Linking prenatal exposure to temperature intensity and variability to child mortality: evidence from 31 low-and...

High child mortality in low-and middle-income countries is compounded by rising temperatures and climate variability. We conducted a retrospective cohort study of singleton live births from DHS birth histories (IPUMS-DHS) linked to gridded climate reanalysis across 31 countries in sub-Saharan Africa, South Asia, and the Middle East and North Africa. The analytical sample comprised births to mothers aged 15-49 years from 1990 to 2018. Using survey-weighted multilevel logistic regression with cluster-level random intercepts, we examined associations between prenatal exposure to average maximum temperature (°C) and its variability with neonatal, infant, and under-5 mortality. Each additional degree of cumulative exposure was associated with 2.7% higher under-5 mortality odds (OR = 1.027; 95% CI: 1.023-1.030; p < 0.001). A dose-response pattern emerged, with the highest temperature quintile associated with 38.3% higher under-5 odds (OR = 1.383; 95% CI: 1.330-1.437; p < 0.001). Infant mortality showed similar trimester-specific associations (from OR = 1.008; 95% CI: 1.005-1.011; p < 0.001 to OR: 1.013; 95% CI: 1.006-1.012; p < 0.001), and neonatal mortality smaller but significant ones. Variability showed divergent associations: positive deviations (>0.5σ) with higher under-5 mortality (OR = 1.033; 95% CI: 1.016-1.051; p < 0.001) but lower neonatal odds (OR = 0.966; 95% CI: 0.941-0.991; p < 0.01). Prenatal heat was consistently associated with higher child mortality, with the first and third trimesters as critical windows and risk intensifying at upper quintiles. These findings underscore the limitations of mean-temperature metrics and support embedding heat-alert counseling and affordable cooling within maternal and child health services.