Systematic review of ambient temperature exposure during pregnancy and stillbirth: Methods and evidence

Associations between ambient temperature exposure during pregnancy and stillbirth have been reviewed and described in the literature. However, there is no existing review of environmental and epidemiologic methods applied to measure stillbirths resulting from exposure to ambient temperatures during pregnancy. The objective of this study is to systematically review published methods, data sources, and data linkage practices to characterize associations between ambient temperature and stillbirth to inform stillbirth prevention and risk management strategies.

A systematic review of published studies that assess the association between ambient temperature exposure during pregnancy using any measures or approach and stillbirth was undertaken in Cochrane Library, PubMed, Medline, Scopus, Embase, and Web of Science of studies (2000-2020, inclusive). Selection of studies were assessed by pre-specified eligibility criteria and documented using PRISMA. Citations were managed using EndNote X8 whilst selection, reviewing, and data extraction were performed using Covidence. The screening, selection, and data extraction process consisted of two blind, independent reviews followed by a tertiary independent review. An adapted Critical Appraisal Skills Program (CASP) checklist was used to assess quality and bias. The main findings and characteristics of all studies was extracted and summarized. Where appropriate, a meta-analysis will be performed for measures of association.

Among 538 original records, 12 eligible articles were identified that analysed associations between ambient temperature exposure and stillbirth for 42,848 stillbirths among 3.4 million births across seven countries. Varied definitions of stillbirth were reported based on gestational age, birthweight, both, or neither. The overall rate of stillbirth ranged from 1.9 to 38.4 per 1000 among six high-income countries and one low-middle-income country. All study designs were retrospective and included ten cohort studies, three case-crossover studies, and two additional case-control subgroup analysis. Exposure data for ambient temperature was mostly derived from standard municipal or country-level monitors based on weather stations (66.6%) or a forecasting model (16.7%); otherwise, not reported (16.7%). Results were not statistically pooled for a meta-analysis due to heterogeneity of methods and models among included studies. All studies reported associations of increased risk of stillbirth with ambient temperature exposures throughout pregnancy, particularly in late pregnancy. One study estimates 17-19% (PAR) of stillbirths are potentially attributable to chronic exposure to hot and cold ambient temperatures during pregnancy. Overall, risk of stillbirth was observed to increase below 15 °C and above 23.4 °C, where highest risk is above 29.4 °C.

Exposure to hot and cold temperatures during pregnancy may increase the risk of stillbirth, although a clear causative mechanism remains unknown. Despite lack of causal evidence, existing evidence across diverse settings observed similar effects of increased risk of stillbirth using a variety of statistical and methodological approaches for exposure assessments, exposure windows, and data linkage. Managing exposure to ambient temperatures during pregnancy could potentially decrease risk of stillbirth, particularly among women in low-resource settings where access to safe antenatal and obstetric care is challenging. To fully understand the effects or dose-response relationship of maternal exposure to ambient temperatures and stillbirth, future studies should focus on biological mechanisms and contributing factors in addition to improving measurement of ambient temperature exposure.