Assessing the impact of heat stress on growth faltering in the first 1000 days of life in rural Gambia

doi:10.21203/rs.3.rs-2358038/v1

Download PDF

Article

Assessing the impact of heat stress on growth faltering in the first 1000 days of life in rural Gambia

https://doi.org/10.21203/rs.3.rs-2358038/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The intersecting crises of climate change, crop failure, food security and under-nutrition are disproportionately impacting children living in the Global South. Understanding the relationship between heat stress exposure and child growth is needed considering current and projected increasing temperatures. We used multilevel, multivariate linear regression models of 60-day heat stress exposure on child growth. Heat stress was defined by Universal Thermal Climate Index (UTCI), and outcomes as: prenatal weight-for-age (WAZ); postnatally to 2 years weight-for-age (WAZ), weight-for-height (WHZ) and height-for-age (HAZ) z-scores, in The Gambia, West Africa.

Postnatal WAZ and WHZ reduced with increasing heat stress exposure. Mean UTCI exposure of 30°C versus 20°C was associated with 0.34 (95%CI -0.49;-0.20) reduction in WHZ between 0–2 years. In contrast, HAZ increased with increasing UTCI, to 29°C, beyond which HAZ plateaued/decreased. Our results suggest that rising global temperatures may impact child growth in vulnerable areas with long-term implications for morbidity and mortality.

Earth and environmental sciences/Environmental sciences/Environmental impact

Scientific community and society/Developing world

Undernutrition is estimated to be associated with 3 million child deaths per year worldwide.¹ West Africa experiences one of the highest rates of child undernutrition in the world with pooled rates in children under 5 years of age estimated at 31.8% stunted, 20.1% underweight and 10% wasted,² defined as less than 2 standard deviations of the international standard (stunting < -2 height-for-age z-score (HAZ); underweight < -2 weight-for-age z-score (WAZ); wasted < -2 weight-for-height z-score (WHZ)). Additional impacts of child undernutrition include a life course increased risk of chronic health conditions such as type 2 diabetes (associated with poor fetal growth), delays in neurocognitive development and reduced economic productivity in adulthood.³

Repeated anthropometric measurements – most commonly weight and length/height - are used to measure a child’s nutritional status and to identify failure to thrive - growth faltering. Growth faltering is known to be multifactorial and is influenced by genetic and epigenetic variants, food insecurity, poverty, poor sanitation, repeated infectious episodes, emotional and physical neglect, and environmental exposures.^4–6 However, recent studies have shown limited benefits for growth faltering from nutrition interventions and disappointingly no impact from sanitation interventions.^7,8

With anthropogenic climate change having already heated the planet by an average 1.2°C above pre-industrial levels, with even greater increases over the populated land masses, there is a growing interest in the impact of ambient temperature on both fetal and infant health.^9,10 The increasing exposure to extreme heat, which is disproportionately affecting those in extreme poverty,¹¹ and the current intersecting crises of climate change, crop failure, drought, and globally rising food costs, results in an urgent need to understand the health impact of heat in more detail.

A recent systematic review of heat and pregnancy outcomes, found 12 out of 19 studies described a reduction in birth weight with increasing temperature exposure, although the effects were small.¹² A cross-sectional study of growth faltering in African children found increased prevalence of wasted and underweight children in those exposed to temperatures above 35°C, but a reduction on the prevalence of stunted children.¹³ However the cross-sectional design made interpretation of these findings difficult, as growth faltering occurs over time.

