The coexistence of stunting and overweight or obesity in Ethiopian children: prevalence, trends and associated factors

**Background:** The double burden of childhood malnutrition, where undernutrition (stunting) coexists with overweight/obesity within the same individual, is an understudied phenomenon in low-income settings. In Ethiopia, stunting has declined from 51% in 2005 to 37% in 2019, while overweight is emerging as a significant issue. This study aimed to assess the prevalence, trends, and factors associated with concurrent stunting and overweight/obesity (CSO) among children aged 0–59 months in Ethiopia using pooled nationally representative data.

**Methods:** The study used pooled data from three consecutive Ethiopian Demographic and Health Surveys (EDHS) conducted in 2005, 2011, and 2016. A weighted sample of 23,756 children aged 0–59 months was included. Stunting was defined as height-for-age Z-score (HAZ) below −2 SD, and overweight/obesity as weight-for-height Z-score (WHZ) above +2 SD using WHO 2006 growth standards. CSO was defined as a child meeting both criteria simultaneously. Multilevel logistic regression models accounting for sampling weights and clustering were used to identify individual- and community-level factors associated with CSO. Four models were fitted: an empty model, individual-level variables, community-level variables, and a combined model. Model fit was assessed using deviance information criterion (DIC), AIC, and BIC.

**Key Results:** The overall prevalence of stunting was 43.12% [95% CI: 42.50–43.75%], overweight/obesity was 2.62% [95% CI: 2.42–2.83%], and CSO was 1.33% [95% CI: 1.18–1.48%]. The prevalence of CSO declined from 2.36% [95% CI: 1.94–2.85%] in 2005 to 0.87% [95% CI: 0.7–1.07%] in 2011, then increased slightly to 1.34% [95% CI: 1.13–1.59%] in 2016. Stunting prevalence declined consistently from 50.8% in 2005 to 38.4% in 2016. Overweight/obesity decreased from 4.2% to 2.81% over the same period. CSO was more common in rural areas than urban areas across all survey years (2.38% vs 2.06% in 2005; 0.88% vs 0.71% in 2011; 1.39% vs 0.89% in 2016) and more common in males than females. In the final multilevel model (Model IV), factors significantly associated with higher odds of CSO included: currently breastfeeding [AOR: 1.64, 95% CI: 1.01–2.72], maternal overweight/obesity (BMI ≥ 25 kg/m²) [AOR: 2.65, 95% CI: 1.19–5.88], household size of 1–4 members [AOR: 1.52, 95% CI: 1.02–2.26], and being surveyed in EDHS-2005 compared to EDHS-2016 [AOR: 4.38, 95% CI: 2.42–7.95]. The intraclass correlation coefficient (ICC) in the final model was 7.99%, indicating significant community-level variation.

**Clinical Implications:** The study found that less than 2% of Ethiopian children had CSO, which is lower than reports from some other low-income settings but shows a rising trend between 2011 and 2016. The high stunting prevalence (43.12%) remains a critical public health concern, exceeding the WHO threshold of ≥30% for very high prevalence. The association between maternal overweight/obesity and CSO suggests that biological, behavioral, and nutritional transition factors are at play, where children in households with better food supply may experience excessive energy intake while still suffering from stunting due to poor nutrient quality. The finding that breastfeeding was associated with higher CSO odds warrants further investigation, possibly related to maternal nutritional status affecting breast milk quality in food-insecure settings. The authors recommend targeted interventions for early identification and management of at-risk children, particularly those born to overweight mothers and those in smaller households. They emphasize the need for strong national nutrition policies addressing both undernutrition and overnutrition simultaneously to achieve Sustainable Development Goals (Goal 2 and Target 3.4). Study limitations include potential recall bias for birth size and infection history, the cross-sectional design precluding causal inference, and the inability to account for factors like food security and pregnancy nutrition due to secondary data constraints.