cashin-Garbutt, 
2017). Over the last twenty years, the typical American diet has changed dramatically both in 
quantity and quality of food intake. Recent studies have shown that the average adult used to 
consume an average of 2,160 kilocalories per day in 1970 and has increased by 20-25 percent to 
2,673 kilocalories per day in 2016. Examples given included a regular muffin weighing 1.5 
ounces and French fries weighing 2.4 ounces both with a caloric content of 210 kilocalories have 
increased in weight and caloric content over the past twenty (20) years to 4 ounces with 500 
kilocalories and 6.9 ounces with 610 kilocalories respectively (Scinta, 2016). Unfortunately, 
people do not see these changes and view the increased portion sizes as normal, and this is 
described as portion distortion.  
2.4.2 Portion Distortion 
Portion distortion is a term used to refer to situations where people tend to regard excessive 
portions as normal amounts (Faulkner et al, 2012). These situations are believed to contribute to 
unhealthy eating habits that may lead to obesity. There are increasing concerns that larger 
portion sizes may encourage over-eating and contribute to the high obesity rates seen across 
developed countries (Faulkner et al., 2012); a case not too dissimilar to our situation in the 
developing world. The proportion of people who are obese and overweight has increased sharply 
13 
 
University of Ghana  http://ugspace.ug.edu.gh
in many countries. In the US for example, the proportion has doubles since 1980 and almost a 
third of all adults are classified as obese (Duffey & Popkin, 2011). A cross-sectional study 
conducted in the US showed an increased intake of energy by both children and adults which is 
largely contributed by an increased portion sizes and increased eating frequency over the last 
three decades. (Livingstone & Pourshahidi, 2014). A study by Piernas & Popkin, 2011 compared 
portion sizes to physiological cues such as hunger and satiety and identified that portion sizes 
have more influence in the quantity consumed resulting in excess intake of high energy foods. 
Larger portions sizes of foods and beverages are thought to affect energy intakes at meals and 
promote overeating (Piernas & Popkin, 2011).  
Nutritional information although may be present on packaged foods based on serving sizes, 
people generally do not correctly assess the amount they are eating. Often consumers are unable 
to tell the differences in portion size when offered different sizes on different days. After 
consumption of larger portions in one eating occasion, people normally compensate that by 
eating fewer calories during the rest of the day or the time period before or following the eating 
occasion. However, this in many cases is difficult for many to do because researchers found that 
the people who ate large portions did not notice the size difference and ate their normal amount 
of food at the following meal. Portion size is therefore a modifiable determinant of energy intake 
that should be considered in the prevention and treatment of obesity (Young & Nestle, 2012).   
In order to curb the spate of high obesity rates and increasing cases of diet related NCDs there is 
the need to ascertain what and how much food people are consuming.    
14 
 
University of Ghana  http://ugspace.ug.edu.gh
2.4.3 Portion Size Measurement Aids 
Maintaining a healthy diet is necessary in preventing, treating and managing diseases such as 
obesity, diabetes and certain cancers. Part of this is done by becoming aware of one’s portion 
size intake. Healthcare professionals use the portion size intake information of their patients to 
undertake nutritional assessment, and to educate them about specific servings and appropriate 
use of portion measurement aids. 
Portion Size Measurement Aids (PSMAs) can be defined as tools that assist in the facilitation of 
food estimation (Ball, 2014). One major barrier to controlling the amount of food consumed may 
be the difficulty consumers have in accurately estimating portion sizes due to the fact that most 
people are unaware or unconscious of the amount of food eaten (Riis, 2014). The accuracy of 
estimating portion sizes of foods is of great importance because it can influence the quality of 
dietary intake data as well (Jia et al., 2014). A major benefit of using portion size measurement 
aids (PSMAs) may improve estimation accuracy, however, it has the  disadvantage of being 
bulky and costly and so tends to make them impractical for regular use (Brrd-Bredbenner & 
Schwartz , 2004).  
With ever-increasing obesity rates and the commensurate rise in portion sizes, it is imperative 
that diet and nutrition-related professionals work with their clients to ensure accuracy in the 
estimation of their dietary intake. There is mis-reporting  of dietary intake due to factors such as 
variation in food consumed from day to day, poor memory recall, inability to estimate portions 
resulting in over or underestimation, and inadequate knowledge of portion sizes. (Hight, 2008). 
The ability to estimate portion size of food eaten appears to be affected by the food type such as 
foods presented in multiple units (Almiron-Roig et al., 2013). Awareness of one’s portion sizes 
15 
 
University of Ghana  http://ugspace.ug.edu.gh
is a great step to maintaining a healthy diet and weight and portion size measurement aids 
(PSMAs) have been identified as being beneficial for subjects trying to accurately describe 
portion sizes (Chaudry, Connelly, Siek, & Welch, 2011). A number of factors play a role in 
reliability of aids used for estimation including the way the aids are presented to the subjects, the 
type of aid used for recall, the food type and shape, characteristics of the subject, and the extent 
to which the aid resembles the size and shape of the food (Hight, 2008).  
Accurate information about oral intake of individuals and populations is particularly difficult to 
obtain; it is usually reliant upon self-report and can be subject to large errors. There is much 
controversy in literature about the value of self-reported food intake, particularly for research 
into energy intake and obesity (Gibson et al., 2016) stresses the importance of the collection of 
data on the volume and kinds of different individuals and groups in clinical and educational 
settings in a bid to affect policies on diet and health. 
 Currently, it has been observed that people can estimate healthy portion sizes either by using 
serving size information printed on food packages or by making visual comparisons with various 
objects. Unfortunately, both methods present problems for low literacy populations (Gibson et 
al., 2016). In a study done by Ollberding and Wolf (2010) it was observed that 47.2% of 
Americans use serving size information on food packages to consume less energy, fat and sugar ( 
Ollberding & Wolf , 2010). However, the information on serving size on packaged foods is 
difficult to understand particularly for low literacy and numeracy populations (Rothman et al., 
2006). The other option will be for people to determine portion sizes by associating food portions 
with visual aids recommended by registered dietitians. For example, an appropriate portion size 
for pasta is ½ a cup, which is visually equivalent to a tennis ball. Such estimation aids, however, 
lack a standard definition (Ball & Friedman , 2010) making their general use challenging. 
16 
 
University of Ghana  http://ugspace.ug.edu.gh
Moreover, people do not always eat portion sizes comparable with these aids, and hence the need 
to track and calculate total daily intake is not completely eliminated. 
2.5 Household Measures and dietary assessment  
One of the keys to maintaining weight and staying healthy is portion control, but serving sizes 
can be rather vague and difficult to measure exactly. Common household measures, such as 
cups, measuring spoons, glasses, plates and bowls are frequently used to quantify portion sizes 
(Peprah Boateng, 2014). These household measures are used in conjunction with other 
assessment methods in assessing dietary intake prospectively or retrospectively.  
A good illustration of the use of household measures can be found in the United States; The 
United States Department of Agriculture (USDA) has come up with a dietary guide called the 
‘MyPlate Food Guide’, which identifies daily meal proportions for the fruit, vegetable, grains, 
protein, and dairy food groups. It provides an illustration of appropriate portion sizes of the 
different food groups, with an emphasis on fruits and vegetables (constituting half the plate) 
(USDA, 2016).  
Currently, in Ghana, household measurement aids are the dietary aids of choice in portion size 
estimation by dietitians and nutrition-related professionals. This, in my opinion is largely the 
case owing to the fact that it is easier for the sampled population to relate to these items. 
Inasmuch as household measures offer simplicity and universality as merits for their use, they 
present the challenge of being cumbersome.  
17 
 
University of Ghana  http://ugspace.ug.edu.gh
2.6 Photographic Atlas 
A food photographic atlas is a type of portion size estimation aid, which can be defined as “a set 
of photograph series of food in portion sizes, usually bound together in a single volume” (Nelson 
& Haraldsdottir, 1998). The photographic food atlas may contain portion sizes ranging between 
3 to 8 different portion sizes of food (Turconi et al., 2005). Photographs of foods are usually 
taken of small, medium and large portions, which are judged to be representative of the range of 
portion sizes actually consumed by a group of people. Individuals are asked to choose from the 
photographs of food the one that best reflects their usual portion size or their actual portion size. 
Alternatively, a food photographic atlas can contain a single photograph of the average portion 
size and individuals asked to estimate their own portion size as a fraction or multiple of the 
single photograph displayed.  
The assessment of food portion sizes from food photographic atlases according to Turconi et al., 
(2005) involves three (3) main elements: perception, conceptualization and memory. Perception 
refers to an individual’s ability to relate a quantity of food that is present in reality to a portion or 
quantity of the same food depicted in a food photographic atlas. Conceptualization involves an 
individual’s ability to make a mental picture of an amount of food that is not present in reality 
but is able to relate that to that quantity of food displayed in the food photographic atlas. 
Memory is one’s ability to recall accurately the amount of food eaten over a period of time and 
can have an effect on the accuracy of conceptualization (Turconi et al., 2005).  All these three 
elements will be affected by a number of factors such as the number of different portion sizes, 
their placement in the food photographic atlas, the dimension of each photograph and the camera 
angle at which each of the food was taken and framed. One advantage of a food photographic 
atlas especially if it is colored can be used by all age groups; however, one disadvantage may be 
18 
 
University of Ghana  http://ugspace.ug.edu.gh
a misreporting of portion sizes especially when compared to weighed foods (Turconi et al., 
2005). 
Various researchers have done extensive work on the use of the photographic atlas all over the 
world notably work done by (Nelson & Haraldsdottir, 1998) and (Tueni et al., 2012), to mention 
a few. In a study by (Lazarte et al., 2012), a comparison was made between the use of a digital 
food photographic atlas and weighed food record, and it was concluded that there was no 
significant differences between the two methods with exception to some food categories such as 
rice.  
In contrast, very little work has been done on the validation and use of the photographic atlas in 
Ghana. Pioneering studies by Peprah Boateng (2014) asserts, inter alia, that access to a 
photographic food atlas of commonly consumed Ghanaian foods will aid in data collection, 
nutrition education and dietary counseling. The findings of the validation of the food 
photographic atlas in Accra, Ghana by Peprah Boateng (2014) showed a statistical significance 
in correct estimation of portion sizes using the photographic atlas, and therefore, being a useful 
dietary measurement aid. 
 
