Batame BMC Public Health (2024) 24:985 BMC Public Health

https://doi.org/10.1186/s12889-024-18204-7

RESEARCH Open Access

Unlocking gender dynamics in food

and nutrition security in Ghana: assessing
dietary diversity, food security, and crop
diversification among cocoa household heads
in the Juaboso-Bia cocoa landscape
Michael Batame1,2*

Abstract
Background Malnutrition is a worldwide problem that impacts every country, affecting one in three individuals,
including Ghana. According to estimates from the Food and Agriculture Organization (FAO), 690 million people were
undernourished globally in 2019. Malnutrition is no exception in rural cocoa communities in Ghana. The expansion
of cocoa is causing food insecurity and low dietary diversity in most cocoa-growing areas. There is little information
on the food security status and dietary diversity among male-headed and female-headed cocoa households in
the Juaboso-Bia cocoa landscape. Thus, this study sought to explore the dietary diversity distribution, food security
status, sources of staple food, food unavailability periods, food production status, themes contributing to low dietary
diversity and food insecurity, and perception of the impact of cocoa expansion on crop diversification among male-
headed and female-headed cocoa households in the study area.
Methods Both quantitative and qualitative research methods were employed to address the research questions. The
study collected 200 semi-structured questionnaire data and 14 in-depth interview data from cocoa household heads
in the Juaboso-Bia landscape. The survey data was cleaned and analysis, such as household dietary diversity status,
food security status, and binary logistic regression were performed in the Statistical Package for Social Scientists
(SPSS). The in-depth interviews were analyzed using thematic analysis.
Results Through this study, it was revealed that 62.8% of the male-headed cocoa households had medium to high
dietary diversity compared to their female counterparts 39.3%. About 47.9% of the male-headed households were
food secure than the female-headed households (29.1%). Moreover, the months that both male and female-headed
households recalled facing severe food unavailability were July and June. In addition, climate change/variability,
unavailable lands, poverty, large household size, and gender stereotypes were themes promoting low dietary diversity
and food insecurity among male and female-headed households. Furthermore, sex, total household income, and

*Correspondence:
Michael Batame
[email protected]
Full list of author information is available at the end of the article

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data.
Batame BMC Public Health (2024) 24:985 Page 2 of 18

cropland conversion to cocoa were the variables influencing household heads’ perception of the impact of cocoa
expansion on crop diversification.
Conclusions The study showed that the male-headed cocoa households had high dietary diversity and were food
secure than the female-headed cocoa households, lending credence to the conceptual framework applied in this
study. There is a significant relationship between household head gender and food security status. Therefore, this
study recommends the following interventions to improve dietary diversity and food security among male and
female-headed cocoa households: raise awareness about the need for diverse diets and provide practical information
on how to incorporate a greater variety of food groups into their daily meals; and promote gender equity and
inclusivity in food security interventions. Future research could investigate how gender mainstreaming policies in
agriculture have helped empower and improve the food security of female-headed households in Ghana.
Keywords Nutrition, Food security, Mixed research methods, Dietary diversity score, Crop diversification, Rural cocoa
household

Introduction specifically, the availability of a variety of agricultural

Globally, one in three people suffers from malnutrition items for self-consumption raises the dietary diversity of
[1]. Approximately 690 million people worldwide were households in agricultural production areas. The study
undernourished in 2019, according to the Food and Agri- by [7] showed that in cocoa-growing areas, cocoa house-
culture Organization (FAO) estimate [2]. In emerging holds usually consume staple foods (roots, tubers, oil,
nations, malnutrition can become worse, particularly in and small dried fish) that are not sufficient in micronu-
rural areas where the poorest people are most concen- trients to counter undernutrition. Malnutrition occurs in
trated [3]. Malnutrition has many different root causes, Ghana’s cocoa regions as a result of the majority of cocoa
but it is more prevalent in areas where there is evidence households’ undiversified diets (consuming less than 5
of rising food insecurity, which is frequently linked to the food groups daily), which also stunts their growth, espe-
occurrence of climate change events [2]. For instance, in cially among women [7].
southern Africa, erratic rainfall that is typically linked Food and nutrition security has several facets, and one
to climate change impacts farmers’ agricultural yields, of them is the quality of diets, and the dietary diversity
which in turn affects their food security and nutrition [4]. score (DDS) is a well-known numerical indicator of diet
Despite major improvements in food and nutrition quality [1, 12–14]. Dietary diversity offers information on
security over the past few decades, particularly in sub- the availability of a variety of foods in a household and
Saharan Africa (SSA) and South Asia, the prevalence can be used as a proxy for a person’s diet’s nutrient suffi-
of undernutrition is still high [1]. In Ghana, there have ciency [15]. For example, an increase in a person’s dietary
been numerous attempts to reduce the rate of malnutri- diversity score is related to an increase in diet nutrient
tion, and these efforts have steadily improved [5]. How- sufficiency [16]. Healthy eating is defined as consum-
ever, Ghana still struggles with the issue of malnutrition, ing a varied, well-balanced diet that includes foods high
which has contributed to 50% of child fatalities [6]. Mal- in vitamins and minerals, as well as fruits, vegetables,
nutrition in Ghana is largely attributed to high intake of and fresh, natural meals [15]. It also requires practicing
carbohydrate-rich food, namely cassava, maize, and rice, healthy eating practices and engaging in activities that
and inadequate food intake rich in vitamins and proteins, advance one’s physical and mental health [7, 17, 18]. The
for example, meat, eggs, milk products, legumes, and number of various foods or food groups in a diet is used
fruits [7]. to calculate dietary diversity. A concept known as house-
The Juaboso-Bia cocoa landscape is a rural district hold dietary diversity (HDD) is used to gauge a house-
where its inhabitants are predominantly cocoa farmers hold’s capacity to pay for a range of foods throughout a
and also engage in subsistence farming [8]. Many cocoa given period [2]. A common tool for evaluating food con-
households in this area have expanded their cocoa pro- sumption, including the degree of variety of foods that a
duction intending to increase their household income household has access to, is the Household Dietary Diver-
and secure their food and nutrition security [8–10]. This sity Score (HDDS), which is generated from the dietary
cocoa production expansion ends up displacing food diversity questionnaire [2]. This HDDS is limited by fac-
croplands hence exacerbating households’ food and tors, such as income, land size, household size, source of
nutrition insecurity in this cocoa rural district [8]. More- staple food, and age and education level of the household
over, [11] argued that food availability in rural places [3, 8, 15, 19, 20].
depends on the environment’s natural resources and There have been extensive studies conducted on cocoa
the ability to produce food through agriculture, more production and its impact on household food security in
Batame BMC Public Health (2024) 24:985 Page 3 of 18

