food & nutrition
research
ORIGINAL ARTICLE
Household food insecurity is associated with child’s dietary 
diversity score among primary school children in two districts in 
Ghana
Janet Antwi1, Esi Quaidoo2, Agartha Ohemeng3* and Boateng Bannerman4
1Department of Agriculture, Nutrition and Human Ecology, Prairie View A&M University, Prairie View, TX, USA; 
2Department of Nutrition, School of Public Health and Health Sciences, University of Massachusetts Amherst, 
Amherst, MA, USA; 3Department of Nutrition and Food Science, University of Ghana, Legon, Accra, Ghana; 
4Nutrition Linkages Project, University of Ghana, Accra, Ghana
Popular scientific summary
•  Among school going children aged 6-12 years in selected urban and rural settings in Ghana, dietary 
diversity was generally low
•  Household food insecurity was found to be negatively associated with the children’s dietary diversity.
•  In line with working to meet Sustainable Development Goal 2, more measures are needed to tackle 
malnutrition of school age children, paying more attention to factors that influence household level 
food availability, accessibility, and quality.
Abstract
Background: Dietary diversity is generally considered as a good indicator of nutrient adequacy and is 
influenced by various factors at the national, household, and individual levels.
Objective: The present study sought to determine the relationships between household food insecurity, primary 
caregivers’ nutrition knowledge, and dietary diversity of school-aged children in Ghana.
Methods: This forms part of  a longitudinal study conducted in the Ayawaso West Municipal district in 
Accra (urban setting) and the Upper Manya Krobo district (rural setting) in Ghana. Data were collected 
from a total of  116 caregiver-child dyads using 24-h dietary recall and a short version of  the US 12-month 
Household Food Security Survey Module. Nutrition knowledge and sociodemographic data were obtained 
using a structured questionnaire. Multivariable logistic regression was used to check for factors associated 
with children’s dietary diversity.
Results: Majority of households reported food insecurity, with a higher percentage of insecure households 
located in the rural area (88.9% vs. 46.5%, P ≤ 0.0001), compared to the urban setting. Diet diversity among 
the study children was low, with a mean (standard deviation [SD]) of 5.8 (2.1) out of 14 food groups. Children 
living in food insecure households were three times more likely to have received low diverse diet compared to 
those from food secure households (adjusted odds ratio [OR] =3.3, 95% confidence interval [CI]: 1.4–8.0). 
Caregivers’ nutrition knowledge was, however, not related to children’s dietary diversity.
Discussion and conclusion: Household food insecurity was a main predictor of dietary diversity among 
school-age children in this study. Thus, caregiver knowledge in nutrition may not be enough, particularly in 
the presence of food insecurity to guarantee adequate nutrition for school-aged children.
Keywords: food insecurity; nutrition knowledge; dietary diversity; school-age children
Received: 8 March 2021; Revised: 25 November 2021; Accepted: 2 December 2021; Published: 7 January 2022
Studies on child nutrition have applied several assess- Dietary diversity can be used as a suitable indicator of nu-ment tools to gauge the presence and extent of mal- trient adequacy (7, 9), which reflects one’s nutritional sta-nutrition within different communities (1–6). One tus. Research shows that a child who consistently consumes 
such tool used to date is the dietary diversity score (7, 8). a variety of meals with basic macro- and micro-nutrients 
Food & Nutrition Research 2022. © 2022 Anet Antwi et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// 1
creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, 
even commercially, provided the original work is properly cited and states its license. Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715
(page number not for citation purpose)
Janet Antwi et al.
is more likely to meet recommended nutrient intakes that arm were included in the parent study, and thus, no food 
help maintain good health (10). Factors that can influ- insecurity data were available for children in the control 
ence household and individual dietary diversity are varied group. The baseline data from the intervention study 
and include nutrition knowledge (11, 12), socioeconomic were used for this paper, to exclude any potential effect of 
factors (13–16), cultural factors (3, 17), and food security the intervention on the variables of interest.
(3, 13, 18). Knowledge and an understanding of nutrition 
can play a role in caregivers’ food ingredient sourcing, por- Study area
tion-sizing of the various food groups, and meal prepara- This study was conducted in Dzorwulu (urban setting), 
tion techniques (11, 12, 19). Socioeconomic factors such as and Brepaw Upper and Fefe (rural setting) from June 
occupation and level of education have also been identified through December 2018. Dzorwulu is a vicinity with well-
as determinants of dietary diversity, with lower levels of planned residential areas that accommodates the working 
education correlated with low dietary diversity (2). class and upper crust of Accra society in the Ayawaso 
Food security has four main dimensions: food avail- West Municipal District, of the Greater Accra Region 
ability, economic and physical access to food, food utili- of Ghana. On the other hand, Brepaw Upper and Fefe 
zation, and stability (20), and thus, it is a concept that is are two villages in the Aseseswa subdistrict of the Upper 
closely linked with a child’s ability to consume a diverse Manya Krobo district in the Eastern Region of Ghana.