 
 
 
 
19 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 3 
3.0 METHODOLOGY 
3.1 Study Design 
This was a cross sectional study design.  
3.2 Study Site  
The study was carried out at Asesewa, specifically in the Nutrition Research Centre and the 
Asesewa Regional Hospital in the Upper Manya Krobo district (UMKD) of the Eastern Region 
of Ghana. Asesewa was chosen because of an ongoing project by the department of Dietetics 
which created a platform for easy access to recruit participants. According to the Ghana 
Statistical Service’s Population and Housing Census for 2010 the district has a population of 
72,092. Asesewa is the district capital and has an estimated population of about 20,291. The 
district shares boundary with the Volta Lake in the north, Fanteakwa District in the west, 
Asuogyaman District in east, Yilo Krobo District in the south-west and Lower Manya Krobo in 
the south-east. The main economic activity in the district is agriculture employing about 80% of 
its population, most of whom are subsistence farmers with very few commercial ones. 
3.3 Study Population 
The study included all males and females within the age group of eighteen (18) years and 
seventy (70) years who visited the Nutrition and Research center and the Asesewa Regional 
Hospital within the time period of collecting the data. 
20 
 
University of Ghana  http://ugspace.ug.edu.gh
3.3.1 Inclusion and Exclusion criteria 
Males and females within the age group of eighteen (18) years and seventy (70) years who 
indicated their willingness to partake in the study by filling a consent form attached to the 
questionnaire used in data collection.  
Children under the age of eighteen (18) years and the elderly above the age of seventy (70) years 
were excluded from the study; this is due to their inability to accurately estimate portion sizes. 
Pregnant women were also excluded because of the comparison of portions size estimation based 
on body mass index (BMI). 
3.3.2 Sample size determination  
The sample size of participants recruited for the study was calculated using the total population 
of Asesewa from the Census Report (2010) to find a percentage of the population within the 
study age bracket. According to the Census Report (2010), the total population of the Upper 
Manya Krobo District is 72,092 and the population above eighteen (18) years is 38,509. The 
sample size was calculated using the formula: 
n = (z2 (p) (1-p))/d2    
n = (〖1.96〗2 (0.15) (1-0.15))/〖0.05〗2  
n = 196 
Where ‘n’ is the sample size, ‘z’ is the percentile of the required confidence interval (1.96), p is 
the population estimate (0.15) and‘d’ is the allowable error (0.05). A minimum sample size 
estimate of 196 was obtained.  
21 
 
University of Ghana  http://ugspace.ug.edu.gh
3.4 Pre-testing of questionnaire 
Twenty (20) people who met the criteria for this study were conveniently selected to test the 
questionnaire for clarity. Pre-testing of questionnaire was done in Asesewa, specifically in the 
Nutrition Research Centre and the Asesewa Regional Hospital in the Upper Manya Krobo 
District (UMKD) of the Eastern Region of Ghana but these participants did not form part of the 
final recruitment for the study. The questionnaire was designed to collect socio-demographic 
information such as date of birth, gender, marital status and educational level; anthropometric 
measurements such as height, weight and BMI; and a three (3) day diet history.  
3.5 Data collection 
Data collection was divided into three (3) main parts: socio-demographic and anthropometry, 24-
hour recall and portion size estimation. A flow chart of data collection is shown below:  
 
22 
 
University of Ghana  http://ugspace.ug.edu.gh
3.5.1 Subject recruitment 
The researcher recruited and trained two (2) personnel to assist in data collection. Prior to the 
distribution of questionnaire, subjects were given a brief introduction of the researcher and the 
trained personnel and the purpose of the study was explained to them. Those who were willing to 
partake in the study and met the criteria were given an informed consent form to fill with the 
help of the researcher by translating in the local dialect where necessary. They were informed of 
their right to leave this study if confidentiality is broken. 
3.5.2 Data collection material 
A pre-tested questionnaire was used in collecting data from 196 respondents which included 
socio-demographic information such as age, gender, educational level, and marital status. The 
researcher was also required to conduct a face-to-face interview to fill out a 3-day 24-hour 
dietary recall of participants’ intake. 
3.5.3 Anthropometric measurements 
Anthropometric measurements are a set of inexpensive, noninvasive, quantitative method of 
assessing an individual’s fat and muscle composition, and can be useful in the nutritional 
assessment of an individual (Sánchez-García et al.,2007). The anthropometric measurement of 
interest to this study was height and weight, which were used to calculate the body mass index 
(BMI) of participants. The calculated BMI was compared to reference standards. For this study, 
the reference values were based on WHO (2012) criteria which define underweight as a BMI 
(<18.5kg/m2), healthy weight (18.5kg/m2 – 24.9kg/m2), overweight (25.0kg/m2 – 29.9kg.m2) 
23 
 
University of Ghana  http://ugspace.ug.edu.gh
and obese (≥30.0kg/m2). Weights and heights were recorded by a researcher and trained research 
assistants for all participants using a stadiometer. 
3.5.3.1 Weight 
The weights of participants were measured by researcher and trained research assistants using a 
weighing scale (Seca 755). Prior to the measurement of weights, participants were asked to 
empty their pockets and remove jackets, scarfs, wrist watches and footwear before standing on 
the scale. Participants were asked to stand upright and look forward when the measurement was 
being recorded. 
3.5.3.2 Height 
The heights of participants were measured using a stadiometer (Seca 755) which was attached to 
a digital scale. Participants were asked to stand upright and inhale for the heights to be recorded. 
The values of the recorded heights and weights were used in the calculation of BMI for each and 
every participant and categorized according to the reference standards. For this study, the 
reference values were based on WHO (2012) criteria which define underweight as a BMI 
(<18.5kg/m2), healthy weight (18.5kg/m2 – 24.9kg/m2), overweight (25.0kg/m2 – 29.9kg.m2) 
and obese (≥30.0kg/m2).  
3.5.4 Dietary Assessment 
An interview was conducted using a 24-hour dietary recall to obtain information on participants’ 
intake in the past 24-hours and then estimation of portion sizes of the meals listed were made 
using household measuring aids and a food photographic atlas. Participants were given codes 
24 
 
University of Ghana  http://ugspace.ug.edu.gh
which matched their questionnaire and were either visited in their homes or returned to the 
nutrition research center or the Asesewa regional hospital for day 2 and day 3 dietary recalls.   
Household measuring aids were used to assist portion size estimation. These included measuring 
spoons, measuring cups, soup ladles, stew ladles, empty small tin tomato, and empty milk can 
(185g), and match box. The estimation of the meals were limited to the carbohydrate foods that 
the researcher and research assistants identified from the list of foods provided by the 
participants in the previous 24-hour period. Estimation of food using food photographic atlas of 
the same carbohydrate foods was also required of participants with the help of the research team. 
The food photographic atlas contained carbohydrate foods in eight different portion sizes. 
Participants were asked to choose from the book how their portion sizes looked like based on the 
estimation made using the household measuring aids. This was repeated for the other two days. 
3.6 Data Analyses 
Analyses and statistical procedures were carried out using the Statistical Package for Social 
Sciences program (SPSS, version 21.0 for Windows). Results were expressed as means ± SD. 
Pearson’s Chi-square tests and one-sample t-test were used to compare individual and overall 
carbohydrate estimation by male and female participants. For all statistical comparisons, a P-
value of <0.05 was considered as statistically significant. Pictorial representations are provided 
where applicable. 
3.7 Ethical approval 
Approval for the study was obtained from the Ethics and Protocol Review Committee of the 
School of Biomedical and Allied Health Sciences for approval. Permission was sought from the 
health directorate and the Regional Hospital of Asesewa and any protocol observed. The purpose 
25 
 
University of Ghana  http://ugspace.ug.edu.gh
and significance of the study was explained to participants. Research Participants Information 
Sheets and written consent were obtained before commencement of data collection. Strict 
confidentiality and anonymity were assured by including a confidentiality clause in the written 
consent form to reassure participants and also served as a reminder to the researcher of her 
professional duty to the respondents. The questionnaire used in data collection required 
participants to indicate their willingness to partake in the study by filling a consent form. Data 
collection procedure was in accordance with approved protocol. 
 
 
 
 
 
 
 
 
 
 
 
 
26 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 4 
4.0 RESULTS 
4.1 Background profile of participants 
Out of the one hundred and ninety-six (196) participants recruited for the study, six (6) 
participants had their questionnaire rejected because they failed to fill some portions of the 
questionnaire. Out of the remaining, one hundred and ninety (190) participants, 53.2% (101) 
were males and 46.8% (89) were females with majority within the age bracket of 18 – 38 
representing 63.1% (120). Ninety-two (48.4%) of the respondents were married or cohabiting, 
seventy-nine (41.6%) were never married and six (3.2%) were divorced or separated (Table 4.1). 
Most of the respondent (42.6%) had some form of education with majority having middle/Junior 
secondary school education and only few (11.6%) had no formal education (Figure 4.1) 
Most participants had monthly income levels ranging from below two hundred (200) to five 
hundred (500) Ghana cedis. Forty-nine percent (49.0%) of participants had monthly income 
below two hundred Ghana cedis (₵200) followed by approximately thirty-five (34.7%) with a 
monthly income between two hundred Ghana cedis (₵200) and five hundred Ghana cedis 
(₵500). Only a few (6.1%) recorded a monthly income level of over one thousand Ghana cedis 
(Figure 4.2).  
 