The postulated mechanisms for in utero heat stress effects on birth weight include: variation in DNA methylation leading to changes in gene expression;¹⁴ change in placental blood flow due to shunting of blood to the skin for thermoregulation resulting in acute and/or chronic fetal stress;¹⁵ and reduction in maternoplacental nutrient delivery.^16,17 The possible biological mechanisms of heat stress affecting growth faltering in childhood include: reduction in food intake to reduce metabolic heat production and so aid thermoregulation (acute alteration); increased risk of acute episodes of diarrhoea;¹⁸ and alteration in surface area to mass ratio to improve heat loss – which would increase wasting but have little or no impact on stunting.¹⁹ Despite the biological plausibility of the impact of heat stress on growth faltering, and growing concern over the health impacts of climate change, there remains a paucity of evidence of the effect.²⁰

In this study we utilise longitudinal data from a randomised trial conducted in the first 1000 days of life – the time from conception to two years of age - where multiple repeated measures of individual pregnant women and their children have been taken prospectively. Using these data we examine the relationship between ambient heat stress exposure and growth faltering.

The in utero to birth dataset included 776 individuals with 2186 measurements (an average of 2.8 measurements per participant – including newborn anthropometry). Final numbers for the 0–2 years dataset included 754 individuals with 5415 measurements (an average of 7.2 measurements per participant).

Complete live birth outcome data was available for 670 participants, 329 females and 341 males, with a mean birth weight of 3.00kg (± 0.40) and 66 (9.9%) with LBW by definition of < 2.5kg. There were 220 (32.8%) infants born small for gestational age (defined as less than 10th percentile for standardised gestational age weight)^21,22. Complete data at 2 years of age was available for 645 participants, with 153 (23.7%) underweight, 80 (12.4%) wasted, 166 (25.7%) stunted and 35 (5.4%) with concurrent wasting and stunting. In Fig. 1 the flow-chart of numbers of participants at each stage of analysis is shown. Individual weight trajectories from 0–2 years and individual WAZ from 0–2 years are reported in Supplementary Materials (Fig S5).

Monthly average heat stress exposure over the study period ranged from 20.4°C to 31.5°C UTCI, with a mean and median value of 27.0°C and 28.0°C UTCI respectively.

Figure 2 demonstrates the seasonal nature of the heat stress exposure for the region (further detail can be visualised in the supplemental video Fig S6).

The in utero multilevel multivariable model demonstrated no clear association between WAZ and UTCI exposure at 30-, 60- or 90-day exposures (Fig. 3), where the 95% confidence intervals include zero at all points. All estimates from the model can be found in the supplement (Table S2) and estimates of WAZ and UTCI exposure are presented in Table 2.

In the infant models there were significant associations between all outcomes and heat stress exposure (Fig. 4 and Supplement Table 2). There was a significant reduction in WAZ associated with increasing mean UTCI exposure over all time periods. The estimates of effect at the 95th percentile (30°C UTCI) compared to 26°C UTCI for all outcomes are presented in Table 2. When 0–2 year olds were exposed to average heat stress temperatures of 30°C UTCI versus 26°C UTCI, WAZ reduced by 0.11 (95% CI -0.17;-0.04), WHZ reduced by 0.22 (95% CI -0.31;-0.14) and HAZ increased by 0.019 (95% CI -0.045;0.083).

Of the growth faltering metrics, WHZ reduced the most with increasing UTCI, including across all time summary values (see supplement for 30- & 90-day model outputs), with mean UTCI exposure of 30°C versus 20°C associated with a 0.34 (95%CI -0.49;-0.20) reduction in WHZ between 0–2 years.

HAZ and heat stress demonstrated a very different association compared with WAZ and WHZ. Increasing UTCI was associated with increasing HAZ up to a threshold of around 29°C UTCI, above which there is uncertainty in the relationship with confidence intervals crossing zero in 30- and 60-day summary exposures, but significantly decreasing when summarised by 90-day exposures (see supplement Fig S8).