diet. Causes of food insecurity in low-income countries 
include poverty, ineffective food policies, inauspicious cli- Study population and sampling
mate events, insufficient food production, and difficulty in For this paper on baseline data of the intervention arm, 
accessing food due to poor transportation infrastructure school children from two conveniently selected public 
(14, 21, 22). basic schools (one urban and one rural) and their pri-
School-age children (typically 6-12 year olds), and mary caregivers were included. The details of sampling 
younger children, have a high risk of becoming malnour- and recruitment processes for the parent study have been 
ished when their diets are not optimal (23, 24). Research described in a previous publication (26). Each arm of the 
in child nutrition has largely focused on children below parent study consisted of one urban school and one rural 
the age of 5 years in Ghana. The few published studies school. The sample size for the parent study was calcu-
that have documented school-age children nutrition sit- lated using a 95% confidence rate, 4% error margin, and 
uation suggest that this is another vulnerable group that 80% power to detect a 10% difference (P < 0.05) in pro-
needs attention if  the country is to overcome hunger and portion in nutrition knowledge between intervention and 
all forms of malnutrition (25, 26). Information gathered control groups. This gives a minimum required sample 
from such research is needed to assist in constructing size of 86 in each study arm. This was adjusted to 100 per 
contextual food policies aimed at promoting adequate nu- group to account for nonresponses. Caregivers of the chil-
trition for school-age children at various administrative dren were informed about the study at a Parent Teacher 
levels. Therefore, the objective of the parent study from Association meeting, in addition to letters and informed 
which data were obtained for this current study was to consent forms that were sent home with eligible chil-
evaluate the effect of a nutrition education intervention dren  for approval. Children within the target age group 
on nutrition knowledge, diet, and nutritional status of (6–12 years) who returned signed informed consent forms 
school-aged children in urban and rural settings in Ghana from their caregivers were included in this study. Caregiv-
(26). The study also engaged the children’s caregivers in ers with children within the target age group (6–12 years) 
assessing nutritional knowledge, attitudes, and practices. who signed informed consent forms for themselves and 
This current study focuses on the factors that are associ- their children’s participation and gave them to the chil-
ated the dietary diversity scores of school-aged children. dren to return to the research team were, thus, included 
in this study.
Methods
Data collection 
Study design All study questionnaires were pretested among individ-
The analysis presented in this paper focused specifically uals with characteristics similar to the study population 
on primary school children and their caregivers for whom to ensure that the assessments were contextual. Thus, 
data were available in the intervention arm of a parent questionnaire pretesting took place in two locations: one 
study conducted to determine the impact of a 6-week school located in an urban area and a second school in the 
nutrition education intervention on the nutrition knowl- rural setting. Research assistants received a 3-day training 
edge, attitudes, and practices of school-age children in in implementing and administering the study tools. So-
Ghana (26). The participants in the intervention arm ciodemographic information that was obtained included 
were included because only caregivers of children in this child’s age, gender, and current class in school. Caregivers 
2 Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715
(page number not for citation purpose)
Dietary diversity among primary school children
were invited to their wards’ schools to interact with the Households were further classified as food secure (score 
research team. Research assistants interviewed caregivers ≤ 1) and food insecure (score > 1).
individually; the interview took approximately 30 min. 
Data on caregivers’ occupation, marital status, level of Statistical analysis
education, residence, nutrition knowledge, and household All sociodemographic characteristics of the study partic-
food security were also collected. Information on the di- ipants underwent descriptive analyses. The Pearson’s chi-
etary intake of the school children was also collected. square test for proportions was first used to evaluate the 
Caregivers’ nutrition knowledge was obtained through possible relationship between factors including household 
interview questions that focused on number of meals to food insecurity and dietary diversity, as well as socioeco-
feed child, what constituted healthy eating, nutrient con- nomic factors including child’s age, sex, and caregiver ed-
tent of foods, food safety, quality of food to prevent ill- ucation. The dependent variable of interest was whether 
ness, and food function. To generate knowledge scores for a child received a high or a low diversified diet in the 24 h 
each caregiver, a correct answer to each question was as- prior to the interview, and the main independent variable 
signed a value of one, and an incorrect answer was coded of interest was household food insecurity status. Multi-
as zero. Next, all scores under this section were summed variable logistic regression modeling was conducted after 
up, and the total value was used to represent the knowl- the Pearson chi-square test to assess the relationship be-
edge score for that person. The maximum possible score tween household food insecurity status and child’s dietary 
for the nutrition knowledge assessment was 20. diversity, adjusting for caregiver nutrition knowledge, 
The dietary intake data were gathered using single 24-h formal education, and sociodemographic factors such as 
dietary recall. Each child was asked to list and describe all the age and sex of children. The choice of variables that 
foods and beverages consumed at school and at home in were included in the adjusted analysis was based on exist-
the past 24-h indicating the time and source of the food. ing literature on the possible related factors of individual 
Visual household measures and food models were used to dietary diversity. Although the ‘location’ variable (urban/
help children estimate the amounts of foods and bever- rural) was significantly associated with dietary diversity 
ages consumed (data on food quantities are not presented (P < 0.0001) in the bivariate analysis, it was not included 
in this paper). The 24-h recall information was used to in the adjusted model because of the very low level of 
calculate dietary diversity scores using the Food and Ag- variability observed in the sample (only two children in 
riculture Organization guidelines (7), by categorizing food the rural setting had high diverse diet). Statistical analyses 
items consumed by the children into 14 food groups. For were performed using SPSS version 20.0, and P-value of 
each food group, a child was given a score of one if  he/she <0.05 was considered statistically significant.
