**Background:** Ethiopia faces high malnutrition rates, with 38% of children under 5 stunted, 10% wasted, and 24% underweight. The country ranks 90th out of 116 countries on the 2021 Global Hunger Index. Ethiopian diets are particularly low in chicken, fruits, vegetables, and red meat, contributing to high burdens of micronutrient deficiencies. While previous studies examined dietary diversity determinants, they were limited in geographical scope. This study aimed to identify dominant food groups consumed and investigate determinants of household dietary diversity across all regions of Ethiopia using nationally representative data.
**Methods:** Data came from the 4th wave (2018/2019) of the Ethiopian Socioeconomic Survey (ESS), collected by the Central Statistics Agency of Ethiopia in collaboration with the World Bank. The survey included 7,527 households from 565 Enumeration Areas selected through multi-stage sampling. This study focused on 3,115 rural households. Household Dietary Diversity Score (HDDS) was calculated using 10 food groups consumed over the past 7 days, categorized as low (≤3 food groups), medium (4-6), or high (≥7) per FAO recommendations. Independent variables included household size, sex, education, marital status, religion of household head, PSNP participation, location, and wealth index (constructed using Principal Component Analysis). Ordinal logistic regression was employed, with the Brant test confirming the proportional odds assumption. Variance Inflation Factor (VIF) testing showed no multicollinearity problems.
**Key Results:** The mean household size was 5 (SD 2.29), mean age of household head was 44.4 years (SD 15.4), 73.5% were male-headed, and 65.55% of household heads were illiterate. Cereals were the most dominant food group consumed by 96.4% of households, followed by pulses (82%). Nutrition-dense foods like lean meat, vegetables, and fruits were the least consumed. Overall, 26% of households fell into low dietary diversity, 65% into medium, and only 9% into high dietary diversity. The mean number of food groups consumed was 5 (range 2-10). Female-headed households had 38% higher odds of consuming diverse foods compared to male-headed households (AOR = 1.38, 95% CI: 1.10-1.73, p<0.01). Household heads with secondary education and above had 62% higher odds (AOR = 1.62, 95% CI: 1.2-2.30, p<0.01), and those with primary education had 35% higher odds (AOR = 1.35, 95% CI: 1.11-1.64, p<0.01) compared to illiterate heads. Single household heads had 37% lower odds compared to married heads (AOR = 0.63, 95% CI: 0.50-0.80, p<0.01). Households with 6+ members had 43% higher odds (AOR = 1.43, 95% CI: 1.17-1.75, p<0.01) and those with 4-5 members had 33% higher odds (AOR = 1.33, 95% CI: 1.09-1.60, p<0.01) compared to households with 1-3 members. Regional differences were substantial: households in Harari and Dire Dawa had 6.56 times higher odds (AOR = 6.56, 95% CI: 4.60-9.37, p<0.01) compared to Tigray and Amhara. Wealth showed a strong gradient: the richest households had 9 times higher odds (AOR = 9.01, 95% CI: 6.79-11.98, p<0.01), richer had 4.77 times (AOR = 4.77, 95% CI: 3.65-6.21), middle had 3.29 times (AOR = 3.29, 95% CI: 2.54-4.26), and poorer had 2.06 times (AOR = 2.06, 95% CI: 1.62-2.62) compared to the poorest. PSNP participation showed no significant effect on dietary diversity.
**Clinical Implications:** The study reveals that the vast majority (91%) of rural Ethiopian households have inadequate dietary diversity, with diets heavily reliant on cereals and pulses while lacking nutrient-dense foods like meat, vegetables, and fruits. This dietary pattern likely contributes to Ethiopia's high burden of micronutrient deficiencies and child malnutrition. The strong wealth gradient in dietary diversity (9-fold difference between richest and poorest) indicates that economic access is a primary barrier. The regional clustering of dietary patterns (similarities between Tigray and Amhara; Somali and Afar; Oromia and SNNP; Benshangul Gumuz and Gambella; Harari and Dire Dawa) suggests that culturally-tailored, region-specific nutrition interventions are needed rather than uniform national programs. The finding that female-headed households have better dietary diversity supports investing in women's empowerment and nutrition education. The lack of significant effect from PSNP participation on dietary diversity raises questions about whether safety net programs adequately address dietary quality or only caloric sufficiency.