To the best of our knowledge, this is the first study to utilise longitudinal data to assess the impact of heat on growth faltering in the first 1000 days of life. We found increasing heat stress exposure associated with a reduction in WAZ and WHZ in infants after controlling for maternal, child and environmental factors. WHZ was the most impacted measurement of growth faltering by heat stress. This is concerning as wasting (WHZ < -2) has both short and long-term health implications and is associated with increased mortality, reduced immunity and developmental delays.^23,24 Also, although wasting has decreased over the last two decades it remains higher than the SDG target of < 5% globally.²⁵ WAZ reduced with both low and high heat stress exposures (although at lower temperatures this failed to meet statistical significance as demonstrated by wide confidence intervals). Exposure of UTCI above 26°C, resulted in a reduction in WAZ.

In utero there was no significant effect of heat stress on WAZ in the multivariate analysis. This lack of effect may be due to several reasons: the in utero environment may protect against the effects of heat on growth; lack of power to detect an effect in utero (far fewer repeated measurements), in keeping with the literature on heat exposure and birth weight which requires large numbers of individuals to demonstrate an effect;^12,26,27 behavioural changes during pregnancy, including increasing rest and shade, reducing the risks of exposure; or the effects of high temperatures on the risks of premature labour and stillbirth may be mediated through short term exposure and acute stressors that have no measurable impact on growth.^12,28,29

We found an increase in HAZ with increasing UTCI up to an apparent limit of 29°C mean UTCI, after which there was no further increase or perhaps a decline (wide 95%CIs are inconclusive), which is consistent with existing studies. For instance, there have been many human studies exploring the ecogeographical phenomenon described as Allen’s rule, where in homeothermic species, individuals from the same species have shorter limbs in cooler climates compared to those living closer to the equator.^30,31 Austin et al. modelled human thermoregulation and found a consistent positive correlation between surface area to volume ratio and heat tolerance, i.e. taller individuals are better adapted to hotter climates, likely due to increased ability to loose heat.³² To corroborate that longer limbed individuals loose heat more easily than shorter limbed individuals, Tilkens et al. demonstrated this on direct measurements.³³

This study builds on extensive work performed by The Medical Research Council Unit in The Gambia. An early study from Rowland et al. established a link between growth faltering in children aged 3 months to 3 years and gastrointestinal infections.³⁴ Although they found both height and weight significantly negatively correlated with diarrheal disease, which impeded any catch-up growth that was seen when diarrhoea was not present, subsequent work found no improvement in growth faltering despite a significant reduction in diarrhoea rates.³⁵ In our study, neither diarrhoea alone, nor any infectious episode in the month preceding anthropometric measurements altered the effect of heat stress on growth faltering metrics. Since the 1970s there have been multiple nutritional interventions, improvements in clinical care, vaccinations, access to water and sanitation and general infrastructure. During this time infant anthropometry has been routinely collected and allows analysis of changes in growth metrics over time. Nabwera et al. analysed this and found that the proportion of children with stunting or underweight halved from 1970s to 2000s, but remained stubbornly high at 30.0% and 22.1% respectively, with almost no change in wasting rates.⁷ This analysis concluded that despite intensive nutrition interventions, there are missing factors in the established understanding of contributors to growth faltering. We suggest, based on the results reported here, that increasing environmental heat stress may be one of those factors.

Furthermore, this study supports the scarce published literature on this topic. Two previous studies have utilised Demographic and Health Surveillance (DHS) data to explore the relationship between heat stress and growth faltering.^13,36,37 Both these cross-sectional studies were based in Sub-Saharan Africa, with Tusting et al. including 656,107 children across 29 countries and Baker et al. including 192,000 across 30 countries. Both studies found an increase in wasting with increasing temperature exposure, as we have shown. Additionally, Tusting et al.¹³ found a decreased odds of stunting in those exposed to temperatures above 35°C compared to below 30°C and Baker et al.³⁶ found a similar shaped curve of the relationship between HAZ and heat as our study. Although our findings are similar, the methodology is very different.