consumed any food item in that group, and a score of zero 
if  child did not consume any item in the group. A sum of Ethical approval
the scores for all the food groups represented the dietary This study was conducted according to the guidelines 
diversity score of participants. To categorize the dietary laid out in the Declaration of Helsinki, and researcher 
diversity variable, the median score of the children com- received ethical approvals and permissions from all the 
puted to be six was used as the cut-off  point. Thus, low relevant institutions before data collection started.
diet diversity in this study was defined as having a total 
score less than six, and high dietary diversity was defined Results
as a total score equal or more than six. A total of  116 caregiver-child dyads were included 
The household food security status was measured using in this analysis. The mean age of  the children was 
the 6-item short form of the US 12-month Household 9.6 (1.8) years, while that of  the caregivers was 38.5 (10.8) 
Food Security Survey Module (27). The shorter version years (Table 1). Majority of  the caregivers (73.3%) who 
was used in order to reduce participants’ response burden, participated in the study were women, and most partici-
in lieu of the longer version. Despite its length, it has been pants were the biological mothers and fathers (82.8%) of 
demonstrated to measure and differentiate food security the study children. Most of  the primary caregivers had 
and food insecurity with sufficient sensitivity, specificity, low level of  formal education. Only about one-third of 
and minimum bias compared to the lengthy module (28). them (31.1%) had gone through the Senior High School 
The affirmative responses (‘often true’, ‘sometimes true’, level or above. Trading was the most common primary 
‘almost every month’, ‘some months but not every month’, occupation among the caregivers (Table 1), followed by 
and ‘yes’) were coded as ‘1’ and negative responses as ‘0’, farming and vocational jobs such as dressmaking and 
and these were added to get a total score for each house- hairdressing. It is, however, important to note that with 
hold. Scores of 0–1 were classified as food secure, while the exception of  one participant, all the caregivers who 
2–4 and 5–6 were classified as food insecurity without indicated farming as their primary occupation were lo-
hunger and food insecurity with hunger, respectively. cated in the rural setting.
Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715 3
(page number not for citation purpose)
Janet Antwi et al.
Table 1. Sociodemographic characteristics of study participants The experience of household food insecurity was re-
(N = 116) ported by majority of the caregivers (62.9%), with only 
Characteristic Mean (SD) n (%) about one-third being classified as food secure (Fig. 1). 
Children Of households that were classified as food insecure, more 
than half  were severely food insecure. Food insecurity was 
Age (years): 9.61 (1.84)
reported by a higher proportion of caregivers in the rural 
 6–9 48 (41.4)
setting (88.9% vs. 46.5%, P ≤ 0.0001), compared to those 
 10–12 68 (58.6)
in the urban setting.
Sex:
Based on a single 24-h recall, the most consumed food 
 Male 53 (45.7)
items consumed by the children were from the grains’ 
 Female 63 (54.3) food group, followed by the ‘other vegetable’ group such 
School levela: as onions and garden eggs (Fig. 2). Among the animal 
 Lower primary 60 (51.7) source foods, the most common type consumed by the 
 Upper primary 56 (48.3) study children during the period of assessment was fish, 
Takes money to school: 104 (89.7) but organ meat was totally absent from their diet. Nota-
Amount (GH¢): 2.37 (1.85) bly, vitamin A rich fruits were the least consumed in the 
 Caregivers vegetable and fruit category.
 Age (years) 38.89 (10.79) Generally, diet diversity among the study children was 
 Sex: low, with a mean score of 5.8 (2.1) out of 14 food groups. 