Ghana. For instance, [21] examined food security among and female-headed cocoa households obtain their staple
male and female cocoa households in Wassa Amenfi food?; 3. What is the food production status among male
West District, Western North Region of Ghana. The and female-headed cocoa households?; 4. Which months
study used the household dietary diversity score to mea- do the male and female-headed cocoa households expe-
sure the food security status of cocoa households. The rience low consumption of dietary diversity?; 5. What
study found that the female-headed cocoa households are the contributing themes for the low dietary diversity
were more food insecure than the male-headed cocoa and food insecurity in the male and female-headed cocoa
households. Also, [8] investigated how farmers convert households’?; and 6. How do cocoa household heads per-
land use from food crops to cocoa, as well as the primary ceive the impact of cocoa expansion on crop diversifi-
factors that influence the transition. According to their cation in the Juaboso-Bia landscape? These answers are
findings, food cropland has been drastically reduced as crucial for comprehending dietary diversity across male
a result of cocoa expansion, affecting farmers’ food self- and female-headed households within the highest lev-
sufficiency. Again, [22] discovered that households with els of food insecurity, allowing policymakers to properly
lower dietary diversity had lower production diversity, implement tailored nutrition interventions as well as
while households with higher dietary diversity experi- informing and providing feedback to development activi-
enced higher production diversity in Indonesia. Dietary ties. This study’s results will contribute to a large body of
diversity was primarily loss due to decreased consump- knowledge on the nexus of cocoa production, food secu-
tion of healthful food groups, such as fruits, vegetables, rity, and dietary diversity.
legumes, and fish. Several food groups, such as dairy,
meats, and eggs, have been heavily consumed over the Literature review
years and are good components of nutritious diets. Fur- This sub-section explores studies on food security and
thermore, [23] revealed that gender, level of education, dietary diversity as well as their determinant factors.
source of food, and wealth of the head of household were
the factors influencing dietary diversity in urban house- Food security and dietary diversity
holds in Accra, Ghana. Their studies also found that veg- Many rural people experience food and nutrition inse-
etables were most consumed followed by cereal-based curity [2]. The study by [2] sought to identify the fac-
and grain products. tors that influence the Household Dietary Diversity
However, these studies did not provide insight into Score (HDDS) in the rural area of the Paute River Basin,
dietary diversity distribution among male and female- Azuay Province, Ecuador. A stratified random sampling
headed households, months in which food insecurity technique with proportional affixation was employed to
occurs, accessibility to sources of stable food, crop diver- determine the sample size of 383 surveys. The HDDS was
sification, and factors accounting for low dietary diver- used to quantify dietary diversity using 12 food groups
sity. Thus, the Juaboso-Bia cocoa landscape in Ghana over a 7-day recall period: cereals, fruits, sugar/honey,
has no information on the distribution of dietary diver- eggs, meat and eggs, legumes or grains, vegetables, oils/
sity and food security patterns among male and female- fats, milk and dairy products, meats, miscellaneous, fish,
headed cocoa households, and its associated factors. and shellfish. The association between the HDDS and
Dietary diversity is necessary to be explored since it sociodemographic factors was evaluated using a Poisson
has multifaceted reasons, such as health, nutrition, and regression model. The findings indicated that carbohy-
socio-economic benefits. Encouraging individuals and drates, fruits, roots and tubers, and sugar/honey were the
communities to embrace a diverse and balanced diet is most often consumed food groups among those evalu-
a key strategy for addressing nutritional challenges and ated. Furthermore, the predictive model’s most effective
promoting holistic well-being [10, 17]. drivers of the HDDS were housing size, household size,
To address these gaps, this study, therefore, examined per capita food expenditure, area under cultivation, edu-
the dietary diversity distribution and food security sta- cational attainment, and the head of household’s marital
tus among male and female-headed cocoa households status.
and its influencing factors in the Juaboso-Bia cocoa land- Furthermore, [24] examined the determinants of
scape. The hypotheses for this study were: H0 = there is dietary diversity score for rural households of Uttar
no significant relationship between food security sta- Pradesh State in India using 248 surveyed households
tus and gender of the household head, and H1 = there is through stratified random sampling. The study assessed
a significant relationship between food security status household dietary diversity score and individual dietary
and gender of the household head. Research questions diversity score, using 12 food groups and 9 food groups
explored for answers were: 1. How is the dietary diver- developed by FAO, respectively. The results indicated
sity and food security status in male and female-headed that cereals, tubers, oils and fats, spices, and condiments
cocoa households distributed?; 2. Where do the male are the most often consumed foods in the household.
Batame BMC Public Health (2024) 24:985 Page 4 of 18

When it comes to dietary diversity, female-headed proportionate random sample of the targeted households
households scored lower than male-headed households. based on the number of households in the village taking
There was a slight positive correlation found between part in the CARI program, yielding 204 respondents in
the household dietary diversity score and the head of the total. The authors utilized cross-sectional survey data
household’s nutritional knowledge, awareness, attitude, from the Bani and Mbarali Districts. There were 101 and
and educational status. 103 participants from Bani and Mbarali Districts, respec-
Ref. [17] explored the determinants of rural household tively. The results indicate that the foods that are most
dietary diversity in the Amatole and Nyandeni districts frequently consumed in the home are cereals, vegetables,
in South Africa. A household cross-sectional survey data oils and fats, spices, condiments, and beverages. Dietary
from 181 rural communities was collected, that is 100 diversity was lower in households led by women than it
from Amatole and 81 from Nyandeni. The household was in households headed by men. From mid-Novem-
dietary diversity score was employed to calculate the ber to March, male-headed households experience food
dietary diversity of the respondents using 12 food groups shortages. Gender and education of the household head
over a 24-hour dietary recall. The results showed that and food preparation and nutrition training were signifi-
these food groups were largely consumed: sugars (16%), cant determinants of household dietary diversity.
condiments (16%), oils (12%), potatoes (12%), grains This study by [26] looked at the connections between
(11%), and beans/peas (9%), whereas these food groups: household income, credit availability, and dietary diver-
milk (6%), vegetables (5%), eggs (4%), meats (3%), fruits sification. Dietary diversity was calculated using the
(3%) and fish (2%) were least consumed. The multinomial Food Diversity Index and Food Consumption Score.
logistic regression model predicted that participation in The authors used 5779 and 8312 household survey data
irrigation schemes, gender, education, income, access to from the 4th and 5th rounds of the Ghana Living Stan-
home gardens, and ownership of small livestock attain- dards Survey. The findings revealed that access to credit/
ment positively influenced high dietary diversity. income was found to be positively correlated with dietary
Ref. [25] examined the factors that determine house- diversity since having an income greatly influences the
hold food security in the Sekyere-Afram Plains District food types households consume in Ghana.
of Ghana. The authors randomly selected 4 communi- Ref. [19] study aimed to investigate, from a gender
ties and interviewed 25 households per community, so in viewpoint, the dietary diversity among Nigerian rural
total interviewed 100 households. The USDA Household households. The Simpson Diversity Index, cross-tabu-
Food Security Scale was utilized to estimate the food lation, and Oaxaca-Blinder decomposition were used to
security status, and a binary logistic regression was used analyze data from the Living Standard Measurement Sur-
to predict the determinant of food security. Household vey-Integrated Survey on Agriculture 2016. The results
size, farm size, off-farm income, credit access, and mari- unveiled that the score for dietary diversity was some-
tal status were positively related to food security. what higher in households headed by women than in
Ref. [3] analyzed dietary diversity in rural areas, espe- households headed by men. In households led by women,
cially in the case of indigenous communities in Sierra food share expenditure was highest for fish and seafood,
Tarahumara, Mexico employing household dietary diver- whereas in homes headed by men, it was highest for
sity score. The 123 survey data was collected in February cereals. Households headed by males aged 31 to 40 and
and March 2015. Cereals (100%) were the food group comprising a minimum of 16 members had low dietary
most consumed by the study sample, followed by legumes diversity, whereas households headed by females aged
or nuts (96.7%), eggs (78.9%), sugar/honey (78%), and 51 to 60 and comprising 11 to 15 members displayed the
oils/fats (77.2%). The best variables that positively influ- highest level of food diversity.
enced dietary diversity were food expenditure per capita, Ref. [27] argued that before the cocoa harvest, many
part-time work, the Prospera program, and the head of households in West Africa that cultivate cocoa go
household’s marital status. through a “lean season,” which makes them susceptible
Moreover, [1] predicted the variables that influence to a range of stressful situations, chief among them being
dietary diversity and the potential role of men in improv- food insecurity. The author investigated the effects of
ing household nutrition in Tanzania. The study used income distribution and intra-household dynamics on
a multi-stage sampling technique, which comprised 4 household resilience during the lean season, primarily
stages: (i) Bani and Mbarali were purposively selected using qualitative data from Côte d’Ivoire. Also, the [28]
because of their prevalence of poverty and malnutrition; study’s goal was to examine the factors that influence
(ii) From the list generated by the Competitive African household dietary diversity and consumption patterns
Rice Initiative (CARI) program administrators and Dis- in the Yayu biosphere reserve in southwest Ethiopia. A
trict Agriculture Irrigation and Cooperatives Officers cross-sectional survey including 183 randomly chosen
(DAICOs), 20 villages were chosen at random; and a households was carried out. Dietary diversity was found
Batame BMC Public Health (2024) 24:985 Page 5 of 18