 
 
27 
 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.1: Socio-demographic characteristics of respondents N= 190 
Variable n (%) 
Gender 
  Male 1 01 (53.2) 
  Female 89 (46.8) 
Age 
  18 – 28  70 (36.8) 
  29 – 39 50 (26.3) 
  40 – 50 39 (20.5) 
  51 – 61 24 (12.6) 
  62 – 70 7 (3.7) 
Marital status 
  Never married 79 (41.6) 
  Married/ Co-habiting 92 (48.4) 
  Separated/Divorced 6 (3.2) 
  Widowed 13 (6.8) 
 
 
 
 
 
28 
 
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.1: Educational Level of Participants 
22, 11.6%
41, 21.6%
None
Middle/ Junior Secondary School
Senior Secondary School
Post-Secondary/ Tertiary Level
46, 24.2% 81, 42.6%
 
 
Figure 4.2: Average monthly income of Participants 
12, 6.3%
20, 
10.5%
< 200
93, 48.9% 200 - 500
500 - 1000
65, 34.2% 1000 <
 
29 
 
University of Ghana  http://ugspace.ug.edu.gh
4.2 Anthropometric assessment of Respondents 
The mean height of all the respondents was 1.62 meters; however, the males with a mean height 
of 1.65 meters were taller than the females with a mean height of 1.58 meters (p-value < 0.001). 
The mean weight was 65.5kg with the females having a higher weight than the males. The mean 
Body Mass Index (BMI) was 24.96kg/m2. The mean BMI for the females was higher than that of 
males (p-value < 0.0001) (Table 4.2). 
 Figure 4.3 shows the BMI classification of participants. A large proportion (53.7%) of the 
respondents had normal BMI, followed by 31.1% who were overweight with very few (2.6%) (3 
females and 2 males) being underweight. However, a higher proportion of females were obese as 
compared to their male counterparts (Figure 4.4). The age range of 18 – 28 years reported the 
highest number of normal weight individuals (68; 35.8%) as well as overweight individuals (39; 
20.5%) as displayed in figure 4.5   
 
Table 4.2: Distribution of mean height, weight and Body Mass Index among respondents 
Mean height P-value Mean weight P-value Mean BMI P-value 
Category 
± SD. ± SD. ± SD. 
Male 1.65 ± 0.089 <0.001 65.12±9.610 <0.001 24.37±4.475 <0.001 
Female 1.58 ± 0.061 <0.001 65.93±11.757 <0.001 25.63±4.157 <0.001 
Total 1.62 ± 0.084 <0.001 65.5±10.648 <0.001 24.96±4.363 <0.001 
One sample T-test; statistical significance P-value<0.001 
 
 
30 
 
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.3: BMI classification of Participants 
5, 2.6%
24, 12.6%
Underweight
Normal weight
59, 31.1% Overweight
102, 53.7%
Obese
 
Figure 4.4: BMI distribution based on gender  
 
70
64.4
60  
50
41.6
40 34.8
30 27.7
20.2
20
10 5.9
2 3.4
0
Underweight Normal weight Overweight Obese
Male Female
BMI Classification
31 
 
Percentage
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.5: BMI distribution of participants based on age groups 
50
45
40
35
30
UNDERWEIGHT
25
NORMAL
20
OVERWEIGHT
15
OBESE
10
5
0
18 - 28 29 - 39 40  - 50 51 - 61 62 - 70
Age
 
4.3 Commonly consumed carbohydrate foods 
From the three day 24-hour food recall, fifty-four (54) commonly consumed carbohydrate foods 
were identified. These were grouped into three (3) main categories: the cereal and grain category 
(61.2%), the roots, tubers and plantain category (30.5%), and the sugar group (8.3%) as 
displayed in table 4.3. Based on Peprah Boateng’s study (2014) the foods that were considered as 
most commonly consumed carbohydrates foods made up 5% or more of the frequency of the 
total carbohydrate foods consumed within the three days, using a 24 hour food recall. There were 
five (5) foods that were identified as being commonly consumed and these are: Banku forming 
19.0%, Fufu (13.8%), Boiled rice (7.5%), Boiled yam (5.7%) and Sugar (granulated) forming 
8.2%.  
 
 
32 
 
Percentage
University of Ghana  http://ugspace.ug.edu.gh
Table 4.3 Commonly Consumed Carbohydrate Based foods  
Cereals and Grains % Roots, Tubers and % Sugars % 
Plantain 
Abolo 0.2 Boiled yam 5.7 Sugar 8.2 
Agidi 0.3 Cocoyam (boiled) 0.5 
Banku 19.0 Fried potato 0.1   
Boiled corn 0.5 Fried ripe plantain  1.0   
Boiled rice 7.5 Fried yam 0.7   
Bread roll 2.6 Fufu 13.8   
Butter bread 1.9 Gari   2.7   
Corn flakes 0.2 Gari foto 0.1   
Corn porridge 4.0 Kakro 0.1   
Crackers 0.4 Kelewele 0.2   
Digestive biscuits 0.1 Kokonte 1.6   
Doughnut 1.1 Plantain chips 0.5   
Fante kenkey 0.4 Ripe plantain (boiled) 0.1   
Fried rice 0.2 Roasted Plantain 0.2   
Fula 0.3 Unripe Plantain (boiled) 3.6   
Ga Kenkey 4.2   
Hausa Koko 3.2     
Jollof rice 0.8     
Meat pie 0.5     
Millet porridge 0.1     
Oats 0.5     
Oblayo 0.7     
Omutuo 0.8     
Pastry chips 0.1     
Pop corn 0.3     
Rice porridge 0.8     
Roasted corn 0.9     
    
33 
 
University of Ghana  http://ugspace.ug.edu.gh
Rock buns 0.1 
Saabo 0.1     
Spaghetti 0.4     
Sugar bread 0.4     
Sweetened biscuits 0.1     
Tea bread 3.2     
Tom brown 0.7     
TZ 0.8     
Waakye 3.5     
Wheat bread 0.1     
Yakeyake 0.1       
   
4.4 Participants’ estimation of common carbohydrate foods 
This section presents results on estimation of common carbohydrate foods by the participants. It 
further illustrates estimation by gender, age and BMI. Overestimation in this study implies that 
the participant’s selection of portion size from the food atlas exceeds that of the portion size of 
an identical food presented using the household measuring aids. Correct portion size estimation 
is when a participant’s choice of a food portion from the food atlas is identical to that presented 
by the household measuring aids. Underestimation is indicated by the choice of a lower portion 
size than that using the household measuring aids. 
4.4.1 Overall carbohydrate estimations by gender 
There was no significant difference in gender estimations of carbohydrate foods (P= 0.295). 
However, a higher proportion of female participants were able to estimate correctly (7.9 %) 
compared to males (3.0 %). No significant differences (P>0.05) were found between the number 
of participants who underestimated or overestimated carbohydrate among gender (Figure 4.6) 
34 
 
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.6: Overall carbohydrate estimation by gender 
 
70 66.3
60.7
60
50
40
29.7 Male
30 28.1
Female
20
10 7.9
3
0
UE CE OE
overall carbohydrate estimation 
 
Pearson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
 
4.4.2 Carbohydrate foods estimation by BMI categories 
A significantly higher proportion (80.0%; P=0.031), of participants that were underweight 
underestimated portion sizes (Figure 4.7). However, the effects of the different BMI categories 
on participants’ ability to correctly estimate or overestimate were not significant. 
 
 
 
 
35 
 
Percent estimation
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.7: Participants’ BMI and overall portion size estimation 
 
90
*P=0.031
80
70
60
50 Underweight
40 Normal weight
Overweight
30
Obese
20
10
0
UE CE OE
BMI and level of estimations 
 
Pearson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
 
4.4.3 Age assessment of carbohydrate foods estimation 
Overall carbohydrate estimations by participants was not statistically significant for age 
(P=0.640). A high proportion, (66.7%), found within the 39-49 year range underestimated 
portion sizes of the carbohydrate food. Correct portion sizes estimation was observed in higher 
proportions of participants aged between 61-70 years. However, the age group of 61-70 years 
showed a higher proportion of overestimation. (Figure 4.8) 
 
 
 
36 
 
Percent estimation
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.8: Age range of participants and portion size estimation 
80
70
60
50
17-27
40 28-38
30 39-49
50-60
20
61-70
10
0
UE CE OE
Age and level of estimations 
 
Pearson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
 
 
4.4.4 Estimation of the types of carbohydrate foods identified 
Presented in figure 4.9 is the estimation of portion sizes in fufu, banku, boiled yam and rice and 
sugar. Overall estimations of the different types of carbohydrate groups were not significant. 
Majority of the participants, 77.8%, underestimated the portion size of sugar, followed by banku 
(68.6%) and fufu (62.8%). Approximately fifty-four (53.8%) of the participants estimated the 
portion sizes of boiled yam correctly, followed by boiled rice (25.3%), fufu (20.5%) and banku 
(19.9%). Overestimation was greater in boiled rice (57.5%), followed by boiled yam (24.6%) and 
less in fufu (16.7%). (Figure 4.9) 
 
 
 
 
37 
 
Percent estimation
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.9: Estimation of common carbohydrate food 
90
80 77.8
68.6
70 62.8
60 57.553.8
50 Under estimation
40 Correct
30 24.6 25.3 estimation
20.5 19.9 21.6
20 16.7 17.2
Over estimation
14.4
11.5
10 7.8
0
Fufu Banku Boiled yam Boiled rice Sugar
 
4.4.5 Gender related estimation of the different types of carbohydrate foods 
Figure 4.10 shows gender specific estimation of portion sizes in some common carbohydrate 
foods. Except for the estimation of banku, the other estimations were not significantly different 
between genders (P> 0.05). 
In estimating the portion sizes, higher proportion of males (55.4%) underestimated fufu 
compared to females (44.6%) whilst higher proportion of females (59.3%) correctly estimated 
fufu compared to the males (40.7%), however this was not statistically significant (P = 0.410). 
Among gender, a statistically significant higher proportion of females (74.2%) correctly 
estimated banku compared to males (25.8%) (P=0.001). Boiled yam on the other hand was 
underestimated by higher proportion of males (71.4%) compared to females (28.6%), whilst 60% 
of males correctly estimated boiled rice compared to females (40.0%) (P=0.464). Similarly, 
sugar was correctly estimated by 69.2% males and 30.8% females (Figure 4.10).  
 
