Introduction: Conflict exacerbates poor complementary feeding and reduces dietary diversity. Before the 2020–2022 war in northern Ethiopia, around 74% of children aged 6–23 months failed to meet minimum dietary diversity (MDD) and post-war prevalence was unknown. This study aims to assess MDD prevalence and associated factors among children aged 6–23 months in a town in northern Ethiopia two years after the ceasefire.

Methodology: A health facility-based cross-sectional study of 584 participants was conducted in a town in northern Ethiopia. Sociodemographic and dietary data were collected using a 24-hour dietary recall questionnaire and analyzed in STATA ® version 15. Pre-war dietary diversity was estimated using data from the 2016 and 2019 Ethiopian Demographic and Health Surveys.

Results: MDD declined from 33.2% pre-war period to 25.2% (95% CI: 21.6-28.7) in the post-war period. Children aged 18–23 months were 3.2 times more likely to achieve MDD than those aged 6–11 months (p = 0.001). Middle- and high-income households had 6.13-fold and 13.58-fold higher odds of meeting MDD, respectively (both p < 0.001). Households with 5–8 members had 3.6-fold higher odds of providing MDD compared to those with 1–3 members (p = 0.017). Children of mothers aged 26–35 years (AOR = 0.48, p = 0.026) and 36–42 years (AOR = 0.29, p = 0.017) had lower odds of meeting MDD. Higher paternal education (AOR = 2.58, p = 0.031) and paternal occupation as a merchant (AOR = 3.17, p = 0.001) were positively associated with MDD. Post-war, grain (77.5% to 92.1%) and legume (33.9% to 69.3%) consumption increased significantly, while flesh foods (20.9% to 7.2%) and vitamin A-rich foods 46.8% to 26%) declined.

Conclusion: Post-war dietary diversity in remains low, influenced by socioeconomic and demographic factors. Significant reductions in nutrient-dense foods highlight critical gaps in child nutrition. Targeted nutrition education for caregivers and interventions promoting dietary diversity are essential in conflict-affected settings.

Study Design and Population

This study employed a facility-based cross-sectional design, conducted in late 2024, in a town in northern Ethiopia. The study population comprised mothers/caregivers with children aged 6–23 months who attended Extended Program on Immunization (EPI) services at one of three randomly selected health centers. Participants were recruited through random sampling. Data on sociodemographic characteristics and dietary intake were collected via face-to-face interviews using a pretested and standardized questionnaire.

Study Setting

This study was conducted in a town in northern Ethiopia two years after the end of the war in northern Ethiopia (2020-2022). The town, located in nothern Ethiopia, has an estimated population of 600,000 as of 2024. It is served by several hospitals and health centers.

Study variables

The outcome variable was Minimum Dietary Diversity (MDD), defined as the consumption of foods and beverages from at least five of eight defined food groups within a 24-hour period, according to the updated WHO/UNICEF definition ((UNICEF) 2021). The eight food groups are: 1) breast (milk); 2) grains (grains, roots, tubers, and plantains); 3) legumes (beans, peas, lentils, nuts, and seeds); 4) dairy products (milk, infant formula, yogurt, cheese); 5) flesh foods (meat, fish, poultry, organ meats); 6) eggs; 7) vit-A (vitamin-A rich fruits and vegetables); and 8) other (fruits and vegetables) (Kennedy, Ballard et al. 2011, UNICEF 2022). Consumption of a specific food groups was coded as 1, and a sum of the eight columns “food-sum” was created by merging the eight columns. A second column “mdd” identified the child’s dietary diversity as “diverse=1” when food-sum >=5, or “not-diverse=0”, when food-sum <5.

Data Collection

Data were collected using a structured, validated questionnaires adapted from FAO and EDHS tools (Kennedy, Ballard et al. 2011, ICF 2016). The questionnaire was initially developed in English, translated into Tigrigna (the local language), and then back-translated to ensure consistency and accuracy. Face-to-face interviews were conducted with mothers/caregivers, focusing on dietary intake over the previous 24 hours and relevant sociodemographic factors. Dietary data were collected based on the updated WHO/UNICEF minimum dietary diversity guidelines. To minimize recall bias caregivers were prompted to recall specific foods and meals consumed by the child during the reference period. Community health workers, who had prior experience in interviewing, were trained and recruited to conduct the data collection. Daily supervision was implemented to ensure the accuracy and completeness of the data. While we did not conduct a calibration study, the UNICEF dietary diversity tool has been validated in similar settings, providing confidence in its reliability for assessing minimum dietary diversity.

Sampling technique

This study used multistage sampling technique to identify the participants of this study. In the first stage of sampling, three three administrative areas were randomly selected. Then during the second stage, one health facility was randomly selected from each area. In the third stage, children were selected from each health center using a non-random, first-come, first-served approach, where the eligible children arriving at the health center on the vaccination day of the week were recruited. The final sample was proportionally allocated based on the number of children visiting each facility. Children who came with non-consenting caregivers and children with severe chronic illnesses or congenital anomalies were excluded from the study.

Data Analysis

Data were entered into Microsoft Excel ®, cleaned, and then analyzed using STATA ® statistical software version 15. Missing data were imputed using median imputation for continuous variables and mode imputation for categorical variables. Descriptive statistics were used to summarize sociodemographic characteristics and dietary diversity. Categorical variables were presented as frequencies and percentages, while continuous variables were categorized based on common frameworks. Chi-square or Fisher’s exact tests were used to assess associations between categorical variables, and binary logistic regression analysis was performed to identify factors associated with MDD. To address potential confounding, we included key sociodemographic variables in the final multivariable logistic regression model. These variables were selected based on prior literature and theoretical relevance to dietary diversity.

Prewar estimation of MDD

The pre-war dietary diversity estimates (EDHS 2016 and EDHS 2019) were obtained from the Ethiopian Demographic and Health Survey (DHS) and are not included in this dataset due to licensing restrictions. Researchers can access these data by registering on the DHS website (https://dhsprogram.com) and requesting permission for use.

References

Kennedy, G., Ballard, T., & Dop, M. C. (2011). Guidelines for measuring household and individual dietary diversity. Food and Agriculture Organization of the United Nations.

ICF. 2016. Ethiopia Demographic and Health Survey 2016. Addis Ababa, Ethiopia, and Rockville, Maryland, USA: CSA and ICF

World Health Organization and the United Nations Children’s Fund (UNICEF). (2021). Indicators for assessing infant and young child feeding practices: definitions and measurement methods.

UNICEF. (2022). United Nations Children’s Fund, Division of Data, Analysis, Planning and Monitoring (2022). Global UNICEF Global Databases: Infant and Young Child Feeding: Egg and/or flesh food consumption, Minimum dietary diversity, Minimum meal frequency, Minimum acceptable diet