The strength of the findings from this study lies in the rigorous methodology, where multiple standardized measurements on individuals over time, adjusted for seasonality and infectious episode, reduces the likelihood of both bias and confounding. Limitations to the study include that the data come from only one area in rural Gambia and generalisability, especially in relationship to potential thresholds may be uncertain. We also used mothers’ residences to geolocate mother-infant pairs, but this did not take into account relocation or potential movement within the region. Nevertheless, this is likely to have a minimal effect as daily temperature variation across the region was small.

Future studies to explore heat stress and growth faltering in different populations would help determine risk thresholds for different populations, which could inform public health measures. Furthermore, exploration of interventions to reduce heat load (through personal or structural building interventions) and the impact that has on appetite, food intake, food availability and infant growth would help determine effective, evidence-based solutions to adapt to global heating.

In conclusion, our study suggests that increasing heat stress is associated with decreasing WAZ and WHZ. With global heat records being broken yearly and currently no net reduction in greenhouse gas emissions, our findings have major implications for the protection of child health and development under a changing climate.

Inclusion and Ethics

Ethics approval for the original study was granted by the Gambia Government/MRC Laboratories joint ethics committee in August 2008 (ref SCC 1126v2) and further ethical approval from the Gambia Government/MRC joint ethics committee and the London School of Hygiene and Tropical Medicine Ethics Advisory Board (ref 25171) for this analysis in April 2021.

Birth cohort data

The Early Nutrition and Immunity Development (ENID) randomised trial was based in the West Kiang region of The Gambia and was performed from 2010–2015. This rural area is a resource-poor region with seasonal variations in climate, nutritional exposures and birth outcomes.^38,39 The ENID trial recruited pregnant participants prior to 20 weeks gestation and used obstetric ultrasound to determine gestational age (GA) at entry into the trial (mean GA at enrolment 13.3 weeks). Pregnancy dating was done using the crown-rump-length or bi-parietal diameter, depending on the GA at enrolment. Full details of the trial can be found in the published protocol.⁴⁰

Briefly, pregnant women were allocated into one of four arms where they received the following nutritional supplementation from enrolment until delivery: i) iron & folate; ii) multiple micronutrients; iii) protein-energy; iv) multiple micronutrients & protein-energy. In addition, infants between 6–18 months of age were randomised to two further nutritional supplement arms: a) lipid-based supplement with micronutrients, b) lipid-based supplement alone. During their pregnancy maternal anthropometry including weight, height, mid-upper arm circumference and skin fold thickness were taken. Repeated fetal ultrasound scans were done at weeks 20 and 30 gestation of the following parameters: head circumference, biparietal diameter, abdominal circumference, femur length, and tibia length. All measurements were taken in triplicate by trained sonographers and averaged for analysis. Estimated fetal weight was calculated using the Intergrowth-21 calculator based on head and abdominal circumference and from this the gestational age specific z-score.⁴¹ All mother-infant pairs were visited within 72 hours of delivery for a health assessment and collection of newborn anthropometries and birth data.⁴² Weight and length measurements were collected at the following weeks post birth: 1, 8, 12, 24, 40, 52, 78 and 104.⁴³ After birth z-scores for weight-for-age (WAZ), height-for-age (HAZ) and weight-for-height (WHZ) from 2 months to 2 years were calculated using the World Health Organization Multicentre Growth Reference Study.⁴⁴ Weekly visits up to 2 years collected questionnaire data on any acute illness, including fever, diarrhoea, vomiting, cough, difficulty in breathing or any other infection. These data were aggregated as any illness and any diarrheal illness in the 4 weeks prior to each anthropometric measurement visit.