 
38 
 
percent estimation
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.10: Gender specific estimation of the different types of carbohydrate foods 
80
*p = 0.001 
70 p = 0.216 
p = 0.410 p = 0.464 p = 0.377 
60
50
40
30
20
10
0
Male Female Male Female Male Female Male Female Male Female
Fufu Banku Boiled yam Boiled rice Sugar
UE 55.4 44.6 61.7 41 71.4 28.6 41.2 58.8 51.4 48.6
CE 40.7 59.3 25.8 74.2 45.7 54.3 60 40 30.8 69.2
OE 50 50 72.2 27.8 43.8 56.2 49.1 50.9 42.9 57.1
 
Pearson’s Chi square: *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
 
4.4.6 BMI related estimation of the different types of carbohydrate foods 
The BMI of the participants were not significantly associated with correct estimations of fufu 
(P=0.425), banku (P= 0.854), boiled yam (P= 0.818), boiled rice (P=0.275) and sugar (P=0.795). 
However, the highest proportion of participants who underestimated fufu and sugar (100%) were 
found among those with BMI <18.5 m/kg² (underweight category) with proportion 23.8% and 
26.1% of overweight participants correctly estimating fufu and banku respectively. Although 
there was no statistically significant association between BMI of participants and correct 
estimation of boiled yam, a higher percentage (57.1%) of participants within the underweight 
category (<18.5 kg/m²) and 53.8% within the obese category (≥ 30 km/kg²) correctly estimated 
(Table 4.4). Higher proportion (66.7) of the participants within the normal weight category 
39 
 
% Estimation
University of Ghana  http://ugspace.ug.edu.gh
overestimated boiled rice followed by 48.5% and 45.5% within the overweight and obese 
categories respectively.  
Table 4.4: BMI specific estimations of the different types of carbohydrates foods 
Under Correct Over 
BMI N estimation estimation estimation P-value 
 Fufu Underweight 2 100 0.0 0.0 0.425 
Normal weight 69 56.5 20.3 23.2 
 Overweight 42 66.7 23.8 9.5  
 Obese 19 73.7 15.8 10.5  
 Banku Underweight 5 80 20 0 0 .854 
Normal weight 86 70.9 17.4 11.6 
 Overweight 46 60.9 26.1 13  
 Obese 19 73.7 15.8 10.5  
 Boiled yam Underweight 2 0 100 0 0 .818 
Normal weight 30 23.3 46.7 30 
 Overweight 21 19 57.1 23.8  
 Obese 12 25 53.8 24.6  
 Boiled rice Underweight 1 0.0 100 0 0 .275 
Normal weight 54 11.1 22.2 66.7 
 Overweight 16 24.2 27.3 48.5  
 Obese 5 27.3 27.3 45.5  
 Sugar Underweight 1 100 0 0  0.795 
Normal weight 47 72.3 19.1 8.5 
 Overweight 33 81.8 12.1 6.1  
 Obese 9 88.9 0 11.1  
P earson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation ; UE: 
underestimation; OE: Overestimation 
40 
 
University of Ghana  http://ugspace.ug.edu.gh
4.4.7 BMI and gender evaluation of the different types of carbohydrate foods 
The results presented in table 4.5 showed that correct estimation of fufu by both gender was not 
statistically associated with BMI. A higher proportion of overweight (69.6%) and obese (83.3%) 
males underestimated portion sizes of banku when compared to females. A significantly 
(p=0.047) higher proportion of overweight females correctly estimated portion sizes of yam 
(63.6%) compared to males ((40.0%)., Correct estimation was significantly higher among normal 
weight males (24.2%) compared to females (19.0%) for boiled rice (0.047). Over estimation of 
boiled rice by gender was higher in the females in both overweight (45.0%) and obese (50.0%) 
categories compared to the males. Sugar on the other hand was also highly underestimated by 
males in the overweight and obese categories. Among normal weight participants, more females 
(33.3%) than males (8.9%) correctly estimated portion sizes of banku (p=0.015) (Table 4.5) 
In table 4.6, 78.3% and 65.0% malnourished males and females respectively underestimated 
portion size of fufu. Malnutrition, in this context, relates to individuals who are within the 
overweight and underweight categories. Banku was also highly underestimated by 74.2% 
malnourished males compared to females (59.0%). Similarly, sugar was also highly 
underestimated by 87.5% and 81.5% malnourished males and females respectively. However, 
60.9% and 58.3% malnourished females and males correctly estimated portion sizes of boiled 
yam respectively (Table 4.6). Significant overestimation of boiled rice was observed in the 
malnourished females (P=0.044). 
 
 
41 
 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.5: BMI and gender specific estimations of the different types of carbohydrates 
foods 
% Estimation  
 B MI  Gender UE CE OE  P-value  
F ufu Underweight Male 1 (100) 0 (0.0) 0 (0.0) -  
Female 1 (100) 0 (0.0) 0 (0.0)  
 N ormal weight Male 28 (62.2) 8 (17.8) 9 (20.0)  0.425  
 Female 11 (45.8) 6 (25.0) 7 (29.2)  
  Overweight Male 13 (72.2) 3 (16.7) 2 (11.1) 0 .636  
 Female 15 (62.5) 7 (29.2) 2 (8.3)  
 O bese Male 4 (100) 0 (0.0) 0 (0.0) 0 .405  
 Female 10 (66.7) 3 (20.0) 2 (13.3)  
 Banku  Underweight Male 2 (100) 0 (0.0) 0 (0.0) 0 .361  
Female 2 (66.7) 1 (33.3) 0 (0.0)  
  Normal weight Male 43 (76.8) 5 (8.9) 8 (14.3) 0 .015  
 Female 18 (60.0) 10 (33.3) 2 (6.7)  
 O verweight Male 16 (69.6) 3 (13.0) 4 (17.4)  0.12  
 Female 12 (52.2) 9 (39.1) 2 (8.7)  
 O bese Male 5 (83.3) 0 (0.0) 1 (16.7) 0 .405  
 Female 9 (69.2) 3 (23.1) 1 (7.7)  
 Boiled yam  Underweight Male 0 (0.0) 2 (100) 0 (0.0)  -  
Female 0 (0.0) 2 (100) 0 (0.0)  
  Normal weight Male 6 (28.6) 9 (42.9) 6 (28.6)  0.580  
 Female 1 (11.1) 5 (55.6) 3 (33.3)  
  Overweight Male 4 (40.0) 5 (50.0) 1 (10.0)  0.047  
 Female 0 (0.0) 7 (63.6) 4 (36.4)  
  Obese Male 0 (0.0) 2 (100) 0 (0.0) 0 .424  
 Female 3 (30.0) 5 (50.0) 2 (20.0)  
 Pearson Chi-S quare *Statistically significant at P<0.05; CE: Correct estim ation; UE: 
underestimation; OE: Overestimation 
 
42 
 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.5 continued: BMI and gender specific estimations of the different types of 
carbohydrates foods 
% Estimation 
Carbohydrate 
food BMI Gender UE CE OE p-value 
Boiled rice Underweight Male 0 (0.0) 1 (100) 0 (0.0) 0.580 
Female 0 (0.0) 1 (100) 0 (0.0) 
  Normal weight Male 5 (15.2) 8 (24.2) 20 (60.6)  0.047 
 Female 1 (4.8) 4 (19.0) 16 (76.2) 
  Overweight Male 2 (15.4) 4 (30.8) 7 (53.8) 0 .424 
 Female 6 (30.0) 5 (25.0) 9 (45.0) 
  Obese Male 0 (0.0) 2 (66.7) 1 (33.3) 0 .216 
 Female 3 (37.5) 1 (12.5) 4 (50.0) 
S ugar U nderweight Male 1 (100) 0 (0.0) 0 (0.0)  0.227 
Female 1 (100) 0 (0.0) 0 (0.0) 
  Normal weight Male 22 (81.5) 3 (11.1) 2 (7.4) 0 .794 
 Female 12 (60.0) 6 (30.0) 1 (10.0) 
  Overweight Male 11 (84.6) 1 (7.7) 1 (7.7) 0 .571 
 Female 16 (80.0) 3 (15.0) 1 (5.0) 
  Obese Male 2 (100) 0 (0.0) 0 (0.0) 0 .377 
 Female 6 (85.7) 0 (0.0) 1 (11.1) 
P earson Chi-S quare *Statistically significant at P<0.05; CE: Correct estim ation; UE: 
underestimation; OE: Overestimation 
 
 
 
 
43 
 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.6: Gender and BMI association with carbohydrate food estimations 
Gender BMI UE CE OE P-value 
F ufu Male Normal 28 (62.2) 8 (17.8) 9 (20.0) 0.369 
Malnourished 18 (78.3) 3 (13.0) 2 (8.7) 
 F emale Normal 11 (45.8) 6 (16.2) 7 (29.2)  0.125 
 Malnourished 26 (65.0) 10 (25.0) 4 (10.0) 
B anku  Male Normal 43 (76.8) 5 (8.9) 8 (14.3)  0.963 
Malnourished 23 (74.2) 3 (9.7) 5 (16.1) 
  Female Normal 18 (60.0) 10 (33.3) 2 (6.7) 0 .986 
 Malnourished 23 (59.0) 13 (33.3) 3 (7.7) 
 Boiled yam M ale Normal 6 (28.6) 9 (42.9) 6 (28.6) 0 .385 
Malnourished 4 (33.3) 7 (58.3) 1 (8.3) 
  Female Normal 1 (11.1) 5 (55.6) 3 (33.3)  0.918 
 Malnourished 3 (13.0) 14 (60.9) 6 (26.1) 
B oiled rice M ale Normal 5 (15.2) 8 (24.2) 20 (60.6)  0.465 
Malnourished 2 (11.8) 7 (41.2) 8 (47.1) 
  Female Normal 1 (4.8) 4 (19.0) 16 (76.2)  0.044 
 Malnourished 9 (32.1) 6 (21.4) 13 (46.4) 
 Sugar  Male Normal 22 (81.5) 3 (11.1) 2 (7.4)  0.853 
Malnourished 14 (87.5) 1 (6.2) 1 (6.2) 
 F emale Normal 12 (60.0) 6 (11.1) 2 (10.0)  0.227 
 Malnourished 22 (81.5) 3 (11.1) 2 (7.4) 
 Pearson Chi- Square *Statistically significant at P<0.05; CE: Correct e stimation; UE: 
underestimation; OE: Overestimation 
 
44 
 
University of Ghana  http://ugspace.ug.edu.gh
4.4.8 Evaluation of the different carbohydrate foods by age ranges 
Apart from sugar which had a significant proportion of participants within the age range of 39 to 
49 years underestimating portion sizes (P=0.041; 93.8%), all other estimations were within age 
ranges were not statistically significant. Although not significant, a higher proportion of 
participants, (69.0%) within the age range of 39 to 49 years underestimated portion size of fufu 
followed by those within 50 to 60 years and 17 to 27 years indicated by 62.5% and 61.7% 
respectively. Similarly, higher proportion (75.0%) within the age range of 39 to 49 years 
underestimated portion size of banku compared to the other age ranges. Boiled yam on the other 
hand was estimated correctly by higher proportion (66.7%) of participants within the age range 
of 39-49 years followed by those within 17 to 27 years and 28 to 38 years as indicated by 47.1% 
and 42.9% respectively. Boiled rice was highly overestimated by the participants with the 61 to 
70-year range (100%) and 17 to 27-year range (65.2%). (Table 4.7) 
Table 4.7: Age specific estimations of the different types of carbohydrates foods 
Age N UE CE OE P-value 
F ufu 17-27 47 61.7 25.5 12.8 0.913 
28-38 36 58.3 22.2 19.4 
 39-49 29 69.0 13.8 17.2  
 50-60 16 62.5 18.8 18.8  
 61-70 4 75 0.0 25  
 Banku 17-27 57 68.4 22.8 8.8 0 .193 
28-38 40 68.4 22.8 8.8 
 39-49 32 75 10 15  
 50-60 21 71.9 15.6 12.5  
 61-70 6 50 16.7 33.3  
B oiled yam 17-27 17 29.4 47.1 23.5  0.694 
45 
 