  Male 31 (26.7) Using the group median of six as cut-off, about half  of the 
  Female 85 (73.3) study children (50.9%) consumed high diverse diets within 
 Marital status: the 24 h prior to the data collection. Bivariate analysis 
  Married/co-habiting 86 (74.2) indicated a higher proportion of older children received 
  Separated/divorced/widowed 15 (12.9) diverse diet compared to younger children, but there was 
  Single 15 (12.9) no difference based on gender (Table 2). Almost all chil-
 Formal educationb: dren (96.2%) living in households, where the main occu-
  None 23 (19.8) pation of the primary caregiver was farming, received low 
  Primary 16 (13.8) diverse diet within the period of observation. In a logistic 
  JHS 41 (35.4) regression model, dietary diversity of the children was 
  SHS 27 (23.3) significantly associated with household food insecurity 
  Above SHS 9 (7.7) and child’s age (Table 3). Children living in food insecure 
 Primary occupation: households were three times more likely to have received a 
  Trading 40 (34.5) low diverse diet, and this was significant for both levels of 
  Farming 26 (22.4) food insecurity (without and with hunger), compared to 
  Vocational 26 (22.4) children from food secure households. On the other hand, 
older children were less likely to have eaten a low diverse 
  Unemployed/student 5 (4.3)
c diet (odds ratio [OR] = 0.8, 95% confidence interval [CI]:   Other 19 (16.4)
0.6 – 0.9). There was also a tendency for children whose 
 Relation to study child:
primary caregivers had formal education up to at least Se-
  Parent 96 (82.8)
nior High to consume a high diverse diet. There was, how-
  Other relative 20 (17.2)
ever, no significant association between dietary diversity 
 Estimated household income/month (GH¢): of children and caregivers’ nutrition knowledge.
  ≤500 71 (61.2)
  Above 500 35 (30.2) Discussion
  Do not know  10 (8.6) Engaging caregivers in an attempt to piece together 
 Residence: school children’s nutrition situation is particularly im-
  Urban 71 (61.2) portant since key determinants of dietary practices ad-
  Rural 45 (38.8) opted by children include both caregivers’ and household 
aLower primary consists of classes One to Three, while Upper primary characteristics. Our study engaged caregivers of school 
is made up of classes Four to Six. aged children in urban and rural Ghana to assess their 
bJHS represents Junior High School, and SHS represents Senior High nutrition knowledge and their households’ food security. 
School in Ghana.
cOther occupations included public servants, pensioners, domestic help, Dietary diversity, in general, was low among both 
and laborer. Values are presented as frequencies (%) or means (standard urban- and rural-based school aged children. Consump-
deviation). tion of grains was high among our sample of children with 
4 Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715
(page number not for citation purpose)
Dietary diversity among primary school children
Fig. 1. Household food insecurity among study participants. The darkest section represents study respondents who were food 
insecure without hunger. The crossed-line section represents study respondents who were food insecure with hunger, and the 
brick section represents the study respondents who were food secure.
Fig. 2. Intake of food groups by the school children, based on a single 24-h dietary recall. Bars represent the groups of food that 
the school children ate in 24 h. DGLV stands for dark green leafy vegetables.
rural-based children consuming more grains than any other food groups such as animal protein, and starch-
other food group when compared to their urban-based based meals are perceived to be quenchers of hunger (15, 
counterparts. This finding is consistent with that from a 30). On the other hand, low consumption of fruits, veg-
study in Uganda (29), indicating that grain consumption etables, animal protein, and dairy was observed among 
is generally high among children in sub-Saharan Africa participants, similar to other studies (29–31). One may 
as they form most of the staple foods. Starchy tubers and expect that with majority of rural households in this 
roots were also largely consumed by both urban- and study engaged in farming, intake of vegetables in general 
rural-based children. Starchy foods are an indispensable would be high due to the cultivation of these food items 
meal ingredient in many African meals as they seem to on their farms. However, it has been noted that subsis-
be readily available and accessible than other food staff  tence farmers in developing countries focus on growing 
(8, 30). Grains and starches are relatively cheaper than few varieties of crops, which are mainly slated for sale (29, 
Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715 5
(page number not for citation purpose)
Janet Antwi et al.
