The relationship between prenatal heat exposure and birth outcomes: How much does the heat metric matter?

The impact of prenatal heat exposure on birth outcomes is well-established, but what is it about heat drives this relationship? Is it exposure to extreme temperatures, to moderate heat, or the confluence of heat and humidity? Despite the large body of research on heat exposure and birth outcomes, the literature lacks consistent measurement. This means we cannot extract practical recommendations around which heat conditions pose the greatest risk, and hence should be avoided during pregnancy. It also means we cannot predict the implications of climate change on neonatal health and healthcare needs at a population level. This paper has two goals: first, to demonstrate that our conclusions around the existence and magnitude of the impact of heat exposure vary dramatically with the choice of heat exposure metric, and second, to make general recommendations for how heat exposure should be measured in future. We present analysis from Australia's Northern Territory - a region spanning tropical and arid climates. We compare commonly used heat exposure metrics, alongside additional metrics supported by theory. We find that a metric based on 'bands' of exposure and incorporating daily minimum as well as maximum measures provides the best fit; this is consistent with our theoretical understanding that both moderate and extreme heat affect fetal development in different ways. Estimates based on our preferred heat metric suggest that the impact of prenatal heat exposure on preterm birth is orders of magnitude larger than what would be implied by some metrics commonly used in the literature. Our findings underscore the importance of getting the measure of heat right, particularly in tropical climates.