Effect of heat stress in the first 1000 days of life on fetal and infant growth: a secondary analysis of the ENID randomised controlled trial

Abstract

Background: The intersecting crises of climate change, food insecurity, and undernutrition disproportionately affect children. Understanding the effect of heat on growth from conception to 2 years of age is important because of mortality and morbidity implications in the near term and over the life course.

Methods: In this secondary analysis, we used longitudinal pregnancy cohort data from the Early Nutrition and Immunity Development (ENID) randomised controlled trial in West Kiang, The Gambia, which occurred between Jan 20, 2010, and Feb 10, 2015. The ENID trial assessed micronutrient supplementation in the first 1000 days of life starting from 20 weeks' gestation, during which anthropometric measurements were collected prospectively. We used multivariable linear regression to assess the effect of heat stress (defined by Universal Thermal Climate Index [UTCI]) on intrauterine growth restriction based on length-for-gestational age Z score (LGAZ), weight-for-gestational age Z score (WGAZ), and head circumference-for-gestational age Z score (HCGAZ) at birth, and assessed for effect modification of supplement intervention on the relationship between heat stress and infant anthropometry. We used multivariable, multilevel linear regression to evaluate the effect of heat stress on infant growth postnatally based on weight-for-height Z score (WHZ), weight-for-age Z score (WAZ), and height-for-age Z score (HAZ) from 0 to 2 years of age.

Findings: Complete data were available for 668 livebirth outcomes (329 [49%] female infants and 339 [51%] male infants). With each 1°C increase in mean daily maximum UTCI exposure, in the first trimester, we observed a reduction in WGAZ (-0·04 [95% CI -0·09 to 0·00]), whereas in the third trimester, we observed an increase in HCGAZ (0·06 [95% CI 0·00 to 0·12]), although 95% CIs included 0. Maternal protein-energy supplementation in the third trimester was associated with reduced WGAZ (-0·16 [-0·30 to -0·02]) with each 1°C increase in mean daily maximum UTCI exposure, while no effect of heat stress on WGAZ was found with either standard care (iron and folate) or multiple micronutrient supplementation. For the postnatal analysis, complete anthropometric data at 2 years were available for 645 infants (316 [49%] female infants and 329 [51%] male infants). Postnatally, heat stress effect varied by infant age, with infants aged 6-18 months being the most affected. In infants aged 12 months exposed to a mean daily UTCI of 30°C (preceding 90-day period) versus 25°C UTCI, we observed reductions in mean WHZ (-0·43 [95% CI -0·57 to -0·29]) and mean WAZ (-0·35 [95% CI -0·45 to -0·26]). We observed a marginal increase in HAZ with increasing heat stress exposure at age 6 months, but no effect at older ages.

Interpretation: Our results suggest that heat stress impacts prenatal and postnatal growth up to 2 years of age but sensitivity might vary by age. In the context of a rapidly warming planet, these findings could have short-term and long-term health effects for the individual, and immediate and future implications for public child health.

Funding: Wellcome Trust.

Figure 1

Schematic of potential biological pathways of heat stress effects on growth in-utero (A) and in infancy (B)

Figure 2

Flowchart of populations included in analysis Underweight: <–2 weight-for-age Z score; stunted: <–2 height-for-age Z score; wasted: <–2 weight-for-height Z score. *Individuals could be included in more than one group.

Figure 3

Geographic region (A), heat stress exposure over the course of the study (B), and seasonal trends in WHZ and WAZ (C) In panel A, black dots represent village locations. Map data ©2021 Google. In panel B, each line is an individual village exposure. UTCI=Universal Thermal Climate Index. WAZ=weight-for-age Z score. WHZ=weight-for-height Z score.

Figure 4

Association between infant anthropometry and UTCI (A) Association between WHZ and mean daily UTCI for the preceding 90-day period, adjusted for season, birthweight, gestational age at birth, infant sex, maternal BMI, and maternal age. (B) Association between WAZ and 90-day mean daily UTCI, adjusted for season, birthweight, infant sex, and parity. (C) Association between HAZ and mean daily lifetime UTCI, adjusted for season, birthweight, gestational age at birth, maternal height, infant sex, and parity. The relationship between heat and growth varied by age and therefore age intervals are presented separately. The solid line represents the mean Z score estimates and the shading represents 95% CIs. The vertical dashed line is median exposure for the cohort. HAZ=height-for-age Z score. UTCI=Universal Thermal Climate Index. WAZ=weight-for-age Z score. WHZ=weight-for-height Z score.