Table 2. Bivariate analysis comparing study school children based on dietary diversity
Independent variables Child’s dietary diversity P
Total High (n = 59) Low (n = 57)
Child’s sex 0.270
 Male 53 (45.7) 24 (45.3) 29 (54.7)
 Female 63 (54.3) 35 (55.6) 28 (44.4)
Child’s age 0.041
 6–9 years 48 (41.4) 19 (39.6) 29 (60.4)
 10–12 years 68 (58.6) 40 (58.8) 28 (41.2)
Caregiver education 0.022
 Below SHS 80 (69.0) 35 (43.8) 45 (56.2)
 SHS and above 36 (31.0) 24 (66.7) 12 (33.3)
Caregiver occupation 0.002
 Trader 40 (34.5) 23 (57.5) 17 (42.5)
 Farmer 26 (22.4) 1 (3.8) 25 (96.2)
 Vocational 26 (22.4) 17 (65.4) 9 (34.6)
 Unemployed 5 (4.3) 4 (80.0) 1 (20.0)
 Other 19 (16.4) 14 (73.7) 5 (26.3)
Area of residence <0.0001
 Urban 71 (61.2) 57 (80.3) 14 (19.7)
 Rural 27 (38.8)  2 (4.4) 25 (95.6)
Household food security 0.006
 Food secure 43 (37.1) 29 (67.4) 14 (32.6)
 Food insecure 73 (62.9) 30 (41.1)  43 (58.9)
Data presented as frequency (%).
Table 3. Factors associated with child diversity as unadjusted and adjusted odds ratios
Independent variables Low child dietary diversity
Unadjusted Adjusted
OR 95% CI OR 95% CI
Child’s age (years) 0.8 0.7, 1.0 0.8 0.6, 0.9
Child’s sex 0.7 0.3, 1.4 0.7 0.3, 1.6
 Male, femalea
Caregiver education 0.4 0.2, 0.9 0.4 0.2, 1.1
 Below Senior High (SHS) Levela
 SHS and above
Caregiver nutrition knowledge 1.1 0.8, 1.4 1.2 0.9, 1.7
Household food security
 Food securea 3.1 1.2, 8.2 2.9 1.0, 8.4
 Food insecure without hunger 2.9 1.2, 6.9 3.1 1.2, 8.0
 Food insecure with hunger
Using the group median of 6 as cut-off, low diversity was defined as a total diversity score less than 6. Diversity score was calculated according to FAO 
guidelines (7).
aReference category of the categorical variables §P < 0.10 and *P < 0.05.
30). As a result, farming households tend to base their that in developing countries, many crops are consumed 
diets on few food groups resulting in low dietary diver- only when they are in season, particularly in households 
sity of household members. Additionally, reports indicate with low incomes (32, 33). In this study, dark green leafy 
6 Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715
(page number not for citation purpose)
Dietary diversity among primary school children
vegetables (DGLV), a rich source of iron, folate, and be- nutrition have a protective impact on their children’s 
ta-carotenes, were not consumed by most of the children. nutrition status as they are better equipped to make 
Even though our study did not collect data on seasonal appropriate dietary decisions for their wards (19). How-
variations in relation to dietary intake, this research took ever, caregiver knowledge alone is insufficient to result 
place when common local DGLV was out of season. Con- in better diet since other important factors such as food 
sidering that most of the households had low incomes, availability and accessibility and intra-household food 
this seasonality might have accounted for the low intake allocation influence what is finally consumed. For ex-
of nutrient-dense dark leafy vegetables among the school ample, milk and milk products were barely consumed 
children in this study. by the children in the current study, even though some 
Our study identified more than half  of  the study’s caregivers knew the benefits of  dairy. Thus, for our 
participating households as food insecure, with about a study sample, caregivers’ knowledge of  nutrition did 
quarter of  them identified as being food insecure with not necessarily translate into diverse diet. This study 
hunger, and household food insecurity was strongly as- illustrates that even though an awareness of  the ad-
sociated with child dietary diversity. Earlier studies have vantages of  specific foods to children’s health may be 
reported varied prevalence of  food insecurity in different present in a caregiver, the availability, accessibility, and 
parts of  Ghana (18, 34, 35), indicating vast differences utilization of  these products would be an issue in a 
with respect to different parts of  the country as well as food insecure home. Thus, food insecurity, as observed 
the seasonality of  food availability. Additionally, almost in this study, can have a greater impact on the dietary 
all farming households were food insecure in the current choices made for children than parental awareness of 
study. This is similar to findings from study that assessed adequate nutrition behaviors.
household food security in farming and nonfarming 
communities in three ecological regions in Ghana (18). Conclusion 
Many of  the rural-based caregivers did not have a steady The findings from this study indicate a significant rela-
income, relying largely on the sale of  produce from their tionship between household food security and dietary 
farms to gain income. In the urban setting, most of  the diversity of school age children, but caregivers’ nutrition 
caregivers reported earning monthly incomes of  less knowledge was not associated with diet diversity of the 
than 500 Ghana cedis (i.e. 85.8 US dollars). Household children. Our study provides additional evidence that 
wage earners with regular incomes have more purchas- various factors such as household level food security 
ing power and may be more likely to purchase food items play important roles in ensuring good quality nutrition 
even when they are out of  season and not as common (2) for school-aged children. Programmes aimed at tackling 
and, thus, can vary the meals consumed at the household malnutrition among school-aged children should, there-
level, compared to households where income is not as fore, tailor interventions that address numerous drivers of 
steady (20). It is also important to note that the observed child malnutrition including food insecurity at household 
association between household food insecurity and child levels.