to be low in 17.5% of the households, medium in 61.2%, resources or assets, such as land, agricultural inputs,
and high in 21.3%. The household head’s age, educa- fiancé, and labor [1, 31]. This situation is no different
tion, income, home gardening skills, access to irrigation, from that of the Juaboso-Bia cocoa landscape (Fig. 1).
and awareness of dietary diversity all had a positive and Female-headed households may have limited or no access
significant impact on household dietary diversity, but to resources, such as credit, labor, land, extension offi-
distance from the market had a negative impact. In addi- cers, and farm inputs, which in the long run may lead to
tion, [29] analyzed household dietary diversity and its low dietary diversity and food insecurity [27]. The male-
determinants in Finote Selam town, north west Ethiopia. headed households may have full access to and control
Cross-sectional data was gathered from 400 households over resources and that will improve and cause those
between 22nd to 30th August 2015. A logistic regres- households to have high dietary diversity and be food
sion model was used to determine the factors related secure.
to household dietary diversity. Low, medium, and high
dietary diversity scores were found in 11.8%, 67.2%, and Methods
21% of the households, respectively. Family head (male- Location of the study area
headed), eating frequency, and lack of cooking water The Juaboso-Bia cocoa landscape is one of the cocoa
were found to be substantially correlated with dietary landscapes in the Westward of Ghana. The Westward
diversity in households. contributed approximately 43% of the total 766,977 tons
Ref. [30] examined 479 smallholder farmers in the Zim- of annual regional cocoa purchases in 2019/2020 to
babwean provinces of Manicaland and Masvingo’s level of Ghana’s cocoa production [34]. The Juaboso-Bia cocoa
crop diversification as well as the factors that contribute landscape encompasses Bia East, Bia West, and Juaboso
to it. Estimating diversity was done using the Herfindahl districts, but for this study, only Bia West and Juaboso
index, and factors related to agricultural diversification districts were studied. The studied area is located in the
were assessed using the Tobit model. According to the Western North Region of Ghana. The area lies within 2°
gender results, households headed by men were some- 40’ W and 3° 20’ W longitude, and 6° 10’ N and 6° 50’ N
what more diverse than households headed by women. latitude (Fig. 2). The studied landscape is endowed with
According to the Tobit model, agro-ecological zone, natural resources, such as the Bia National Park and the
household income, farmer-to-farm extension, routine Krokosua Hills forest reserves [35]. The population of the
extension, farming experience, access to markets, mem- landscape is estimated to be 147, 374 by [35]. The major
bership in a farmers group, gender of the head of the rainfall season is between May and June and the minor
household, education, number of livestock units, access season is from September to October [8]. The occupa-
to irrigation, and farms on level terrain are all important tion of the inhabitants is predominately agriculture, with
factors in increasing crop diversification. cocoa being the largest cash crop cultivated by almost
Lastly, [18] examined factors affecting dietary diver- every household [35]. The other food crops grown are
sity in the Usambara Mountains, Tanzania. The authors plantain, cassava, yam, cocoyam, oil palm, maize, rice,
employed in-depth interviews, focus group discussions, etc. Some households occasionally intercrop these food
and participant observation. Researchers found that par- crops with cocoa farms, while others produce them on
ticipants strongly agreed on the importance of dietary a small parcel of land for livelihood. The Juaboso-Bia is
diversity in maintaining and enhancing hunger over days, regarded as one of the hotspot zones for cocoa expan-
months, and seasons. According to locals, having enough sion, which is affecting food crop production [8–10, 36];
financial resources, agrobiodiversity, terrain heterogene- thus was selected to study the food security and dietary
ity, and a diversity of livelihoods all helped them be able diversity status of male and female-headed cocoa house-
to eat a diversified diet and maintain decent nutritional hold heads.
status. Seasonality (climate change), household size, tra-
dition/cultural beliefs, and gender were other factors that Research design
influenced nutrition and dietary diversity. Mixed research methods integrate qualitative and quan-
titative research designs employed to analyze phenom-
Conceptual framework ena and may occur in different approaches [37]. With
This study employed the conceptual framework by [31] to this technique, it is possible to assess the consistency of
explain the relationship between gender and food secu- data collected using a variety of instruments, compare
rity. Food and nutrition security status of female and and contrast one method’s findings with another, and
male-headed households are diverse globally, nation- show how the results of one approach can affect other
ally, and locally [31–33]. The differences in the food or research strategies, for example, qualitative data could
nutritional security status of male and female-headed help researchers to well describe their findings from
households are influenced by access to and control over quantitative data [37, 38]. The mixed methods ensure a
Batame BMC Public Health (2024) 24:985 Page 6 of 18

Fig. 1 Conceptual framework illustrating factors influencing food security and dietary diversity status, with a focus on differences between female and
male-headed households. Source: (adapted 31])

Fig. 2 Map showing the Juaboso-Bia cocoa landscape within the Western Region of Ghana
Batame BMC Public Health (2024) 24:985 Page 7 of 18

comprehensive understanding and enhance the validity as give way to the power dynamics, desires, and motives
and credibility of the research [37]. For the mixed meth- that shape how certain places, people, and events are
ods, the quantitative aspect through the use of question- portrayed [38]. The purposive sampling technique was
naires was used to find answers to the questions, namely used to select 14 participants to engage in the in-depth
household head dietary diversity and food security sta- interviews, which comprised 7 male and 7 female-headed
tus distribution, sources of food, food production status, cocoa households that had over 20 years of cocoa farm-
factors affecting crop diversification, and food shortage ing experience. The study selected 2 participants (a male
periods. The qualitative approach (in-depth interviews) and female) from each of these purposively selected com-
was used to find answers to these questions, such as rea- munities namely; Juaboso, Bonso Nkwanta, Benchema,
sons/themes for low household dietary diversity/food Adjuafua, Yawmatwa, Essam-Debiso, and kojoaba.
security status, opinion on crop diversification, and their
food unavailability experiences and production status. Semi-structured questionnaires analysis
This qualitative approach was necessary since it captures The survey data was numbered, cleaned, and coded in
the experiences and views of respondents. Statistical Package for Social Sciences (SPSS) for the fol-
lowing analysis.
Data collection
This study collected primary data from the cocoa farmers Dietary diversity and food security status analysis
in the Juaboso-Bia landscape in the form of quantitative The food security and dietary diversity status of the cocoa
and qualitative data. With the quantitative method, the household heads were measured using the Household
sample size was calculated using the formula by [39], that Dietary Diversity Score (HDDS) [17, 21, 45]. The HDDS
is n = (N / 1 + N(a)2) because this formula ensures that included 12 local food groups (see Appendix 1 ) that
the sample size is sufficient to provide reliable and accu- were used to analyze the dietary diversity and food secu-
rate results. Where the margin error (a) was 5%, the sam- rity of the male and female-headed cocoa households in
ple size (n), and the sample population (N) was 15,491 the study area. A household was scored 1 if that house-
cocoa households. The confidence level was 95% since hold consumed any of the 12 food groups in the past 24-
this study involved human beings whose truthfulness of hours, otherwise 0. The total HDDS was 12, which was
the information is vulnerable to prejudices. Now, sam- further categorized into low dietary diversity (0–4 food
ple size = (15,491 / 1 + 15,491 (0.05)2) = 390 households. groups); medium dietary diversity (5–8 food groups); and
The study interviewed 200 out of the 390 households high dietary diversity (9–12 food groups). For the food
due to largely limited financial resources and road inac- security status estimation, any household that scored
cessibility. The study administered 200 semi-structured less than 6 HDDS was classified as food insecure, while
questionnaires in the Juaboso-Bia landscape. Follow- more than 6 HDDS was classified as food secure [16, 21].
ing [1, 40–42], and [43], the multi-stage sampling tech- Also, other concepts of food security, such as food avail-
nique was used to select the cocoa household heads from ability and accessibility were analyzed. The results were
these communities, particularly, Yawmatwa, Adjuafoa, presented in descriptive statistics. Chi-Square was used
Kojoaba, Asuopiri, Akaatiso, Nkrabea, Bechema, Bonso to test the association between food security status and
Nkwanta, and Juaboso. With the multi-sampling tech- household head gender.
nique, firstly, the study area was divided into 4 groups
namely north, south, east, and west to ensure geographi- Variables affecting crop diversification
cal representation and reduce biases. Secondly, purposive A binary logistic regression model was utilized to predict
sampling was used to select the aforementioned commu- the surveyed household heads’ socio-economic factors
nities based on road accessibility. Lastly, the simple ran- influencing their opinions on the impact of cocoa expan-
dom technique was applied to select the cocoa household sion on crop diversification in the study area, following
heads to be interviewed. The semi-structured question- several studies [8, 21, 25]. The binary logistic regression
naires (see supplementary material 1) covered the house- model was CECDi = βi + β1 × 1 + β2 × 2 + β3 × 3 + β4 × 4 + β5
hold heads’ dietary diversity score, food shortage periods, × 5 + β6 × 6 + β7 × 7 + β8 × 8 + β9 × 9 + β10 × 10+e….(1) where;
food production status, and demographic data (sex, age, CECD = Cocoa expansion affects crop diversification (No
education, etc.). “0”, Yes “1”); βi = the constant term; β1−10 = the coefficient
With the qualitative method, an in-depth interview data of the independent variables. The independent variables
collection method was utilized, for instance, the usage of used in the model included:
an in-depth interview approach aids in understanding X1 = Sex of the household head ( Female “0”, Male “1”).
the realities and social worlds that actors encounter daily, X2 = Age of the household head in years.
including their lived experiences, complexities, negotia- X3 = Education status of the household head (No formal
tions, views, conflicts, and shared meanings [44], as well education “0”, Formal education “1”).
Batame BMC Public Health (2024) 24:985 Page 8 of 18

