Maternal acute thermophysiological stress and stillbirth in Western Australia, 2000–2015: A space-time-stratified case-crossover analysis

Abstract

Background: The extreme thermal environment driven by climate change disrupts thermoregulation in pregnant women and may threaten the survival of the developing fetus. Objectives: To investigate the acute effect of maternal exposure to thermophysiological stress (measured with Universal Thermal Climate Index, UTCI) on the risk of stillbirth and modification of this effect by sociodemographic disparities. Methods: We conducted a space-time-stratified case-crossover analysis of daily UTCI and 2835 singleton stillbirths be- tween 1st January 2000 and 31st December 2015 across multiple small areas in Western Australia. Distributed lag non- linear models were combined with conditional quasi-Poisson regression to investigate the effects of the UTCI exposure from the preceding 6 days to the day of stillbirth. We also explored effect modification by fetal and maternal sociodemographic factors. Results: The median UTCI was 13.9 °C (representing no thermal stress) while the 1st and 99th percentiles were 0.7 °C (slight cold stress) and 31.7 °C (moderate heat stress), respectively. Relative to median UTCI, we found positive associations between acute maternal cold and heat stresses and higher risks of stillbirth, increasing with the intensity and duration of the thermal stress episodes. The cumulative risk from the preceding 6 days to the day of stillbirth was stronger in the 99th percentile (RR = 1.19, 95% CI: 1.17, 1.21) than the 1st percentile (RR = 1.14, 95% CI: 1.12, 1.15), relative to the median UTCI. The risks were disproportionately higher in term and male stillborn fetuses, smoking, un- married, ≤19 years old, non-Caucasian, and low socioeconomic status mothers. Discussion: Acute maternal exposure to both cold and heat stresses may contribute to the risk of stillbirth and be exacerbated by sociodemographic disparities. The findings suggest public health attention, especially for the identified higher-risk groups. Future studies should consider the use of a human thermophysiological index, rather than surrogates such as ambient temperature.