dietary diversity in this study has also been noted among 
younger children (18, 20). Thus, there is the need to in- Study limitations
clude food security issues at the household level as an in- These findings should be interpreted with some level of 
tegral part of  policies and strategies addressing nutrition caution. This study establishes possible associations, not 
of  all children. causality. Although the researchers used validated ques-
There was a tendency for children whose primary care- tionnaires and rigorous assessment tools, not all factors 
givers had less than secondary education to receive low that could possibly impact a child’s dietary diversity, 
diverse meals during the observation period. A study in such as cultural practices and seasonal variations of food 
Algeria reported that the educational level of caregivers availability, were assessed. Additional information on the 
played a significant role in the dietary diversity of chil- actual nutrient quality of the meals consumed by the chil-
dren in their study (36). Considering that education qual- dren was not recorded.
ification can influence caregivers’ income (2), the higher 
a caregiver’s education level, the more likely a household Acknowledgments
would have high household income, which could trans- JA and AO designed the research. EQ participated 
late into a more diversified diet for children within the in data collection. BB and AO analyzed the data. JA, 
household. EQ, and AO wrote the first draft of  the manuscript. JA, 
Caregivers’ nutrition knowledge, however, was not EQ, AO, and BB reviewed and edited the manuscript. 
associated with dietary diversity of  the school chil- All authors read and approved the final manuscript. We 
dren studied, contrary to other studies (11, 12). It has are indebted to the research assistants for their untiring 
been suggested that mothers who have knowledge on support during data collection. We cherish the school 
Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715 7
(page number not for citation purpose)
Janet Antwi et al.
principals, primary school teachers, school children,  6. Vandevijvere S, De Vriese S, Huybrechts I, Moreau M, Van 
caregivers, and school cooks/vendors of  the four pri- Oyen H. Overall and within-food group diversity are associated 
mary schools that participated in the parent study for with dietary quality in Belgium. Public Health Nutr 2010 Dec; 
13(12): 1965–73. doi: 10.1017/S1368980010001606 
the immerse contribution and support during the imple-  7. Food and Agriculture Organization. Guidelines for measuring 
mentation of  the study. household and individual dietary diversity. Rome: The United 
Nations; c2010–2013. Nutrition and Consumer Protection 
Conflict of interest and funding Division; [about 60 screens]. Available from: http://www.fao.
All authors declare no conflict of interest. This work org/3/a-i1983e.pdf [cited 25 July 2017].
was supported with funding from the Institute of Inter-  8. Chagomoka T, Drescher A, Glaser R, Marschner B, Schlesinger J, 
Nyandoro G. Women’s dietary diversity scores and childhood an-
national Education (IIE) through the Carnegie African thropometric measurements as indices of nutrition insecurity along 
Diaspora Fellowship Program (CADFP) to JA and AO the urban–rural continuum in Ouagadougou, Burkina Faso. Food 
for the fellowship and funding to undertake this research. Nutr Res 2016 Jan 1; 60(1): 29425. doi: 10.3402/fnr.v60.29425 
The IIE or CADFP had no influence in the study design,  9. Ruel MT. Is dietary diversity an indicator of food security or 
conduct of the study, analysis of data, interpretation of dietary quality? International Food Policy Research Institute 
findings, or writing of this article. The authors would like (IFPRI) Discussion Paper Brief; 2002. Available from: https://
www.ifpri.org/publication/dietary-diversity-indicator-food-se-
to thank the Cooperative Agriculture Research Center of curity-or-dietary-quality-0 [cited 25 July 2017].
the College of Agriculture and Human Sciences at Prairie 10. Fanzo, J. The nutrition challenge in sub-Saharan Africa. 
View A&M University for supporting cost of publication United Nations Development Programme [UNDP] Work-
of this work. ing  Paper. 2012. Available from: http://www.undp.org/con-
tent/dam/rba/docs/Working%20Papers/Nutrition%2 [cited 12 
January 2020].
Ethical standards disclosure 11. Ali S, Chaudry T, Naqvi QU. Effect of maternal literacy on 
This study was conducted according to the guidelines laid child health: myth or reality. Ann PIMS-Pak Inst Med Sci 2011; 
down in the Declaration of Helsinki, and all procedures 7: 100–3.
involving research study participants were approved by 12. Shanshan GE, Jingqiu MA, Shanshan LI, Jie Zhang XS. Lack 
the College of Basic and Applied Sciences, University of dietary diversity contributes to the gaps in micronutrient sta-
of Ghana Institutional Review Board (IRB, {study # tus and physical development between urban and rural infants. 