Environmental exposure data

The ERA5-HEAT (Human thermal comfort) dataset from the Copernicus Climate Change Service (C3S) was used to determine location specific heat stress (Universal Thermal Climate Index – UTCI).⁴⁵ ERA5 uses quality-controlled historical meteorological station observations to give global gridded estimates by advanced modelling and data assimilation. Data is gridded at 0.25 by 0.25 degrees, equivalent to approximately 27km² at the equator, and available hourly.⁴⁶ Use of UTCI (a composite measure of heat, humidity, wind speed and solar radiation) was considered of importance due to the wide variation in humidity in The Gambia and because UTCI more effectively captures the physiological response to heat stress exposure compared to air temperature alone.^47–49 Hourly data were first summarised to daily mean UTCI and then further averaged over 30, 60 and 90 days. The village of residence for each mother-infant pair was matched by geolocation to the UTCI grid. The date of each anthropometric measurement (in utero and in infancy) was taken as the end point of 30, 60 or 90 days UTCI exposure. This allowed long-term exposures to be explored, which in the case of growth faltering was considered important.

Statistical analysis

Four outcomes, WAZ in utero (gestational age 20 and 30 weeks and birth) and WAZ 0–2 years; HAZ 0–2 years; and WHZ 0–2 years, were evaluated with regards to heat stress exposures by performing longitudinal multilevel, multivariate linear regression analyses. Multilevel modelling was adopted given the hierarchical nature of the data, with repeated measurements within individuals, who were further clustered by village and study arm.

Thus, we included three random effects, at individual, village and study arm level, aiming to take into account unmeasured but likely existing heterogeneity, between individuals, villages and study arms.

First, we explored the relationship between UTCI and each outcome (WAZ in utero, WAZ 0–2, WHZ 0–2, and HAZ 0–2) using linear and non-linear models. Non-linear models were tested by natural and cubic splines with different knot placement and 2 degrees of freedom. The model with cubic splines and one knot at 0.5 had the lowest Akaike Information Criterion (AIC) and was used in all subsequent models.

Secondly, multivariate multilevel models of a priori determined covariates, including both time varying and time constant covariates were used to evaluate heat stress exposure on outcomes of interest. The covariates are presented in Table 1 and included maternal, child and environmental exposures that have been linked to growth faltering, including seasonality. Despite no clear signal on seasonal decomposition (Supplement Fig S1-4), due to strong a priori evidence of seasonality on the outcomes considered here,^39,50 we used natural spline and two degrees of freedom on day of the year to model seasonality.^51,52 Model sensitivity for inclusion of any infectious episode or only diarrhoeal illness in the preceding 4 weeks was assessed, with no difference in heat stress estimates or model performance.

Finally the dlnm R package was used to synthetize the exposure response curve for UTCI and WAZ, HAZ and WHZ, centred at 26°C UTCI, above which is defined as various degrees of heat stress.⁵³ All models were assessed for compliance with the following model assumptions: linearity, homogeneity of variance and normal distribution of model residuals. All analyses were performed in R version 4.1.0 and R code for all models can be found in the supplement.

Data availability

Anonymised data will be made available on reasonable request from the corresponding author and subject to relevant approvals.

Conflict of interest

None declared

Funding

This project was funded by the Wellcome Trust through the Wellcome Trust Global Health PhD Fellowship awarded to AB (216336/Z/19/Z). The funders had no role in study design, data collection, analysis, manuscript writing or decision to submit.

Acknowledgements

We would like to acknowledge the communities in West Kiang who participated and supported the original data collection and the whole study team involved in the original study.