X4 = Marital status of the household head ( Other “0”, food insecurity among male and female-headed cocoa
Married “1”). households.
X5 = Household size in numbers.
X6 = Cocoa cooperation membership ( No “0”, Yes “1”). Results
X7 = Cropland conversion (No “0”, Yes “1”). Respondent’s socio-economic characteristics
X8 = Credit access (No “0”, Yes “1”). The responses (Table 1) from the surveyed household
X9 = Total household income in cedis. heads showed that 39.5% and 60.5% were headed by
X10 = Access to extension officers (No “0”, Yes “1”). females and males, respectively. Most of the interviewed
household heads aged between 46 and 66 years (58.5%).
Respondents’ in-depth interviews analysis For the educational levels of the household heads, 10%,
The information from the interviews was categorized 26.5%, 13.5%, and 17.5% obtained primary, JHS/Form 4,
into themes and used to validate and support the quan- SHS/SSS, and Tertiary, respectively, as well as 32.5% had
titative findings. The thematic analysis was employed to no formal education. Moreover, most of the respondents
identify the factors influencing low dietary diversity and had access to extension officers (87%) and credit (70.5%).
Also, 81.5% of the household heads had converted their
Table 1 Socio-economic characteristics of the household heads cropland to cocoa farms, and 75.5% of them had joined
in the Juaboso-Bia cocoa landscape cocoa cooperative groups. In terms of the total house-
Socio-economic variables Valid Valid Mean hold income, the majority of the surveyed household
frequency per- heads earned less than 10,000 cedis (52.5%). Further-
cent more, 90.5% of the household heads perceived that cocoa
(%)
expansion affects crop diversification in the study area.
Sex Female 79 39.5 0.61
Male 121 60.5
Distribution of dietary diversity and food security pattern
Age between 25 and 45 44 22 56.14
among cocoa household heads in the Juaboso-Bia cocoa
between 46 and 66 117 58.5
landscape
above 67 39 19.5
In the Juaboso-Bia cocoa landscape, it was observed that
Education Primary 20 10 0.68
there was a disparity distribution in the amount of the 12
JHS/Form 4 53 26.5
SSS/O or A-level 27 13.5
food groups consumed in households headed by women
Tertiary 35 17.5
than that households headed by men. According to Fig. 3,
No formal education 65 32.5
the male-headed households recalled eating roots/tubers
Marital status Married 148 74 0.76 (58%), vegetables (55.5%), oil/fat/butter (49.5%), local
Single 8 4 grains (42.5%), seafood (36.5%), and sugar/honey (30%),
Divorced 14 7 respectively. These food groups were averagely consumed
Widow/widower 30 15 in male-headed households, particularly organic meat,
Household size between 1 and 5 55 27.5 8.14 fruits, spice/tea/condiments, and eggs, ranging from 22
between 6 and 10 98 49 to 29.5%. Legumes/nuts (13%) and milk products (18.5%)
above 10 47 23.5 were consumed less. In an interview with a participant, it
Cocoa No 43 21.5 0.79 was indicated that;
cooperation Yes 157 78.5
membership “We eat vegetable stew with cassava, yam, cocoyam,
Cropland con- No 37 18.5 0.81 and maize most of the time because it is our basic
version to cocoa Yes 163 81.5 diet. I grow these foods on my family farm, so I don’t
Credit access No 59 29.5 0.70 always have to buy them. I get these food items
Yes 141 70.5 from the market when I run out because they are
Access to exten- No 26 13 0.87 inexpensive”(Key informant interview, 23 January
sion officers Yes 174 87 2023).
Total household less than 10,000 105 52.5 1.82
income Between 10,000 and 26 13
30,000
Above 30,000 69 34.5
The households headed by women also reported these
Household heads’ perception
food groups being consumed the most namely roots/
Cocoa expan- No 19 9.5 0.90 tubers (38%) and vegetables (37%). The food groups con-
sion affects crop Yes 181 90.5 sumed on average were seafood (27%), and local grains
diversification (22%). The least consumed food groups were sugar/honey
Source: Fieldwork, 2023 (17.5%), oil/fat/butter (17%), spices/tea/condiments
Batame BMC Public Health (2024) 24:985 Page 9 of 18

Fig. 3 Recalled household food groups eaten in male and female-headed cocoa households. Source: Fieldwork, 2023

Table 2 Dietary diversity status among the male and female- Table 3 A Chi-Square test statistics on the relationship between
headed cocoa households in the Juaboso-Bia cocoa landscape food security status and gender of the household head
Household dietary diversity status Gender Food security status and gender cross tabulation
Female Male Food security status Gender
Low dietary diversity (0–4) 60.8% 37.2% Male Female
Medium dietary diversity (5–8) 19% 30.6% Secure 47.9% 29.1%
High dietary diversity (9–12) 20.2% 32.2% Insecure 52.1% 70.9%
Total 100% 100% Total 100% 100%
Source: Fieldwork, 2023 Chi-Square Tests Statistics
Value df Asymp. Sig. Exact Sig. Exact Sig.
(16.5%), eggs (12.5%), organic meat (8.5%), fruits (7.5%), (2-sided) (2-sided) (1-sided)
legumes/nuts (7%), and milk product (4.5%). For exam- Pearson 9.575a 1 0.002
Chi-Square
ple, a female headed-household noted that;
Continuity 8.691 1 0.003
“I never really consider the meats or eggs (quality) Correctionb
in the meals I offer to my kids because I’m always Likelihood Ratio 9.774 1 0.002
more concerned about the quantity. We can buy a Fisher’s Exact Test 0.002 0.001
lot of cassava or maize and that will provide us with Linear-by-Linear 9.527 1 0.002
enough food to last us for approximately a month Association
Number of Valid 200
with the money that will be used to purchase meats,
Cases
milk, eggs, fruits, and legumes. These food groups
are extremely expensive to buy here on the market,
Following Table 3, 29.1% of the female-headed cocoa
and they won’t even satisfy my kids as much as local
households were food secure as compared to food inse-
cereals and roots/tubers do”(Key informant inter-
cure (70.9%). Conversely, 52.1% of the male-headed
view, 20 January 2023).
households were food insecure while 47.9% were food
secure. With the chi-square test, the gender group com-
According to Table 2, low dietary diversity was more prised male and female-headed households, which was
prevalent in female-headed households (60.8%) than compared to food security status that is secure and inse-
the male-headed households (37.2%). Medium and high cure. The p-value for the chi-square analysis was sta-
dietary diversity categories were high in the male-headed tistically significant, therefore the null hypothesis was
households, specifically 30.6% and 32.2%, compared to rejected. There is a significant relationship or differ-
the female-headed households, such as 19% and 20.2%, ence between food security status and the gender of the
respectively. household head.
Batame BMC Public Health (2024) 24:985 Page 10 of 18

Food unavailability experience of male and female-headed Table 4 Food unavailability experience of male and female-
cocoa households headed cocoa households in the Juaboso-Bia cocoa landscape
Some of the male-headed households (66.1%) and female- in 2023
Food unavailability experience Gender
headed households (83.5%) indicated that they face food
Female Male
shortages or unavailability in the study areas (Table 4).
Yes 83.5% 66.1%
They expressed that food shortage or unavailability expe-
No 16.5% 33.9%
rience could be severe, average, and low according to cer-
Total 100% 100%
tain months.
Source: Fieldwork, 2023
The months that the male-headed households recalled
facing severe food unavailability were July, June, April,
Table 5 Sources of staple food for male and female-headed
and May, however, for the female-headed households,
households
the severe months were June, July, March, February, and
Sources of staple food Gender
January (Fig. 4). The average severe food shortage months Female Male
were March, August, February, January, and September Self-production 10.1% 30.6%
for the male-headed households, and May, April, and Market 8.9% 4.1%
December for female-headed households. The low severe Self-production and market 81.0% 64.5%
food unavailability months for female-headed house- Others 0% 0.8%
holds were August, September, October, and November. Total 100% 100%
On the other hand, December, November, and October Source: Fieldwork, 2023
were less recorded food shortage months for households
headed by men. For instance, in an interview with the Sources of staple food, accessibility to food, food
head of a cocoa household, he shared that; production status, and dietary diversity nexus among
cocoa household heads
“When the rainy season approaches, especially
Households headed by males and females in the stud-
in June and July, I am really frightened because
ied area expressed that they obtain their staple foods
the rains are exceptionally heavy and threaten to
through self-production, market, both self-production
destroy my food crops. Also, the food I store for the
and market and others. According to Tables 5 and 81% of
rainy season is consumed quickly, causing my fam-
the female-headed households got their food from both
ily to eat fewer meals each day. We occasionally
self-production and the market, 10.1% from self-produc-
go without food to eat. My home experiences food
tion, 8.9% from the market, and no household indicated
shortages throughout the dry season in December to
through others (0%). In addition, for the female-headed
March because the weather is so dry that the crops
households that said they obtained their food from the
are unable to flourish” (Key informant interview, 20
market, 53.6% of the households walked less than 30 min
January 2023).
within the community and 26.1% got it from the neig-
bouring communities (Fig. 5).