University of Ghana  http://ugspace.ug.edu.gh
28-38 14 35.7 42.9 21.4 
 39-49 15 6.7 66.7 26.7  
 50-60 15 20 53.3 26.7  
 61-70 4 0 75 25  
B oiled rice 17-27 46 15.2 19.6 65.2  0.226 
28-38 24 12.5 25 62.5 
 39-49 15 20 40 40  
 50-60 11 36.4 36.4 27.3  
 61-70 3 0 0 100  
 Sugar 17-27 41 63.4 22 14.6 0 .041 
28-38 26 92.3 7.7 0 
 39-49 16 93.8 0 6.2  
 50-60 7 71.4 28.6 0  
 61-70 0 0 0 0  
P earson Chi-Square *Statistically significant at P<0.05; CE: Correct estim ation; UE: 
underestimation; OE: Overestimation 
4.4.9 Evaluation of the different carbohydrate foods by age ranges and gender 
Among the age range of participants, all males within the age range 50 to 60 years and 61 to 70 
years underestimates portion sizes compared to females in the same years range. The males 
within the age range 39 to 49 years and females within the age range of 28 to 38 years highly 
underestimated portion size of banku compared to the other age ranges. Boiled yam on the other 
hand was estimated correctly by higher proportion (62.8%) of male participants within the age 
range of 28 to 38 years and females (75.0%) within the age range of 39 to 49 years (Figure 
4.11a).  
Boiled rice was highly overestimated by the male (100%) and female (100%) participants with 
the 61 to 70-year range followed by 17 to 27-year range (68.2%) for males and 28 to 38 year 
46 
 
University of Ghana  http://ugspace.ug.edu.gh
range for females. Underestimation of sugar was seen among a statistically significant 
percentage (P=0.041; 100%) of female participants within the 28 to 38-year range compared the 
males of same age range (Figure 4.11b).  
Table 4.8 shows that higher proportion of young adult (76.9%) and older adult (55.2%) males 
underestimated portion size of fufu with no significant association (P = 0.231). Young adults 
refers to participants within the age bracket of 17 – 49 years, and older adults within the ages of 
50 – 70 years. Similarly, higher proportion of young and older adult males underestimated 
portion size of banku with no significant association for young adult (P = 0.059) but significant 
for old adults. Portion size of boiled yam was correctly estimated by 54.5% young females and 
70.0% old females with no significant association. Boiled rice on the other hand was 
overestimated by 62.2% young adult females and 70.8% older adult males. Approximately 83% 
and 84% of young and older adult males underestimated portion size of banku with no 
significant association (Table 4.8). 
 
 
 
 
 
 
 
47 
 
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.11(a) Age and gender specific estimations of the different types of carbohydrates 
foods 
a 120
100
80
60
40
20
0
Male Female Male Female Male Female
Fufu Banku Boiled yam
Food/Gender/Age
UE CE OE
 
Pearson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
 
Figure 4.11(b) Age and gender specific estimations of the different types of carbohydrates 
foods
b 120
100
80
60
40
20
0
Male Female Male Female
Boiled rice Sugar
Food/Gender/Age
UE CE OE
 
Pearson Chi-Square *Statistically significant at P<0.05; CE: Correct estimation; UE: 
underestimation; OE: Overestimation 
48 
 
Percent estimation Percent estimation
17-27 17-27
28-38
28-38 39-49
39-49 50-60
61-70
50-60
17-27
61-70 28-38
39-49
17-27
50-60
28-38 61-70
39-49 17-27
28-38
50-60 39-49
61-70 50-60
61-70
17-27 17-27
28-38 28-38
39-49
39-49
50-60
50-60 61-70
17-27
61-70
28-38
17-27 39-49
28-38 50-60
61-70
39-49 17-27
50-60 28-38
39-49
61-70 50-60
61-70
University of Ghana  http://ugspace.ug.edu.gh
Table 4.8: Age and gender specific estimations of the different types of carbohydrates foods 
 
Age categories Gender UE CE OE P-value 
 Fufu Young Adult Male 30 (76.9) 5 (12.8) 4 (10.3) 0.231 
Female 28 (59.6) 11 (23.4) 8 (17.0) 
 O ld Adult Male 16 (55.2) 6 (20.7) 7 (24.1)  0.756 
 Female 9 (52.9) 5 (29.4) 3 (17.6) 
 Banku  Young Adult Male 40 (75.5) 5 (9.4) 8 (15.1)  0.059 
Female 29 (61.7) 13 (27.7) 5 (10.6) 
  Old Adult Male 26 (76.5) 3 (8.8) 5 (14.7)  0.003 
 Female 12 (54.5) 10 (45.5) 0 (0.0) 
B oiled yam Y oung Adult Male 7 (35.0) 7 (35.0) 6 (30.0) 0 .36 
Female 4 (18.2) 12 (54.5) 6 (27.3) 
  Old Adult Male 3 (23.1) 9 (69.2) 1 (7.7)  0.14 
 Female 0 (0.0) 7 (70.0) 3 (30.0) 
 Boiled rice  Young Adult Male 3 (11.5) 12 (46.2) 11 (42.3)  0.12 
Female 6 (16.2) 8 (21.6) 23 (62.2) 
 O ld Adult Male 4 (16.7) 3 (12.5) 17 (70.8) 0 .44 
 Female 4 (22.2) 5 (13.9) 23 (63.9) 
 Sugar Y oung Adult Male 20 (83.3) 2 (8.3) 2 (8.3)  0.668 
Female 25 (73.5) 5 (14.7) 4 (11.8) 
 O ld Adult Male 16 (84.2) 2 (10.5) 1 (5.3) 0 .274 
 Female 9 (69.2) 4 (30.8) 0 (0.0) 
P earson Chi-Sq uare *Statistically significant at P<0.05; CE: Correct estima tion; UE: 
underestimation; OE: Overestimation 
 
 
 
49 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 5 
5.0 DISCUSSION 
5.1 General profile of the participants 
5.1.1 Demographics 
Although the distribution of male participants (53.2%; 101) in the study as compared to the study 
carried out in Accra by Peprah Boateng (2014) (49.6%; 401) was higher, they were comparable 
due to the smaller sample size in this study. However there is a slightly higher male population 
(50.6%; 36,500) in the entire Upper Manya Krobo District as compared to the female population 
(49.4%; 35,592). Majority of the study population were in their youthful years and had some 
level of formal education thus had a fair knowledge and understanding of the purpose of the 
study when explained to them by the researcher.  
5.1.2 Anthropometric measurement 
The females showed higher weight than the males. Similarly mean BMI for the females was 
higher than that of males indicating that the average female among the study participant was 
overweight and the average male was within the normal range of BMI. Obesity was observed in 
a greater proportion of females as compared to their male counterparts. Participants within the 
age group of 18 – 28 years reported the highest number of normal weight individuals as well as 
overweight individuals. Birirtwum, Gyapong, & Mensah (2005) reported in their study that 
obesity trends were highest in the Greater Accra region as compared to the other regions in the 
country, and this could have accounted for the differences in the mean BMI of this study which 
recorded an average of 24.94 kg/m² in relation to the study done in Accra by Peprah Boateng 
(2014) of an average BMI of 26.14 kg/m². 
50 
 
University of Ghana  http://ugspace.ug.edu.gh
5.2 Common carbohydrates foods consumed by the participants 
The Ghanaian diet is mainly composed of carbohydrate based foods; starchy roots (43%) and 
cereals excluding beers (29%) providing almost three- quarters (72%) of the daily energy 
expenditure (FAO, 2009).This huge contribution of carbohydrates in the daily energy supply 
(DES) of Ghanaians indicates the important role starchy roots and cereals play in the Ghanaian 
diet. Although according to (FAO, 2009) sugar contributes only 3% of DES it has been linked to 
the incidence of NCD’s and therefore its inclusion in the collation of the common carbohydrate 
foods (FAO, 2009). The study identified Banku, Fufu, boiled rice, boiled yam and Sugar 
(granulated) as commonly consumed carbohydrate foods. The foods identified vary from the 
study conducted in Accra where a larger pool of carbohydrate foods was used. This is indicative 
of varied tastes of the sampled population. In comparison to the study that was done in Accra, 
the foods that were considered as commonly consumed carbohydrates foods made up 5% or 
more of the frequency of the total carbohydrate foods consumed within the three day 24-hour 
recall. 
5.3 Overall estimation of carbohydrate food 
Accurate information about intake of individuals and populations is particularly difficult to 
obtain; it is usually reliant upon self-report and can be subject to large errors. Currently, it has 
been observed that people can estimate healthy portion sizes either by using serving size 
information printed on food packages or by making visual comparisons with various objects. 
Unfortunately, both methods present problems for low literacy populations (Gibson et al., 2016). 
The other option will be for people to determine portion sizes by associating food portions with 
visual aids recommended by registered dietitians (Brrd-Bredbenner & Schwartz , 2004). In 
Ghana, household measurement aids are the dietary aids of choice in portion size estimation by 
51 
 
University of Ghana  http://ugspace.ug.edu.gh
dietitians and nutrition-related professionals. An overall correct estimation of foods has been 
reported in other studies to range from 40% to 70% (Lazarte C. , 2013, Nelson et al., 1994). The 
findings in this study shows that a statistically significant number of participants (63.7%) 
underestimated portion sizes, with 28.9% overestimating and 5.3% correctly estimating (P< 
0.0001). This outcome is contrary to the findings of the study done in Accra by Peprah Boateng 
(2014) which showed that a statistically significant number of participants (54.17%) correctly 
estimated their portion sizes, with 29.16% underestimating and 16.67% overestimating with a P-
value of 0.003. The difference between the results from the two studies may lie in the fact that 
this study relied on the participants’ ability to recall portion sizes with household measures 
against the food photographic atlas which is in contrast to that of Peprah Boateng (2014) where 
the actual foods were present for participants to pick from. The differences furthermore highlight 
the findings of Subar et al., (2010), that portion size estimation is affected by perception, 
conceptualization and memory.  Perception refers to one’s ability to relate an amount of food 
present in reality to the amount of the same food presented by the portion-size aid. 
Conceptualization can be defined as the ability to create a mental picture of a food portion which 
may not be present by the aid of a portion size estimation tool, and is dependent on memory. 
Memory is one’s ability to recall accurately the amount of food eaten over a period of time 
(Subar et al., 2010) 
Ovaskainen et al., (2008) also reported a 50% correct estimation in their study.  However, in 
contrast with the study results, Ovaskainen et al., (2008) claimed a tendency for subjects to 
overestimate more than underestimate. Studies by Hernandez et al., (2006), Turconi et al., 
(2005) and Frobisher & Maxwell, (2003) recorded a higher level of overestimates than 
underestimation.    
52 
 
