Climate, maternal mortality, and stillbirth in Senegal: Associations not identified from spatiotemporal modeling with routinely collected national health data

Abstract

Pregnant women and fetuses are vulnerable to high temperatures. Few studies of climate impacts on obstetric outcomes have been conducted in places with ongoing high heat. We tested whether we could detect associations between rates of maternal mortality, stillbirth, temperature, and precipitation in Senegal using routinely collected, aggregated monthly health data and remotely sensed climate data (2018–2022). We assigned Senegalese health districts to three climate zones based on average temperature and annual precipitation. We used spatiotemporal modeling of records from 2,097,903 births to assess associations between rates of maternal mortality, stillbirth, temperature, and precipitation by climate zone, testing lag options and controlling for health district (as a random effect), year, and temperature-precipitation interaction. Despite modeling efforts across three climate zones, no models identified significant associations between same-month temperature and either stillbirth or maternal mortality. One of six models of precipitation and stillbirth and maternal death identified a statistically significant association between precipitation and stillbirth 3 months later. In this context characterized by consistent high temperatures, we did not find evidence of an association between heat exposure in late pregnancy and population-level changes in stillbirth or maternal mortality. This negative finding differs from reports from other contexts. It may reflect low temperature variation, health data quality, sample size, or the imprecision of monthly averages. Exposure to ongoing high heat may prompt physiological and behavioral adaptation, possibilities that merit future investigation.