Fig. 4 Food unavailability months per gender of the household head in the Juaboso-Bia landscape. Source: Fieldwork, 2023
Batame BMC Public Health (2024) 24:985 Page 11 of 18

Fig. 5 Travel distance covered by the sex of the household head to get staple food. Source: Fieldwork, 2023

Table 6 Sources of local food and dietary diversity status nexus my household and I consume” (Key informant inter-
Sources of staple food Dietary diversity status view, 26 January 2023).
Low Medium High
Self-production 34.4% 19.3% 5.5% Concerning the food production status for the past 5
Market 10.8% 3.8% 0% years (Fig. 6), the study revealed that the female-headed
Self-production and Market 53.8% 76.9% 94.5% households had 3.8%, 19%, and 77.2% as not noticeable,
Others 1.0% 0% 0% increased, and decreased, respectively. The male-headed
Total 100% 100% 100% households, on the other hand, had 48.8% as decreased,
Source: Fieldwork, 2023 13.2% as not noticeable, and 38% as increased.
With regards to Table 7 results, the households that
Interestingly, for the male-headed households, 0.8% of experienced “not noticeable” food production status had
them obtained food through others, while 4.1%, 30.6%, 5.4%, 25%, and 1.8% as low, medium, and high dietary
and 64.5% had food via market, self-production, and both diversity, respectively. Again, households with decreased
market and self-production, respectively. Also, out of the food production status witnessed 94.6% as low, 55.8% as
male-headed households that stated they purchased their medium, and 5.5% as high dietary diversity status. The
food at the market, 46.4% claimed to have walked less majority of the households that obtained increased food
than 30 min within the community to get food and 73.9% production status had 92.7% high dietary diversity sta-
got their food from neigbouring communities (Fig. 5). tus, with only 0% and 19.2% as low and medium dietary
According to Table 6, within the households that diversity status, respectively.
obtained their foodstuffs solely from the self-production In an interview with a household head who had
source, 34.4% had low, 19.3% had medium, and 5.5% increased food production status disclosed that;
obtained high dietary diversity categories. For house-
“Over the past 5 years, my household members and
holds that relied on the market for staple food, 10.8%,
I have managed our piece of land very well by prac-
3.8%, and 0% encountered low, medium, and high dietary
ticing sustainable farming methods. We managed to
diversity categories, respectively. Also, 53.8%, 76.9%, and
grow cocoa and a variety of food crops on the small
94.5% experienced low, medium, and high dietary diver-
land. Because of that, my household eats different
sity categories, respectively in households that depended
diets daily and has enough food always, which has
on both the market and self-production for foodstuffs.
ensured my household’s food and nutrition security”
A respondent shared that;
(Key informant interview, 2 February 2023).
“Hahaha! I cannot cultivate all the food crops in
this whole place. So, I grow some staple crops, such
as cassava, plantains, cocoyam, tomatoes, garden
eggs, and pepper, and buy protein, condiments, and
fruits like chicken, meat, milk, fish, and oranges from
the market. Thus, supplementing my self-production
with goods from the market boosts the quality of diet
Batame BMC Public Health (2024) 24:985 Page 12 of 18

Fig. 6 Food production status for the past 5 years among male and female-headed cocoa households. Source: Fieldwork, 2023

Table 7 Food production status and dietary diversity status “Climate change has resulted in the death of our
nexus cocoa trees and food crops since the crops receive too
Food production status Dietary diversity status much sunlight hence declining the yields generated
Low Medium High from cocoa and food production. Due to the sudden
Not noticeable 5.4% 25% 1.8% arrival of the rains and dryness, we are currently
Decreased 94.6% 55.8% 5.5% unsure of what to plant. Even some of our poultries
Increased 0% 19.2% 92.7% die as a result of excessive heat and do not get to eat
Total 100% 100% 100% their products” (Key informant interview, 22 Janu-
Source: Fieldwork, 2023 ary 2023).

Multifaceted themes influencing low dietary diversity

and food insecurity in male and female-headed cocoa Poverty
households in the Juaboso-Bia cocoa landscape The results of the study showed that many of the cocoa
Climate change/ variability farmers expressed that they lacked the money to diver-
Climate change/ variability was among the common sify their diets and also to buy farm equipment to boost
themes that emerged from the interviews, which has their food production in the study area. In the course of
negatively impacted many of the cocoa households. For the interviews, it was disclosed that financial constraints
instance, through this study, it was revealed that the food limited the variety of food the household heads provided
diversity of both male and female-headed households is to their members. For example, these two participants
impacted by some seasonal variations that do not make respectively stipulated that:
certain crop types flourish. One of the interviewees
“Hmm!, I have been farming cocoa for almost 3
stated that;
decades and I cannot boast of eating a well-bal-
“I remember when I was young, my father used to anced diet every day. My household only eats starchy
cultivate a lot of foodstuffs even to the point of sell- foods, such as cassava, plantain, yam, and vegeta-
ing some to other households. This was because there bles, such as tomatoes and garden eggs, because that
was an abundant and reliable rainfall pattern. is what I can afford” (Key informant interview, 26
However, in this era, the rains fall unexpectedly and January 2023).
heavily to the extent of destroying most grown crops “Sometimes, I feel shy to tell my non-cocoa farmers
on my farms. Due to this, my household eats what is friends that I have not even eaten chocolate before,
available and mostly not rich in vitamins and not a and I consume chocolate drinks occasionally. I can-
balanced diet” (Key informant interview, 20 January not even feed my household members with cheap
2023). food, for example, “abom” (literally means vegeta-
ble stew with either yam or cocoyam or cassava) let
alone talk of eating fruits, eggs, and meat. Due to
Another household head mentioned that;
traders’ perception that this area is a wealthy cocoa
district, these food items are very expensive to pur-
Batame BMC Public Health (2024) 24:985 Page 13 of 18

chase. Also, buying modern farm tools, fertilizers, “I have about 10 household members and due to
and hiring labor is expensive, which affects our food that I barely buy food items rich in vitamins because
crop and cocoa production ” (Key informant inter- those food items do not satisfy me and my family.
view, 2 February 2023). Thus, I regularly buy less expensive food usually
starchy food, which is not well rich in vitamins since
all my children can eat and get satisfied” (Key infor-
Unavailable lands
mant interview, 31 January 2023).
Lack of land was also a key dimensions that was revealed
from the interviews with the participants. The study
found that the unavailability of land has hindered many Gender stereotypes/ beliefs
household heads from growing a variety of crops, hence One of the key facets that played a major role in promot-
has influenced the diets consumed by these households. ing low dietary diversity among female and male-headed
The interviewees noted that the Juaboso-Bia landscape households was gender stereotypes/ beliefs, according to
is full of forests and cocoa plantations, and they cannot the study’s findings. The interviews revealed that both
either go into the forest to cultivate food crops or cut men and women face gender stereotypes and that has
down their cocoa trees to grow food crops. For example, prevented many of them from working hard to provide
in an interview with a respondent, it was revealed that; good diets for their household members.
One participant shared that;
“Currently, we have used all the lands for cocoa pro-
duction and there is no available land in this district “Before the demise of my husband, we ate a well-
to grow food. I gave up my land for cocoa produc- balanced diet almost every day. After his death, the
tion to increase the income and food diversity for my household responsibility has become very heavy for
family, but that goal was never realized, and now me to feed myself and my children balanced diets.
my family is starving and unable to consume cocoa Because I am the head of the family, I am most times
because it is not edible” (Key informant interview, 1 denied access to some of the cocoa farms, extension
February 2023). officers, and other resources to increase my cocoa
yields, hence boosting my household dietary diver-
sity” (Key informant interview, 24 January 2023).
In addition, one participant said that;
“The government through the forestry commission
In another interview with one of the participants, he
allocated portions of the off-reserved forest to us to
revealed that;
cultivate our food crops some time back. The condi-
tion was to replant trees after harvest before shifting “I do occasionally find it challenging to give my fam-
to a new place to begin another cultivation. Failure ily a variety of foods to consume every day to grow
to do so will lead to the prevention of cocoa farm- well and strong, even though I am a man and the
ers from entering the off-reserved to grow food crops. leader of my household. This is because some of my
Some farmers failed to comply with the condition male friends point hands at me and call me all sorts
regardless of the number of warnings given, there- of names for having a small land size, being unable
fore we were restricted from growing food crops there to cultivate different crops, and being a cocoa care-
and this has affected my household food production” taker for other households headed by men” (Key
(Key informant interview, 26 January 2023). informant interview, 28 January 2023).