University of Ghana  http://ugspace.ug.edu.gh
5.3.1 Gender and carbohydrate food estimations 
According to this current study overall underestimation by males although not statistically 
significant (66.3%), was higher than females (60.7%), confirming studies by Ovaskainen et al., 
(2008) and Nelson et al., (1994) who also claimed in their studies that men tend to underestimate 
food portions when compared to women. Although males (29.7%) in this study had a tendency to 
overestimate more than females (28.1%), this was also statistically insignificant. This, however, 
conforms to a study by Burger et al., (2007) who claimed that overestimation of certain foods is 
seen more in males than females.  
Ovaskainen et al., (2008), observed that, significantly higher proportion of females correctly 
estimated food portions. This is somewhat similar to the finding in this study, where 7.9% of the 
females correctly estimated the food portions compared to the 3.0% of males. This can be 
attributed to females’ experience in cooking resulting in a better knowledge of food items and 
portion sizes. The Ghanaian female subjects’ knowledge in shopping and cooking obviously may 
have had an effect on their ability to correctly estimate portion sizes of a variety of foods. Males 
on the other hand, in a typical Ghanaian society are served their meals; hardly go grocery 
shopping and rarely cook. These may have contributed to their reduced ability to correctly 
estimate portion size (Peprah Boateng, 2014).  
The study results also showed underestimation of bigger portions of foods by a higher proportion 
of male and female participants except for banku and sugar. Among gender, a statistically 
significant higher proportion of females (74.2%) correctly estimated banku compared to males 
(25.8%) (P=0.001). Boiled yam on the other hand was underestimated by higher proportion of 
males (71.4%) compared to females (28.6%), whilst 60% of males correctly estimated boiled 
53 
 
University of Ghana  http://ugspace.ug.edu.gh
yam compared to females (40.0%). A significantly higher proportion of males either 
underestimated or overestimated banku and fufu. The different sizes in which these foods are 
molded might have contributed to the poor estimations by males because males usually do not 
cook or serve themselves; these foods are usually cooked and served as one big portion by 
females.   
5.3.2 Body Size and estimation of portion size 
The current study showed that BMI categories were not significantly associated with correct 
estimations and under estimations of all carbohydrate foods. Underestimation was identified 
among a statistically higher proportion (80%, P<0.05) of underweight participants (BMI < 18.5 
kg/m²) when compared to the other BMI categories. In comparison to the study carried out in 
Accra by Peprah Boateng (2014) there was also no statistical significance in correct and under 
estimations of carbohydrate foods based on BMI categories, however a record of a significantly 
higher proportion of obese participants overestimated.  Contrary to results of this study, findings 
from other studies such as Okubo & Sasaki (2004) reported an association between BMI and 
portion size estimated or consumed stating that obese participants highly underestimated portion 
sizes. Wasink (2007) in his book, wrote that even people of normal weight underestimate their 
food intake by about 20 per cent and those who are overweight, can underestimate what they eat 
by more than 50 per cent. The reason he gave was that it was possible for people to be easily 
fooled into consuming more than they believe they have eaten due to what he termed as ‘portion 
creep’ which simply refers to the gradual increase of food portion sizes over the years. This 
might somewhat account for the higher underestimation (79.2%) among obese participants. 
Correct estimation although not significantly related to food portion sizes, was higher in 
underweight (20.0%) participants followed by overweight participants (6.8%). According to 
54 
 
University of Ghana  http://ugspace.ug.edu.gh
Lara, Scott, & Lean, (2004) overweight people may possibly under-report food recall over and 
over again to avoid feeling embarrassed. Vinai (2011) asserted that overweight and obese 
individuals were inclined to underestimate their portion sizes more than normal weight people. 
Although from the present study results, overall estimation of carbohydrate foods was not 
significantly dependent on BMI, the higher proportion of overweight (3.9%) and obese (4.2%) 
people correctly estimated food portions. 
In assessing participants BMI specific estimation of the different carbohydrate foods in the study, 
obese females tended to underestimate fufu as did normal weight males. An equal proportion 
(100%) of males to females underestimated portion sizes of fufu with no statistical significance. 
Contrarily to the study results, some previous studies indicated that underweight individuals had 
the inclination to overestimate their food intake as indicative of patients with anorexia nervosa 
(Vinai, 2011). These participants tend to unusually exaggerate their food intake (Milos et al., 
2012) due to their reported weight problems.  
Male participants within all BMI categories in this study exhibited a higher percentage of 
underestimation of fufu and banku with the exception of boiled yam, boiled rice and sugar where 
portion sizes were overestimated. In the obese (≥30 kg/m2) category, when compared to females 
however, this was not statistically significant. The male participants’ underestimation of portion 
sizes may be because males have been known to consume more food than females (Jeffrey et al., 
2007) and so they perceived the actual portion size presented as small when compared to their 
female counterparts. 
55 
 
University of Ghana  http://ugspace.ug.edu.gh
5.3.3 Age and food portion size estimation 
Findings form the study shows higher proportion of participants aged between 61 to 70 years to 
correctly estimated portion sizes. Participants with this age range have probably had a longer 
period to either grocery shop, cook, serve food or observe these actions being done as stated by 
Ovaskainen et al., (2008). Least correct estimation was found among participants aged 17 to 27 
years. Participants aged 61-70 years old in this study highly overestimated portion sizes. This 
conforms with a study by Nelson et al., (1996) who concluded that portion size overestimation 
on average was significantly higher for over 65year old men. Over estimation and 
underestimations were not significant statistically.  
Underestimation of sugar was seen among a statistically significant percentage (P=0.041; 100%) 
of female participants within the 28 to 38-year range. This is because it is said that women have 
higher cravings to sugar and sugary products such as chocolates and pastries whereas men would 
rather choose meat and meat products over sugary products (Wiseman, 2010). 
 
 
 
 
 
 
 
56 
 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER 6 
6.1 Conclusion   
The demographic profile showed more males than females across the ages of 1 to 70 years with 
majority of the participants aged between 18 – 38 years. Majority of the participants had 
middle/Junior secondary school education. Forty-nine percent of participants had monthly 
income below two hundred Ghana cedis (₵200). Anthropometric measures showed a higher 
population within overweight and obesity category representing 31.1% and 12.6% respectively. 
Obesity was prevalent among participants within the age brackets of 61-70 years (28.5%) and 
39-49 years (25.6%). 
The study identified Banku, Fufu, boiled rice, boiled yam and Sugar (granulated) as commonly 
consumed carbohydrate foods. Portion size estimation of these carbohydrate foods using the 
photographic food atlas and the household measures showed an overall correct estimation of 
5.3%, under and overestimation of 63.7% and 28.9% respectively.  
Gender showed no significant effect on portion size estimation although more females did 
estimate carbohydrate food correctly compared to the males. 
The effects of the different BMI categories on participants’ ability to correctly estimate or 
overestimate were not significant. However, significant effect of BMI on underestimation of 
portion size in some foods was observed.  
Age association showed no significant effect on portion size estimation of carbohydrate food. 
57 
 
University of Ghana  http://ugspace.ug.edu.gh
6.2 Limitations 
A number of factors that could not have been effectively controlled during the study may have 
had an effect on a participant’s ability to correctly estimate portion sizes, hence the resulting 
differences in portion size estimations. 
These factors include: 
 Participants reported that they had difficulty in estimating portion sizes of some of the 
foods in the photographic atlas because the pictures were not clear. 
 The use of household measures in estimating portion sizes of some carbohydrate foods 
were different to the type used in the food photographic atlas.  
6.3 Recommendations 
The following recommendations to be made:  
1. The pictures of portion sizes should be clear enough to make correct estimations. 
2. The household measure used in the portion sizes of the study should correspond to 
household measures used in the development of the photographic food atlas. 
3. The food photographic atlas should be used together with the house hold measuring aids 
in diet counseling. 
4. In order to help curb the rising NCDs in Ghana, appropriate portion size education should 
be incorporated into nutrition intervention programs. 
 
 
 
58 
 
University of Ghana  http://ugspace.ug.edu.gh
REFERENCES 
Fagúndez, L. M., Torres, A. R., Sánchez, M. G., Lourdes de Torres Aured, M., Rodrigo, C., & 
Irles Rocamora, J. (2015). Diet History: Method and applications. Nutrition Hospital , 
57-61. 
Ollberding, N. J., & Wolf , R. L. (2010). Contento I. Food label use and its relation to dietary 
intake among US adults. . American Dietetics Association, 1238-1241. 
Almiron-Roig, E., Solis-Trapala, I., Dodd , J., & jebb, S. (2013). Estimating food portions. 
Influence of unit number, meal type and energy density. Elsevier, 95-103. 
Anti, C. (2008, March 31). Household dietary practices and family nutritional status in rural 
Ghana. Retrieved March 26, 2016, from http://www.ncbi.nlm.nih.gov: 
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2815306/ 
Atkinson , M., Darbyshire, S., & Nelson, M. (1994). Food Photography I: the perception of food 
portion size from photographs. British Journal of Nutrition. 
Ball, G. (2014, February 19). Efficacy of Portion Size Measurement Aids (PSMA). Retrieved July 
17, 2016, from ClinicalTrials.gov: https://clinicaltrials.gov/ct2/show/NCT01262768 
Ball, G. D., & Friedman , A. (2010). Dice, golf balls, and CDs: assumptions about portion size 
measurement aids. Canadian Journal of Dietetic practice and research, 146-9. 
Birirtwum, R., Gyapong, J., & Mensah, G. (2005). The Epidemiology of Obesity in Ghana. 
Ghana Medical Journal, 82-85. 
Bosu, W. K. (2010). Epidemic of hypertension in Ghana: a systematic review. BMC Public 
Health, 10;418. 
Bosu, W. K. (2013, MARCH). ACCELERATING THE CONTROL AND PREVENTION OF 
NON-COMMUNICABLE DISEASES IN GHANA: THE KEY ISSUES. POST 
GRADUATE MEDICAL JOURNAL OF GHANA, 2. 
Boushey, C. J., Kerr, D. A., Wright, J., Lutes, K. D., Ebert, D. S., & Delp, E. J. (2009). Use of 
technology in children's dietary assessment. European Journal of Clinical Nutrition. 
Brent, M. (2016, November 9). 6 factors that influence our food choices. Retrieved January 2, 
2017, from Leaf: https://www.leaf.tv/articles/6-factors-that-influence-our-food-choices/ 
Brrd-Bredbenner, C., & Schwartz , J. (2004). The effect of practical portion size measurement 
aids on the accuracy of portion size estimates made by young adults. Joournal of human 
nutrition and dietetics, 351-7. 
59 
 
