Ambient temperature extremes and neonatal mortality: A time-stratified case-crossover analysis using Demographic and Health Survey data from East Africa (2011-2022)

IntroductionDue to their underdeveloped thermoregulatory system, neonates are at increased risk of morbidity and mortality from hot and cold temperatures. Our study aimed to analyse the effects of environmental temperature on overall, very early, early and late neonatal acute mortality in 5 East African countries using the Demographic and Health Surveys (DHS) data. MethodsWe obtained neonatal mortality data from the DHS conducted between 2016 and 2022. Our outcomes were (i) overall neonatal mortality (days 0-27), (ii) very early (day 0); (iii) early (days 1-6); and (iv) late neonatal mortality (days 7-27). Daily mean temperature was constructed from ERA5-Land and assigned at household-level. We used a time-stratified case-crossover design with distributed lag non-linear models (0-6-day lag) to estimate odds of mortality with exposure to the 5th and 95th temperature percentiles (versus the median). Country-level estimates were generated and then pooled to assess the overall association. ResultsA total of 1373 neonatal deaths were included, over 80% of which occurred within the first 6 days. The association between ambient temperature and neonatal mortality was heterogenous. In pooled analysis, the 95th and 5th percentiles were associated with increased and decreased mortality odds, respectively, although estimates were imprecise. In Uganda, there was strong evidence of association between high ambient temperature (95th percentile) and overall neonatal mortality (OR=3.54; 95% CI: 1.73-7.28) as well as early neonatal mortality (OR=3.75; 1.70-8.28), while odds of very early neonatal mortality increased with exposure to low temperatures (5th percentile) (OR=5.65; 1.89-16.69). There was no strong evidence of association in other countries. ConclusionTemperature-related neonatal mortality risk differs across East African countries. Other factors may play a significant role. Future research should consider the effects of environmental temperature on neonatal mortality across different climate zones. What is already known on this topicO_LIHot and cold indoor environments are associated with a risk of neonatal mortality C_LIO_LINeonatal vulnerability to temperature extremes is most pronounced in the early neonatal period (days 0-6) compared with the late neonatal period (days 7-27) C_LI What this study addsO_LIThis study adds to the body of evidence on the effect of heat on neonatal mortality in different climate regions, allowing for between-country assessment of the association between temperature and neonatal mortality in East Africa. C_LIO_LIOur analysis provides country-level estimates for the association between temperature and overall neonatal mortality, as well as the three distinct phases of the neonatal period (very early, early and late) with differing mortality risk from temperature C_LI How this study might affect research, practice or policyO_LIThere is a need to incorporate heat mitigation strategies into the care for newborn policies and practices, especially for the very early and early neonatal periods C_LI