Iran J Public Health 2018 Jul; 47(7): 958.
ECBAS 029/17-18}) and State University of New York at 13. Ali NB, Tahsina T, Hoque DM, Hasan MM, Iqbal A, 
Oneonta IRB (study # 529). Approvals were sought from Huda TM, et al. Association of  food security and other 
the Accra Metropolitan Assembly directorates, Upper socio-economic factors with dietary diversity and nutritional 
Manya Krobo District of the Ghana Education Service, statuses of  children aged 6–59 months in rural Bangladesh. 
and school principals of the four participating schools. A PLoS One 2019 Aug 29; 14(8): e0221929. doi: 10.1371/jour-
voluntary written informed consent was obtained from all nal.pone.0221929 
14. Usfar AA, Fahmida U, Februhartanty J. Household food secu-
participants. rity status measured by the US-Household Food Security/Hun-
ger Survey Module (US-FSSM) is in line with coping strategy 
indicators found in urban and rural Indonesia. Asia Pac J Clin 
References 
Nutr 2007 Jun 1; 16(2): 368–74.
 1. Arimond M, Wiesmann D, Becquey E, Carriquiry A, Daniels 15. Bukania ZN, Mwangi M, Karanja RM, Mutisya R, Kombe 
MC, Deitchler M, et al. Simple food group diversity indicators Y, Kaduka LU, et al. Food insecurity and not dietary diver-
predict micronutrient adequacy of women’s diets in 5 diverse, sity is a predictor of nutrition status in children within semiarid 
resource-poor settings. J Nutr 2010 Nov 1; 140(11): 2059S–69S. agro-ecological zones in eastern Kenya. J Nutr Metab 2014 Jan 
doi: 10.3945/jn.110.123414 1; 2014: 907153. doi: 10.1155/2014/907153 
 2. Codjoe SN, Okutu D, Abu M. Urban household character- 16. Anarfi JK & Ahiadeke C. Improving the health of children among 
istics and dietary diversity: an analysis of food security in the urban poor in the city of Accra, Ghana. Accra, Ghana: Afri-
Accra, Ghana. Food Nutr Bull 2016 Jun; 37(2): 202–18. doi: can Population and Health Research Center, ISSER; 2006.
10.1177%2F0379572116631882 17. Ochieng J, Afari-Sefa V, Lukumay PJ, Dubois T. Determinants 
 3. Kennedy GL, Pedro MR, Seghieri C, Nantel G, Brouwer I. of dietary diversity and the potential role of men in improving 
Dietary diversity score is a useful indicator of micronutrient household nutrition in Tanzania. PLoS One 2017 Dec 12; 12(12): 
intake in non-breast-feeding Filipino children. J Nutr 2007 Feb e0189022. doi: 10.1371/journal.pone.0189022 
1; 137(2): 472–7. doi: 10.1093/jn/137.2.472 18. Christian AK, Marquis GS, Colecraft EK, Lartey A, Soue-
 4. Steyn NP, Nel JH, Nantel G, Kennedy G, Labadarios D. Food ida R. Household food insecurity but not dietary diversity is 
variety and dietary diversity scores in children: are they good associated with children’s mean micronutrient density ade-
indicators of dietary adequacy? Public Health Nutr 2006 Aug; quacy in rural communities across Ghana. Nutrition 2019 Sep 
9(5): 644–50. doi: 10.1079/PHN2005912 1; 65: 97–102. doi: 10.1016/j.nut.2019.03.006
 5. Torheim LE, Ouattara F, Diarra MM, Thiam FD, Barikmo I, 19. Fadare O, Amare M, Mavrotas G, Akerele D, Ogunniyi A. 
Hatløy A, et al. Nutrient adequacy and dietary diversity in Mother’s nutrition-related knowledge and child nutrition out-
rural Mali: association and determinants. Eur J Clin Nutr comes: empirical evidence from Nigeria. PLoS One 2019 Feb 
2004 Apr; 58(4): 594–604. doi: 10.1038/sj.ejcn.1601853 28; 14(2): e0212775. doi: 10.1371/journal.pone.0212775 
8 Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715
(page number not for citation purpose)
Dietary diversity among primary school children
20. Faber M, Schwabe C, Drimie S. Dietary diversity in rela- 30. Chakona G, Shackleton C. Minimum dietary diversity scores 
tion to other household food security indicators. Int J Food for women indicate micronutrient adequacy and food insecu-
Saf Nutr Public Health 2009 Jan 1; 2(1): 1–5. doi: 10.1504/ rity status in South African towns. Nutrients 2017 Aug; 9(8): 