Large household size Perceptions of the cocoa household heads on the

According to the findings of this study, large household implication of cocoa expansion on crop diversification in
size was among the themes that promoted low dietary the Juaboso-Bia cocoa landscape
diversity in the Juaboso-Bia landscape. The respondents Table 8 below provides the parameter estimates for the
expressed that household heads with large family sizes binary logistic model. Among the 10 variables used to
find it cumbersome to provide quality and different food estimate the logistic model, only 3 variables were statis-
groups for their households compared to household tically significant, namely sex, cropland conversion to
heads with small household sizes. For instance, during cocoa, and total household income.
the interview, one participant indicated that; In line with this, a respondent transpired that;
Batame BMC Public Health (2024) 24:985 Page 14 of 18

Table 8 Logistic regression model predicting the influence households, whereas only roots/tubers and vegetables
of household heads’ socio-economic characteristics on were mostly eaten in female-headed households. This
their perceptions of the impact of cocoa expansion on crop food group pattern indicates that both male and female-
diversification in the Juaboso-Bia landscape
headed households heavily rely on roots/tubers, local
Variables Coefficient Significance Odd ratio
grains, and vegetables because of their low cost and self-
Constant 4.691 0.019 108.938
production [2]. This finding was similar to that of [2] in
Sex (Male) -0.982 0.10* 0.405
Ecuador and [18] in Tanzania (East Usambara Moun-
Age -0.041 0.196 0.960
tain). However, it disagreed with the findings of [24], who
Education status -0.338 0.562 0.713
Marital status -1.215 0.164 0.279
revealed that green vegetables are hardly eaten in rural
Household size 0.096 0.718 1.101
households.
Cocoa cooperation -0.552 0.452 0.576
The seafood group, such as fish which is rich in pro-
membership tein and is good for human growth and development
Cropland conversion 1.079 0.093* 2.942 was highly consumed in male-headed households and
Credit access 0.681 0.206 1.975 average in female-headed households. This could be
Total household income 0.645 0.050** 1.906 partially explained by the fact that male-headed house-
Access to extension officers -1.290 0.263 0.275 holds had more income sources and financial capacity
Pseudo R2 0.71 than female-headed households [31]. This present find-
-2 Log Likelihood 85.46 ing was inconsistent with the findings of [1] in Tanzania,
Significance 0.007 [46] in Botswana, and [3] in Mexico. Also, [19] found that
Observation 200 female-headed households consumed more seafood than
Significance * (10%), **(5%), ***(1%) male-headed households, which was contrary to this
present study.
“The cocoa has eaten our croplands, so I do not have Food groups namely eggs, milk products, organic meat,
a variety of food to eat. The land I could have used legumes/nuts, and fruits are essential for the human
to grow crops, such as cassava, vegetables, yam, body daily because of their protein, vitamins, and miner-
and rice, I have used it to expand my cocoa farm. I als content [47]. Regardless, these food groups were least
believe when we continue to expand cocoa, we will eaten in female-headed households, which is largely due
one day wake up with no different kinds of food to to lack of income or poverty, and was corroborated by
consume. Because of this, the prices of foodstuffs on the findings of [2]. The lack of this food group consump-
the market are expensive to buy” (Key informant tion in female-headed households could imply that such
interview, 28 January 2023). households experience malnutrition [15]. On the other
hand, these food groups were relatively highly consumed
Another household head indicated that; in male-headed households, and this finding was con-
trary to that of [1].
“A few of the cocoa households here do crop diversi-
Concerning the household dietary diversity status of
fication. The reason why this has not been achieved
this study (Table 2), the general dietary diversity pat-
is that the cocoa sector is lucrative that is we earn
tern distribution was skewed towards the males and
more than cultivating only food crops. I remem-
unfairly towards the female-headed households. Specifi-
ber when I used to do more food crop farming, I
cally, the low dietary diversity was more widespread in
used to find it difficult to get people to come and
female-headed (60.8%) households than in their male
buy the foodstuffs so most of them ended up spoil-
counterparts (37.2%). The combination of medium and
ing, and that affected my income. Hmm, so I decided
high dietary diversity categories in male-headed house-
to do more cocoa farming, and I earn a lot from it.
holds was 62.8%, which is statistically greater than what
Regardless, this to a large extent is affecting the vari-
female-headed households obtained (39.3%). Gener-
ety of diets my household eats” (Key informant inter-
ally, the male-headed households had a relatively good
view, 31 January 2023).
dietary diversity (62.8%) even though 37.2% had low
dietary diversity. This connotes that the male-headed
households consumed relatively high dietary diversity,
Discussion which improve their food and nutrition security than
Household head gender and its implication on dietary female-headed households. This evidence confirmed
diversity distribution and food security pattern the findings of [24] and [1], who found that low dietary
The findings from this present study suggest that roots/ diversity is more frequent in female-headed households
tubers, vegetables, oil/fat/butter, local grains, seafood, than in men-headed households in Uttar Pradesh State,
and sugar/honey were highly consumed in male-headed India, and Tanzania (Bahi District and Mbarali District),
Batame BMC Public Health (2024) 24:985 Page 15 of 18

respectively. In addition, this was not in agreement with hardship during times of famine than households that
the findings of [19], who revealed that female-headed depend on both self-production and the market. This
households had relatively higher dietary diversity than implies that households need to supplement their self-
male-headed households. production with produce from the market to ensure their
Moreover, findings from the study (Table 3) suggested food and nutrition security all year round. This finding
that food insecurity is prevalent in female-headed cocoa is in line with that of [20], who revealed that while farm
households (70.9%) compared to their male counter- production diversity alone has a negligible impact on
parts (52.1%). This finding conforms to that of [21], who dietary diversity, it has a higher impact when combined
found that 61% of the female-headed cocoa households with goods bought from the market.
in Wassa Amenfi West District were more food insecure
than male-headed households (33%). Also, the results in Exploring the multidimensional nature of low dietary
Table 3 confirmed the conceptual framework adapted diversity and food insecurity in male and female-headed
from [31] in Fig. 1, which suggested that households cocoa households in the Juaboso-Bia cocoa landscape
headed by men are more food secure than female-headed Climate change/ variability
households because male-headed households have con- One of the key determinants of low dietary diversity
trol over resources, such as farm inputs, land, and labor. identified by the study was climate change. The abrupt
The findings also implied that cocoa households with changes in the weather in the study area have negatively
food-secure status mostly have high dietary diversity and affected many cocoa household heads in diverse ways,
low dietary diversity is usually associated with house- leaving the afflicted households vulnerable to food inse-
holds with food-insecure status. curity and low dietary diversity. Seasonal variations usu-
ally do not make certain crop types flourish, which has
Food unavailability experiences in relation to gender of impacted the food diversity of cocoa household heads.
the household head This finding was aligned with that of [18], who discovered
Food shortage or unavailability was prevalent in April, that food scarcity or unavailability caused by changes in
May, June, and July for male-headed households, but seasonality affects the diversity of the locals’ diets in the
female-headed households experienced it in February, East Usambara Mountains, Tanzania.
March, June, and July. Both households shared simi-
lar months of food unavailability, such as June and July. Poverty
May, June, and July happen to be the major rainy season, Both male and female-headed households agreed that
which usually destroys food crops and make poor house- poverty is an important factor that affects a household’s
holds vulnerable to food insecurity [8]. Again, due to the dietary diversity. A household with a higher income
rainy months, from July to September, [27] discovered could afford to buy different crop seeds, buy more land,
that households in the cocoa industry in Ghana go with- and hire more farm laborers on his or her farms, thereby
out food during those months, and the income generated increasing the dietary diversity and food security of the
by cocoa sales also runs out before the following harvest. household. Many of the cocoa household heads expressed
The dry season which is from December to March dries that they lacked the money to diversify their diets in the
most food crops and makes households food insure. study area. Because of poverty, most of these households
Some households eat one meal, usually starchy or car- experienced low dietary diversity and food insecurity,
bohydrate foods (such as cassava, yam, and cocoyam) a and in other studies, similar results have been discovered
day during the long dry season. Similarly, [1] found that [2, 8, 18].
male-headed households in Tanzania’s Bahi District and
Mbarali District had food shortages from mid-November Unavailable lands
to March, to the point that such households consume just Lands are a major resource needed for agricultural activi-
one meal per day, typically consisting of ugali, cassava, ties. A household’s dietary diversity can be influenced by
and rice. the size of the land since a greater land area would allow
for the growth of more different crops, protecting the
Implications of sources of staple food on household household’s food security and dietary diversity [2]. The
dietary diversity status local people in the Juaboso-Bia landscape complained
The types or sources of food households rely on greatly that there was no available land for them to cultivate food
determine the status of the dietary diversity of the house- crops since they used all their lands for cocoa production.
hold either low, medium, or high. As evident in Table 6, Most of them cultivate food crops under small cocoa
households that rely primarily on the market or self- trees, which is not enough for their sustenance, thus
production have a higher propensity to have a low level influencing their dietary diversity status. Similar findings
of dietary diversity because they may experience greater were made by [8], who found that no land was available
Batame BMC Public Health (2024) 24:985 Page 16 of 18