University of Ghana  http://ugspace.ug.edu.gh
Burger, K., Kern, M., & Coleman, K. (2007). Characteristics of self-selected portion size in 
young adults. American Dietetic Association, 611-8. 
Burrows, T. L., Martin , R. J., & Collins, C. E. (2010). A systematic review of the validation of 
dietary assessment methods in children when compared with the method of doubly 
labeled water. Journal of the Academy of Nutrition and Dietetics, 1501-1510. 
Cashin-Garbutt, A. (2017, October 30). NEWS MEDICAL LIFE SCIENCES. Retrieved 
December 31, 2017, from Obesity and caloric intake: www.new-
medical.net/health/Obesity-and-Fast-Food.aspx 
Chaudry, B., Connelly, K., Siek, K. A., & Welch, J. L. (2011). The Design of a Mobile Portion 
Size Estimation Interface for a Low Literacy Population. PMC, 160-167. 
Collins, C. E., Watson , J., & Burrows, T. (2009). Measuring dietary intake in children and 
adolescents in the context of overweight and obesity. International Journal of Obesity, 
1103-1115. 
Cypel, Y. S., Guenther, P. M., & Petot, G. J. (1997). Validity of Portion-Size Measurement Aids. 
Journal of the American Dietetic Association, 289-92. 
Dake, F., Tawiah, E., & Badasu, D. (2010). Socio-demographic correlates of obesity among 
Ghanaian women. Public Health Nutrition, 1285-1291. 
de Onis, M., Blössner, M., & Borghi, E. (2010). Global prevalence and trends of overweight and 
obesity among preschool children. The American Journal of Clinical Nutrition, 1257-
1264. 
Duffey, K. J., & Popkin, B. M. (2011). Energy Density, Portion Size, and Eating Occasions: 
Contributions to Increased Energy Intake in the United States, 1977–2006. PLOS Med, 
8:e1001050. 
FAO. (2009, November 29). Nutrition Country Profile. Accra: Nutrition and Consumer 
Protection Division, FAO, 2009. 
Faulkner GP et al. (2012). Serving size guidance for consumers: is it effective? Proc Nutr Soc,. 
Frobisher, C., & Maxwell, S. (2003). The estimation of food portion sizes: a comparison between 
using descriptions of portion sizes and a photographic food atlas by children and adults. 
Journal of Human Nutrition and Dietetics, 181-188. 
Ghana Statistical Service. (2010). Population and Housing Census- Final Results. Accra: Ghana 
Statistical Service. 
60 
 
University of Ghana  http://ugspace.ug.edu.gh
Gibson, A. A., Hsu, M. S., Rangan, A. M., Seimon, R. V., Lee, C. M., Das, A., et al. (2016). 
Accuracy of hands v. household measures as portion size estimation aids. Journal of 
nutritional Science. 
Godwin, S., Chambers, E., & Cleveland, L. (2004). Accuracy of reporting dietary intake using 
various portion-size aids in-person and via telephone. US National Library of Medicine 
National Institutes of Health , 585-594. 
Harris-Fry, H., Puskar , P., Karn, M., Mishra, N., Paudel, V., Harrisson, T., et al. (2015). 
Development and validation of a photographic food atlas for. Public Health Nutrition, 1-
13. 
Hernandez, T., Wilder, L., Kuehn, D., & Robotsky, K. (2006). Portion size estimation and 
expectation of accuracy. Journal of food composition and analysis, 14-21. 
Hight, K. (2008). Food Portion Size Perceptions of Various Athletes at the University of 
Wisconsin - Stout. Wisconsin. 
Hill AG, D. R. (2007). Health of urban Ghanaian women as identified by the Women's Health 
Study of Accra. . Int J Gynaecol Obstet, 150-156. 
Jeffrey, R., Rydell, S., Dunn, C. L., Harnack, L. J., Levine, A. S., Pentel, P. R., et al. (2007). 
Effects of portion size on chronic energy intake. International Journal of Behavioral 
Nutrition and Physical Activity. 
Jia, W., Chen, H.-C., yue, Y., Li, Z., Fernstrom, J., Ba, y., et al. (2014). Accuracy of food portion 
size stimation from digital picturs acquired by a chest-worn camera. Public Health 
Nutriition, 1671-1681. 
Johansson, G. (2006). Dietary assessments: Use, design concepts, biological markers, pitfalls and 
validation. 
Karvetti, R., & Knuts, L. (1985). Validity of the 24-hour dietary recall. American Dietitics 
Association, 1437-1442. 
Kristal, A. R., Shattuck, A. L., & Williams, A. E. (1994). Food Frequency Questionnaires for 
Diet Intervention Research. The American Journal . 
Lara, J., Scott, J., & Lean, M. (2004). Intentional mis-reporting of food consumption and its 
relationship with body mass index and psychological scores in women. Journal of 
Human Nutrition and Dietetics, 209-18. 
Lazarte, C. (2013). Nutritional Assessment in a Rural Area of Bolivia. A Study of Zinc and Iron 
Deficiencies and. Lund: Lund University. 
61 
 
University of Ghana  http://ugspace.ug.edu.gh
Lazarte, C. E., Encinas, M. E., Alegre, C., & Granfeldt. (2012). Validation of digital 
photographs, as a tool in 24-h recall, for the improvement of dietary assessment among 
rural populations in developing countries. Nutrition Journal. 
Ledikwe, J. H., Ello-Martin , J. A., & Rolls, B. J. (2005). Portion sizes and the Obesity 
Epidemic. The Journal of Nutrition. 
Lillegaard, I. T., Øverby, N. C., & Andersen, L. F. (2005). Can children and adolescents use 
photographs of food to estimate portion sizes? European Journal of Clinical Nutrition, 
611-617. 
Livingstone, M. E., & Pourshahidi, L. (2014). Portion Size and Obesity. American Society of 
Nutrition, 829-834. 
Livingstone, M., & Robson, P. (2000). Measurement of dietary intake in children. National 
Center for Biotechnology Information, U.S. National Library of Medicine, 279-293. 
Macdiarmid JI, B. J. (1997). Dietary under-reporting: what people say about recording their food 
intake. European Journal of Clinicial Nutrition, 51, 199-200. 
Martinovich-Stanford, M. (2017, June 13). Scribd. Retrieved 3 July, 2017, from These words get 
people to eat more: https://www.scribd.com/article/351218167/These-Words-Get-People-
To-Eat-More-Vegetables 
McGuire, B., Chambers, E., Godwin, S., & Brenner, S. (2001). Size Categories Most Effective 
For Estimating Portion Size of Muffins. Journal of the American Dietetic Association, 
470-472. 
Ministry of Health. (2011). National Policy for the Prevention and Control of Chronic Non-
Communicable Diseases in Ghana. Accra. 
Muller, O., & Krawinkel, M. (2005). Malnutrition and health in developing countries. Canadian 
Medical Association Journal, 173,279-286. 
Naska, A., Lagiou, A., & Lagiou, P. (2017). Dietary assessment methods in epidemiologic 
research: current state of the art and future prospects. F1000 Research. 
Nelson, M., & Haraldsdottir, J. (1998). Food photographs: practical guidelines I. Design and 
analysis of studies to validate portion size estimates. Public Health Nutrition, 219-230. 
Ofori-Asenso, R., & Garcia, D. (2016). Cardiovascular diseases in Ghana within the context of 
globalization. Cardiovascular Diagnosis and Therapy, 67-77. 
62 
 
University of Ghana  http://ugspace.ug.edu.gh
Okubo, H., & Sasaki, S. (2004). Underreporting of energy intake among Japanese women aged 
18-20 years and its association with reported nutrient and food group intakes. Public 
Health Nutrition, 911-7. 
Ovaskainen, M., Paturi, M., Reinivuo, H., Hannila, M., Lehtisalo, J., Pynnönen-Polari, O., et al. 
(2008). Accuracy in the estimation of food servings against the portions in food 
photographs. European Journal of Clinical Nutrition, 674-681. 
Owusu, A. A., Ankrah, M., Boateng , P. G., & Anteh, G. (1995). Food weights/handy measures 
tables. Accra: The Dreyfus Health Foundation of New York. 
Peprah Boateng, G. (2014). Develpoment of a photographic food atlas with portion sizes of 
commonly consumed carbohydrate foods in Accra, Ghana. Accra. 
Piernas, C., & Popkin, B. M. (2011). Increased portion sizes from energy-dense foods affect total 
energy intake at eating occasions in US children and adolescents: patterns and trends by 
age group and sociodemographic cbaracteristics. The American Journal of Clinical 
Nutrition. 
Raina, S. K. (2013). Limitations of 24-hour Recall Method: Micronutrient Intake and the 
Presence of the Metabolic Syndrome. North American Journal of Medical Sciences. 
Riis, J. (2014). Opportunities and barriers for smaller portions in food service: lessons from 
marketing and behavioral economics. International Journal of Obesity, S19-S24. 
Rolls, B., Roe, L., & Meengs, J. (2006). Larger portion sizes lead to a sustained increase in 
energy intake over 2 days. National Center for Biotechnology Information, U.S. National 
Library of Medicine, 543-549. 
Rothman , R. L., Housam , R., Weiss, H., Davis , D., Gregory, R., Gebretsadik, T., et al. (2006). 
Patient understanding of food labels: the role of literacy and numeracy. American Journal 
of Preventive Medicine, 391-8. 
Sánchez-García,, S., García-Peña, C., Duque-López, M., Juárez-Cedillo, T., Cortés-Núñez, A., & 
Reyes-Beaman, S. (2007). Anthropometric measures and nutritional status in a healthy 
elderly population. BMC Publlic Health. 
Scinta, W. (2016, April 28). The History of Portion Sizes: How they've changed over time. 
Retrieved July 17, 2016, from Your Weight Matters: 
http://www.yourweightmatters.org/portion-sizes-changed-time/ 
Shim, J.-S., Oh, K., & Chang Kim, H. (2014). Dietary assessment in epidemiologic studies. 
Epidemiol Health. 
63 
 