IJFSNPH.2009.026915 812. doi: 10.3390/nu9080812 
21. Agbadi P, Urke HB, Mittelmark MB. Household food security 31. Leyna GH, Mmbaga EJ, Mnyika KS, Hussain A, Klepp KI. 
and adequacy of child diet in the food insecure region north in Food insecurity is associated with food consumption pat-
Ghana. PLoS One 2017 May 11; 12(5): e0177377. doi: 10.1371/ terns and anthropometric measures but not serum micronutrient 
journal.pone.01773377 levels in adults in rural Tanzania. Public Health Nutr 2010 Sep; 
22. Sasson A. Food security for Africa: an urgent global 13(9): 1438–44. doi: 10.1017/S1368980 010000327 
challenge. Agric Food Sec 2012 Dec; 1(1): 1–6. doi: 32. Abizari AR, Azupogo F, Nagasu M, Creemers N, Brouwer ID. 
10.1186/2048-7010-1-2 Seasonality affects dietary diversity of school-age children in 
23. Best C, Neufingerl N, Van Geel L, van den Briel T, Osendarp northern Ghana. PLoS One 2017 Aug 14; 12(8): e0183206. doi: 
S. The nutritional status of school-aged children: why should 10.1371/journal.pone.0183206 
we care? Food Nutr Bull 2010 Sep; 31(3): 400–17. doi: 33. Savy M, Martin-Prével Y, Traissac P, Eymard-Duvernay S, 
10.1177/156482651003100303 Delpeuch F. Dietary diversity scores and nutritional status 
24. Danquah AO, Amoah AN, Steiner-Asiedu M, Opare-Obisaw of women change during the seasonal food shortage in rural 
C. Nutritional status of participating and non-participating Burkina Faso. J Nutr 2006 Oct 1; 136(10): 2625–32. doi: 
pupils in the Ghana School Feeding Programme. J Food Res 10.1093/jn/136.10.2625 
2012 Aug 1; 1(3): 263. doi: 10.5539/jfr.v1n3p263 34. Ministry of Food and Agriculture (MoFA). Food security situa-
25. Owusu JS, Colecraft EK, Aryeetey R, Vaccaro JA, Huffman tion in Ghana. Ghana: Northern Region Agricultural Develop-
FG. Nutrition intakes and nutritional status of  school age ment Unit; July 2015. Available from: https://mofafoodsecurity.
children in Ghana. J Food Sci 2017; 6(2): 11–23. doi: 10.5539/ wordpress.com/food-security-situation-in-ghana [cited 20 July 
jfr.v6n2p11 2020].
26. Antwi J, Ohemeng A, Boateng L, Quaidoo E, Bannerman B. 35. Ohemeng A, Marquis GS, Lartey A. Household food insecurity 
Primary school-based nutrition education intervention on is associated with respiratory infections among 6–11-month-
nutrition knowledge, attitude and practices among school- old infants in rural Ghana. Pediatr Infect Dis J 2015 Aug 1; 
age children in Ghana. Glob Health Promot 2020 Dec; 27(4): 34(8): 821–5. doi: 10.1097/INF.0000000000000743 
114–22. doi: 10.1177/1757975920945241 36. Morseth MS, Grewal NK, Kaasa IS, Hatloy A, Barikmo 
27. Bickel G, Nord M, Price C, Hamilton W, Cook J. I, Henjum S. Dietary diversity is related to socioeconomic 
Guide  to  measuring household food security. Alexandria, status among adult Saharawi refugees living in Algeria. 
VA: US Department of  Agriculture, Food and Nutrition BMC Public Health 2017 Dec; 17(1): 1–9. doi: 10.1186/
Services; 2000. Available from: https://www.fns.usda.gov/ s12889-017-4527-x
guide-measuring-household-food-security-revised-2000 
[cited 25 July 2017].
28. Blumberg SJ, Bialostosky K, Hamilton WL, Briefel RR. The *Agartha Ohemeng
effectiveness of a short form of the Household Food Secu- Department of Nutrition and Food Science
rity Scale. Am J Public Health 1999 Aug; 89(8): 1231–4. doi: University of Ghana
10.2105/AJPH.89.8.1231 P.O. Box LG 134
29. Mayanja M, Rubaire-Akiiki C, Morton J, Young S, Greiner Legon
T. Diet diversity in pastoral and agro-pastoral households in Accra
Ugandan rangeland ecosystems. Ecol Food Nutr 2015 Sep 3; Ghana
54(5): 529–45. doi: 10.1080/03670244.2015.1041135 Email: anohemeng@ug.edu.gh
Citation: Food & Nutrition Research 2022, 66: 7715 - http://dx.doi.org/10.29219/fnr.v66.7715 9
(page number not for citation purpose)