High local temperatures reduce dietary diversity in children under 2: variation along the urban spectrum in Kenya and Burkina Faso.

Measuring the impact of climate change on food security and the implications for child nutrition remains a challenge. While urban areas typically have better nutritional indicators, substantial heterogeneity exists and is masked by the commonly used urban-rural binary. Linking four Demographic and Health Surveys from Burkina Faso (2010 and 2021), and Kenya (2014 and 2022) with a high frequency temperature dataset, we implement a cross-sectional secondary analysis to measure recent exposure to high temperatures (average maximum daily temperature as well as the number of days above 32 °C in the 14 and 30 days preceding each survey). We further incorporate a satellite-based degree of urbanization measure. We estimate logistic regression models, adjusting for demographic characteristics, to understand how temperature and urbanicity affect minimum dietary diversity (MDD) and wasting in children 6-23 months. In all four surveys, as local average maximum temperature over the preceding 30 days increased, the odds of attaining MDD reduced 6% in Burkina Faso (2010) (Odds Ratio = 0.94; 95% confidence interval 0.89, 0.99), 9% in Burkina Faso (2021) (OR = 0.91; 95% CI 0.85, 0.99), 10% in Kenya (2014) (OR = 0.90; 95% CI 0.88, 0.93) and 4% in Kenya (2022) (OR = 0.96; 95% CI 0.92, 0.99). This was driven mainly by children in rural areas. Programs to prevent malnutrition will need to be geographically targeted and tailored to local climatic and urban contexts. Dietary diversity is an easy to collect, standardized measure that may serve as an early warning indicator of acute malnutrition and key time point for intervention.