for crop farming in the Juaboso District because the cocoa farms to reduce their burden in meeting their
cocoa farmers had turned their food cropland into cocoa needs. This finding is supported by [30] who found that
farms for a variety of reasons that harmed their ability to female-headed households are negatively associated with
feed themselves. Their studies further noted that some crop diversification, that is they do not grow different
cocoa farmers grow food crops in portions under cocoa types of crops as compared to male-headed households
rehabilitation, and some households barely get food to in Zimbabwe.
eat in a day. Cropland conversion to cocoa had a statistical sig-
nificance with the dependent variable at 5%. This result
Large household size suggests that an increase in the rate of cropland conver-
The size of a household may increase or decrease the sion to cocoa would have a positive influence on the per-
household head’s ability to provide more diverse food for ception of household heads on the implication of cocoa
his or her household members. The large household size expansion on crop diversification. That is, the odd ratio
of some of the male and female-headed cocoa households increases by a factor of 2.942; therefore influencing the
accounted for the low dietary diversity in the Juaboso-Bia household heads to perceive that cocoa expansion affects
landscape. Many studies have shown that large house- crop diversification. Cropland conversion to cocoa dis-
hold size stands a higher chance of having food insecurity places croplands, limiting the ability of household heads
since the head of the household would have many mouths to grow other types of food crops. As a result, this
to feed [8, 18, 28, 29]. However, larger households may be affects household food security and diet quality of the
better able to maintain a diverse diet since they will have cocoa household heads. The study by [48] disclosed that
access to more sources of income and hence more money when cocoa household heads increase the size of their
to spend on different types of food [2]. cocoa farms, they are not able to grow different crops
and, hence become more dependent on the markets for
Gender stereotypes/ beliefs agricultural products, which raises the cost of food and
The study’s findings implied that gender beliefs had a jeopardizes their ability to feed their families. Also, this
significant impact on the food and nutrition security of study’s finding was similar to that of [8], who revealed
the male and female-headed cocoa households in the that about 80% of the cocoa farmers in the Juaboso Dis-
Juaboso-Bia landscape. This finding supports the notion trict had converted their croplands into cocoa farms,
that both men and women in rural communities do which affected the variety of food they consume.
face gender stereotypes or beliefs, which go a long way Total household income was positively related to the
to affect their production, hence leading to low dietary dependent variable at 10%. The finding implies that when
diversity [1, 18]. For example, [1] disclosed that men in the unit of household income increases, the odd ratio
Tanzania also experience gender-based stereotypes, such increases by a factor of 1.906; thus influencing the per-
as, the local people considered that men who take food ception of those household heads that the expansion of
home, namely vegetables or meat are controlled by their cocoa farms hinders crop diversification. An increase in
wives. Moreover, [18] asserted that although women who the incomes of the household heads enables them to pur-
are household heads go through gender-based obstacles, chase foodstuffs from the markets to feed their families
many of them have broken the barriers to provide good without necessarily expanding their cocoa farms to gen-
diets for their households. erate higher incomes.

Household heads’ socio-economic variables influencing Summary, conclusions and recommendations

their perception of the impact of cocoa expansion on crop This study employed both quantitative and qualita-
diversification in the Juaboso-Bia cocoa landscape tive research methods to examine the dietary diver-
According to Table 8, sex was 10% statistically significant sity and food security among male and female-headed
in influencing the perception of the surveyed household cocoa households. The study rejected the null hypoth-
heads’ of the effect of cocoa expansion on crop diversi- esis because there was a significant relationship between
fication. Sex had a negative relationship with the depen- food security status and the gender of the household
dent variable. The result implies that for every rise in the head. Through this study, it was revealed that a higher
number of households headed by females, the odd ratio percentage of male-headed households had high dietary
decreases by a factor of 0.405, influencing those house- diversity and were food secure compared to female-
holds to perceive that cocoa expansion does not affect headed households. Also, roots/tubers, local grains,
crop diversification. This can be partially explained by and vegetables happened to be the top three consumed
the fact that in rural cocoa areas, households headed by foods, while organic meat, eggs, fruits, legumes/nuts, and
women usually experience financial constraints and many milk products were the least consumed. Furthermore,
responsibilities; hence they are motivated to expand their April, June, and July were generally the food shortage
Batame BMC Public Health (2024) 24:985 Page 17 of 18

Data availability
or unavailability months. Moreover, the majority of the For data protection purposes, the datasets that contain the personal
male and female-headed households reported having information of the participants cannot be shared publicly. Other non-
obtained their food from both the market and self-pro- traceable data is available with the corresponding author based on a
reasonable request.
duction. In addition, the study discovered that more of
the male-headed households had increased food produc-
Declarations
tion status, high dietary diversity, and were food secure
than female-headed households. Again, the study iden- Ethical approval and Informed consent
tified several themes, such as climate change/variabil- Approved by the Geo-Information Sciences (GEO) Ethics Committee at the
University of Twente (REF- 220158). All methods were carried out following
ity, poverty, large household size, unavailable lands, and relevant guidelines and regulations in the Declaration of Helsinki for research
gender stereotypes/ beliefs that emerged as influential involving human participants, human material, or human data. All participants
in promoting low dietary diversity and food insecurity in the study voluntarily agreed to be interviewed. The data used in this study
was anonymized before its use. Informed consent was obtained from all the
among male and female-headed households. Sex, total participants t​ o use their information for educational purposes, such as journal
household income, and cropland conversion were the publications.
variables influencing household heads’ perception of the
Consent for publication
impact of cocoa expansion on crop diversification. The Not applicable.
study concludes that male-headed households are more
food secure and food diverse than female-headed house- Conflict of interests
I hereby declare that I am not associated with or otherwise involved in any
holds in the study area. Policymakers can use the find- organization or other body that has a financial or non-financial stake in the
ings to determine appropriate public policies in the battle topics or materials discussed in this manuscript.
against low dietary diversity and food insecurity in rural
Author details
communities. Thus, the study recommends the following 1
Warnell School of Forestry and Natural Resources, University of Georgia,
interventions to improve household dietary diversity and Athens GA 30602, USA
2
food security: implement nutrition education programs Department of Natural Resources Management, Faculty of
Geoinformation Science and Earth Observation, ITC, University of Twente,
to raise awareness about the need for diverse diets and Enschede, The Netherlands
to provide practical information on how to incorporate
a greater variety of food groups into their meals; putting Received: 7 July 2023 / Accepted: 24 February 2024
in place gender-sensitive programs to empower women
in cocoa producing communities; giving credit assistance
to cocoa farming households, particularly those headed
by women to address poverty; and encourage crop diver- References
1. Ochieng J, et al. Determinants of dietary diversity and the potential
sification. Future research could investigate how gen- role of men in improving household nutrition in Tanzania. PLoS ONE.
der mainstreaming policies in agriculture have helped 2017;12(12):e0189022.
empower and improve the food security of female- 2. Cordero-Ahiman OV, et al. Factors that determine the dietary diversity score
in rural households: the case of the Paute River Basin of Azuay Province,
headed households in Ghana. Ecuador. Int J Environ Res Public Health. 2021;18(4):2059.
3. Cordero Ahimán OV, Santellano E, Estrada, Garrido A, Colmenero. Dietary
Supplementary Information diversity in rural households: the case of indigenous communities in Sierra
The online version contains supplementary material available at https://doi. Tarahumara, Mexico. J Food Nutr Res. 2017;5(2):86–94.
org/10.1186/s12889-024-18204-7. 4. Mallikarjuna K. Food security and climate change. Int J Res Appl Nat Social
Sci. 2013;1(1):45–52.
5. Frempong RB, Annim SK. Dietary diversity and child malnutrition in Ghana.
Questionnaires Heliyon. 2017;3(5):e00298.
6. Asiedu B, Nunoo F, Iddrisu S. Prospects and sustainability of aquaculture
development in Ghana, West Africa. Volume 3. Cogent Food & Agriculture;
Acknowledgements
2017. p. 1349531. 1.
I am thankful to the cocoa farmers in the Juaboso-Bia cocoa landscape
7. Kenkhuis M. Nutritional status among cocoa farming families and underlying
for giving me information on their food and nutrition security, and the
causes in Ghana Student Nutrition and Health at Wageningen University,
authorization to publish the findings of this study. I also appreciate the
intern at GAIN. Global Alliance for Improved Nutrition; 2016.
field facilitators’ unwavering assistance during the fieldwork. My heartfelt
8. Ajagun EO et al. Cocoa eats the food: expansion of cocoa into food croplands in
appreciation goes to the editor and the anonymous reviewers for their helpful
the Juabeso District, Ghana Food Security, 2021: p. 1–20.
reviews and comments.
9. Ashiagbor G, et al. Pixel-based and object-oriented approaches in segregat-
ing cocoa from forest in the Juabeso-Bia landscape of Ghana. Remote Sens
Author contributions
Applications: Soc Environ. 2020;19:100349.
The author developed the study, collected the data, analyzed the data, and
10. Batame M. Global Trade and Local Food Security: Mapping and Monitoring
drafted the manuscript.
Cocoa Expansion and its impact on Household Food Security in the Bia West
Disctrict, Ghana. University of Twente; 2023.
Funding
11. Chegere MJ, Stage J. Agricultural production diversity, dietary diversity
Not applicable.
and nutritional status: panel data evidence from Tanzania. World Dev.
2020;129:104856.
12. Heim A, Paksi A. Low dietary diversity and its influencing factors among a San
group in Namibia. BMC Res Notes. 2019;12(1):1–7.
Batame BMC Public Health (2024) 24:985 Page 18 of 18