University of Ghana  http://ugspace.ug.edu.gh
Shim, J.-S., Oh, K., & Chang Kim, H. (2014). Dietary assessment methods in epidemiologic 
studies. Epidemiol Health. 
Subar, A. F., Crafts, J., Zimmerman, T. P., Wilson, M., Mittl, B., Islam, N. G., et al. (2010). 
Assessment of the Accuracy of Portion Size Reports Using Computer-Based Food 
Photographs Aids in the Development of an Automated Self-Administered 24-Hour 
Recall. American Dietetic Association, 55-64. 
Thompson, F. E., & Subar, A. F. (2013). Nutrition in the prevention and treatment of diseases. In 
A. M. Coulston, C. J. Boushey, & M. G. Ferruzi, Nutrition in the Prevention and 
Treatment of Diseases (pp. 5-12). 32 Jamestown Road, London NWI 7BY, UK: Elsevier 
Inc. 
Tontisirin, K., & Henry, F. J. (2004). Uses of Food Consumption and Anthropometric Surveys in 
the Caribbean. Rome: FAO. 
Truesdell, D. (2004). diet.com. Retrieved october 20, 2015, from diet.com: 
http://www.diet.com/g/dietary-assessment 
Tueni, M., Mounayar, A., & Birlouez-Aragon, I. (2012). Development and evaluation of a 
photographic atlas as a tool for dietary assessment studies in Middle East cultures. Public 
Health Nutrition. 
Turconi, G., Guarcello, M., Berzolari, G. F., Carolei, A., Bazzano, R., & Roggi, C. (2005). An 
evaluation of a colour food photography atlas as a tool for quantifying food portion size 
in epidemiological surveys. European Journal of Clinical Nutrition. 
USDA. (2015). HEALTHY PORTION SIZES. Washington, DC: USDA Center for Nutrition 
Policy and Promotion. 
USDA. (2016, January 7). ChooseMyPlate. Gov. Retrieved March 26, 2016, from 
http://www.choosemyplate.gov/: http://www.choosemyplate.gov/ 
Vinai, P. (2011). Influence of Cognitive Biases in Visual Evaluation of Food Amount in Patients 
Affected by Eating Disorders. In V. R. Preedy, R. W. Watson, & C. R. Martin, Handbook 
of Behavior, Food and Nutrition (pp. 1245-). New York: Springer Science+. 
Wasink, B. (2007). Mindless Eating Why We Eat More Than We Think. New York: Bantam Dell. 
Weber, J. L., Tinsley, A. M., Houtkooper, L. B., & Lohman, T. G. (1997). Multi method training 
increases portion-size estimation accuracy. Journal of the American Dietetic Association, 
176-179. 
WHO. (2014). GLOBAL STATUS REPORT on noncommunicable diseases. Geneva: WHO. 
64 
 
University of Ghana  http://ugspace.ug.edu.gh
Wiseman, E. (2010, October 17). The truth about men, women and food. Retrieved December 
10, 2017, from the guardian: 
https://www.theguardian.com/lifeandstyle/2010/oct/17/gender-eating-men-women 
Wojtusiak, J., CA, G., & LR, P. (2011). DIETARY ASSESSMENT IN AFRICA: 
INTEGRATION WITH INNOVATIVE TECHNOLOGY. African Journal of Food, 
Agriculture, Nutrition and Development. 
World Health Organization. (2010). Global Status Report on Non-Communicable Diseases. 
Geneva: World Health Organization. 
Wrieden, W., Peace, H., Armstrong, J., & Barton, K. (2003). A short review of dietary 
assessment methods used in National and Scottish Research Studies. Working Group on 
Monitoring Scottish Dietary Targets Workshop.  
Wunderlich, S. (2013). The Importance of appropriate Nutrition Assessment and Nutrition 
Education for Older Adults. Journal of Nutrition and Food Sciences. 
Young, L. R., & Nestle, M. (2012). Reducing portion sozes to prevent obesity: a call to action. 
Am J Prev Med., 43(5):565-8. 
Yunsheng, M. A., Olendzi, B. C., Pagoto, S. L., Hurley, T. G., Magner, R., Ockene, I. S., et al. 
(2009). Number of 24-Hour Diet Recalls Needed to Estimate Energy Intake. National 
Institute of Health, 553-559. 
Ziraba, A., JC, F., & R, O. (2009, December 15). Overweight and obesity in urban Africa: A 
problem of the rich or the poor? Retrieved from National Center for Biotechnology 
Information, U.S. National Library of Medicine: 
https://www.ncbi.nlm.nih.gov/pubmed/20003478 
 
 
 
 
 
 
65 
 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX 1 
Research Participant Information Sheet 
DEPARTMENT OF NUTRITION AND DIETETICS 
 SCHOOL OF ALLIED HEALTH SCIENCES 
 COLLEGE OF HEALTH SCIENCES UNIVERSITY OF GHANA 
Title of study: “A comparative study between house hold measures and food photographic atlas 
in dietary assessment in a rural area in Ghana” 
 Kate Opoku, MSc. Student and Dr. Gladys Peprah Boateng of the Department of Nutrition and 
Dietetics, School of Biomedical and Allied Health Sciences are conducting a research project 
titled “A comparative study between house hold measures and food photographic atlas in dietary 
assessment in a rural area in Ghana.” The main aim of the study is for participants to provide a 3 
day dietary history using house hold measures and identify their food portions from the food 
photographic atlas. The information will then be used to analyze the nutrient intakes of 
participants using these two methods. 
We would like to invite you to participate in this study. This study is not expected to cause any 
medical or social risk to you. The information gathered will be kept strictly confidential and any 
reports of the findings of this research will not contain your name or any other identifying 
information. Your participation in this project is completely voluntary. If at any time you wish to 
withdraw from this research, you may do so without coercion or prejudice. Just inform any of the 
researchers. Once the study is completed, the analyzed findings would be available to you upon 
66 
 
University of Ghana  http://ugspace.ug.edu.gh
request. Questions or concerns about participation in the research or subsequent complaints 
should be addressed first to the researchers or research advisors. 
Dr. Gladys Peprah Boateng on telephone number: 024-4265-4436: email- 
nitramharpep@yahoo.com 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67 
 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX II 
Research Participant Informed Consent Form  
DEPARTMENT OF DIETETICS 
SCHOOL OF ALLIED HEALTH SCIENCES 
COLLEGE OF HEALTH SCIENCES UNIVERSITY OF GHANA 
 I …………………………………………… understand that my participation in this study is 
strictly voluntary and I may discontinue my participation at any time without prejudice. I 
understand that the purpose of this study is to provide a 3 day dietary history using house hold 
measures and identify food portions from the food photographic atlas, which will be used in 
analyzing nutritional intakes. I further understand that any information about me that is collected 
during this study will be held in the strictest confidence and will not be part of my permanent 
record. I understand that in order for this research to be effective and valuable, some 
demographic information will need to be collected. I also understand that the strictest 
confidentiality will be maintained throughout this study and that only the researchers will have 
access to information that I supply on surveys or in interviews. I understand that at the 
conclusion of this study all records will be destroyed. I am aware that I will not be waiving my or 
any legal or human rights by agreeing to this participation. By signing below, I verify that I am 
18 years of age or older, in good mental and physical condition, and that I agree to and 
understand the conditions listed above.  
----------------------------------------    ----------------------------------------  
Participant’s Signature    Date 
 ----------------------------------------    ---------------------------------------- 
Witness’s Signature     Date 
 
68 
 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX III 
Socio-economic Demographic Information 
Participant’s ID: ___________________________ Date: ___________________________ 
 DEMOGRAPHIC INFORMATION RESPONSE 
1 Gender Male                        Female  
  
2 Date of Birth  
Never married                         
 
Married/ Co-habiting              
3 Current Marital Status 
Separated/ Divorced     
 
Widowed  
None 
 
Middle/ Junior Secondary School    
 
4 What is your highest level of education? 
Senior Secondary School 
 
Post-Secondary/ Tertiary Level  
< GH ₵200.00 
 
Over the past year, what has been your GH ₵200.00 to GH ₵500.00  
5 
 
average monthly income? GH ₵500.00 to GH ₵1000.00 
 
GH ₵1000.00 > 
6 Recorded Height  
7 Recorded Weight  
8 Calculated BMI  
 
 
69 
 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX IV 
Dietary History Form 
Please record all food, drink, vitamin, mineral/ supplemental intake using specific amounts with 
product brands for the next three (3) days. Include at least one (1) weekend day if possible. 
Day 1: 
Participant’s ID: ____________________________ Day/Date: ___________________________ 
Gender: Male    Female    Age: ______________________ 
Corresponding 
Food Amount 
Meal/ Snack amount in food Estimation 
Item (in handy measures) 
atlas 
Breakfast/     
morning 
Mid-morning     
Lunch/     
Afternoon 
Mid-afternoon     
Supper     
Before Bedtime     
supplements     
 
70 
 
University of Ghana  http://ugspace.ug.edu.gh
Day 2: 
Participant’s ID: ____________________________ Day/Date: ___________________________ 
Gender: Male    Female    Age: ______________________ 
Corresponding 
Amount  
Meal/ Snack Food Item amount in food Estimation 
(in handy measures) 
atlas 
Breakfast/     
morning 
Mid-     
morning 
Lunch/     
Afternoon 
Mid-     
afternoon 
Supper     
Before     
Bedtime 
supplements     
71 
 
University of Ghana  http://ugspace.ug.edu.gh
Day 3: 
Participant’s ID: ____________________________ Day/Date: ___________________________ 
Gender: Male    Female    Age: ______________________ 
Corresponding 
Amount  
Meal/ Snack Food Item amount in food Estimation 
(in handy measures) 
atlas 
Breakfast/     
morning 
Mid-     
morning 
Lunch/     
Afternoon 
Mid-     
afternoon 
Supper     
Before     
Bedtime 
supplements     
72 
 
University of Ghana  http://ugspace.ug.edu.gh
 
73