13. Oldewage-Theron W, Kruger R. Dietary diversity and adequacy of women 31. Meinzen-Dick RS et al. Gender, assets, and agricultural development programs: A
caregivers in a peri-urban informal settlement in South Africa. Nutrition. conceptual framework CAPRi Working Paper, 2011.
2011;27(4):420–7. 32. Kanter R, et al. A conceptual framework for understanding the impacts of
14. Kadiyala S, Rawat R. Food access and diet quality independently predict agriculture and food system policies on nutrition and health. Food Secur.
nutritional status among people living with HIV in Uganda. Public Health 2015;7:767–77.
Nutr. 2013;16(1):164–70. 33. Pieters H, Guariso A, Vandeplas A. Concept Framew Anal Determinants food
15. Isabirye N, et al. Dietary diversity and associated factors among adolescents Nutr Secur. 2013.
in eastern Uganda: a cross-sectional study. BMC Public Health. 2020;20(1):1–8. 34. Bank G. SECTOR INDUSTRY ANALYSIS 2022 COCOA SECTOR REPORT 2022.
16. Kennedy G, Ballard T, Dop MC. Guidelines for measuring household and 35. Ghana Statistical S. 2010 Population and Housing Census; Juaboso District
individual dietary diversity. Food and Agriculture Organization of the United Analytical Report. 2014.
Nations; 2011. 36. Ashiagbor G, et al. Monitoring cocoa-driven deforestation: the contexts of
17. Taruvinga A, Muchenje V, Mushunje A. Determinants of rural household encroachment and land use policy implications for deforestation free cocoa
dietary diversity: the case of Amatole and Nyandeni districts, South Africa. Int supply chains in Ghana. Appl Geogr. 2022;147:102788.
J Dev Sustain. 2013;2(4):2233–47. 37. Wheeldon J. Mapping mixed methods research: methods, measures, and
18. Powell B, et al. The determinants of dietary diversity and nutrition: ethnonu- meaning. J Mixed Methods Res. 2010;4(2):87–102.
trition knowledge of local people in the East Usambara Mountains, Tanzania. 38. Teye JK. Benefits, challenges, and dynamism of positionalities associated with
J Ethnobiol Ethnomed. 2017;13(1):1–12. mixed methods research in developing countries: evidence from Ghana. J
19. Idowu O, Olusayo AO. Dietary diversity status of rural households in Nigeria: A Mixed Methods Res. 2012;6(4):379–91.
gendered perspective Dietary diversity status of rural households in Nigeria: a 39. Miller RL, Brewer JD. The AZ of social research: a dictionary of key social sci-
gendered perspective, 2019: p. 613–636. ence research concepts. Sage; 2003.
20. Koppmair S, Kassie M, Qaim M. Farm production, market access and dietary 40. Sultana S, et al. Prevalence and factors associated with depression among the
diversity in Malawi. Public Health Nutr. 2017;20(2):325–35. mothers of school-going children in Dhaka city, Bangladesh: a multi stage
21. Dei Antwi K, Lyford CP, Nartey RY. Analysis of food security among cocoa sampling-based study. Heliyon. 2021;7(7):e07493.
producing households in Ghana. J Agric Sustain, 2018. 11(2). 41. Taylor B. Assessing the food security status and production constraints of cocoa
22. Mehraban N, Ickowitz A. Dietary diversity of rural Indonesian households farming households in the Ashanti region. 2017, Thesis of master of philosophy,
declines over time with agricultural production diversity even as incomes University of Ghana, Legon, Ghana.
rise. Global Food Secur. 2021;28:100502. 42. Ajagun EO, et al. Cocoa eats the food: expansion of cocoa into food crop-
23. Codjoe SNA, Okutu D, Abu M. Urban household characteristics and dietary lands in the Juabeso District, Ghana. Food Secur. 2022;14(2):451–70.
diversity: an analysis of food security in Accra, Ghana. FoodNutr Bull. 43. Aneani F, et al. Adoption of some cocoa production technologies by cocoa
2016;37(2):202–18. farmers in Ghana. Sustainable Agric Res. 2012;1(1):103.
24. Kumar I, Gautam M. Determinants of Dietary Diversity score for the rural 44. Limb M, Dwyer C. Qualitative methodologies for geographers: issues and
households of Uttar Prades State. Int J Food Nutr Diet. 2022;10(1):9–16. debates. Oxford University Press; 2001.
25. Aidoo R, Mensah JO, Tuffour T. Determinants of household food security in the 45. Al-Zabir A, et al. Food security status of farming households in Bangladesh: a
Sekyere-Afram Plains district of Ghana. Eur Sci J, 2013. 9(21). comparison of recipients and non-receivers of institutional support. Afr J Sci
26. Annim SK, Frempong RB. Effects of access to credit and income on dietary Technol Innov Dev. 2021;13(4):449–62.
diversity in Ghana. Food Secur. 2018;10:1649–63. 46. Legwegoh AF, Hovorka AJ. Assessing food insecurity in Botswana: the case of
27. Kiewisch E. Looking within the household: a study on gender, food Gaborone. Dev Pract. 2013;23(3):346–58.
security, and resilience in cocoa-growing communities. Gend Dev. 47. Hernandez A. Why Are Fruits & Vegetables Important? 2018 [cited 2023 27– 02];
2015;23(3):497–513. Available from: https://healthyeating.sfgate.com/fruits-vegetables-impor-
28. Jebessa GM, Sima AD, Wondimagegnehu BA. Determinants of household tant-4703.html.
dietary diversity in yayu biosphere reserve, Southwest Ethiopia. Ethiop J Sci 48. Anderman TL, et al. Synergies and tradeoffs between cash crop production
Technol. 2019;12(1):45–68. and food security: a case study in rural Ghana. Food Secur. 2014;6:541–54.
29. Mekuria G, Wubneh Y, Tewabe T. Household dietary diversity and associated
factors among residents of finote selam town, North West Ethiopia: a cross
sectional study. BMC Nutr. 2017;3:1–6.
30. Dube L, Guveya E. Factors influencing smallholder crop diversification: a case Publisher’s Note
study of Manicaland and Masvingo provinces in Zimbabwe. Int J Reg Dev. Springer Nature remains neutral with regard to jurisdictional claims in
2016;3(2):1–25. published maps and institutional affiliations.