Black, R. E. et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. The Lancet 382, 427–451, doi:https://doi.org/10.1016/S0140-6736(13)60937-X (2013).
Akombi, B. J., Agho, K. E., Merom, D., Renzaho, A. M. & Hall, J. J. Child malnutrition in sub-Saharan Africa: A meta-analysis of demographic and health surveys (2006–2016). PLoS One 12, e0177338, doi:10.1371/journal.pone.0177338 (2017).
Martins, V. J. B. et al. Long-lasting effects of undernutrition. International journal of environmental research and public health 8, 1817–1846, doi:10.3390/ijerph8061817 (2011).
Prentice, A. M., Moore, S. E. & Fulford, A. J. Growth faltering in low-income countries. World Rev Nutr Diet 106, 90–99, doi:10.1159/000342563 (2013).
Rogol, A. D. Emotional Deprivation in Children: Growth Faltering and Reversible Hypopituitarism. Front Endocrinol (Lausanne) 11 (2020).
Quilter, C. R. et al. Identification of methylation changes associated with positive and negative growth deviance in Gambian infants using a targeted methyl sequencing approach of genomic DNA. FASEB Bioadv 3, 205–230, doi:10.1096/fba.2020-00101 (2021).
Nabwera, H. M., Fulford, A. J., Moore, S. E. & Prentice, A. M. Growth faltering in rural Gambian children after four decades of interventions: a retrospective cohort study. The Lancet Global Health 5, e208-e216, doi:10.1016/S2214-109X(16)30355-2 (2017).
Pickering, A. J. et al. The WASH Benefits and SHINE trials: interpretation of WASH intervention effects on linear growth and diarrhoea. The Lancet Global Health 7, e1139-e1146, doi:10.1016/S2214-109X(19)30268-2 (2019).
The state of the global climate 2020. (World Meteorological Organization, 2021).
IPCC. Climate Change 2021: the physical science basis. Contribution of working group I to the sixth assessment report of the Intergovernmental Panel on Climate Change. (IPCC, IPCC, 2021).
Ebi, K. L. et al. Hot weather and heat extremes: health risks. Lancet 398, 698–708, doi:10.1016/s0140-6736(21)01208-3 (2021).
Chersich, M. F. et al. Associations between high temperatures in pregnancy and risk of preterm birth, low birth weight, and stillbirths: systematic review and meta-analysis. BMJ 371, m3811, doi:10.1136/bmj.m3811 (2020).
Tusting, L. S. et al. Environmental temperature and growth faltering in African children: a cross-sectional study. The Lancet Planetary Health 4, e116-e123, doi:https://doi.org/10.1016/S2542-5196(20)30037-1 (2020).
Abraham, E. et al. Pregnancy exposure to atmospheric pollution and meteorological conditions and placental DNA methylation. Environment international 118, 334–347, doi:10.1016/j.envint.2018.05.007 (2018).
More, V. S. Fever in pregnancy and its maternal and fetal outcomes. 6, 5523–5527 (2017).
Wells, J. C. Thermal Environment and Human Birth Weight. J Theor Biol 214, 413–425, doi:10.1006/jtbi.2001.2465 (2002).
Yates, D. T. et al. ASAS-SSR Triennnial Reproduction Symposium: Looking back and moving forward—how reproductive physiology has evolved: Fetal origins of impaired muscle growth and metabolic dysfunction: Lessons from the heat-stressed pregnant ewe. J Anim Sci 96, 2987–3002, doi:10.1093/jas/sky164 (2018).
Chua, P. L. C. et al. Association between Ambient Temperature and Severe Diarrhoea in the National Capital Region, Philippines. International journal of environmental research and public health 18, doi:10.3390/ijerph18158191 (2021).
Wells, J. C. Environmental temperature and human growth in early life. J Theor Biol 204, 299–305, doi:10.1006/jtbi.2000.2015 (2000).
Lieber, M., Chin-Hong, P., Kelly, K., Dandu, M. & Weiser, S. D. A systematic review and meta-analysis assessing the impact of droughts, flooding, and climate variability on malnutrition. Glob Public Health 17, 68–82, doi:10.1080/17441692.2020.1860247 (2022).
de Onis, M. & Habicht, J. P. Anthropometric reference data for international use: recommendations from a World Health Organization Expert Committee. The American journal of clinical nutrition 64, 650–658, doi:10.1093/ajcn/64.4.650 (1996).
Schlaudecker, E. P. et al. Small for gestational age: Case definition & guidelines for data collection, analysis, and presentation of maternal immunisation safety data. Vaccine 35, 6518–6528, doi:10.1016/j.vaccine.2017.01.040 (2017).
Karlsson, O., Kim, R., Guerrero, S., Hasman, A. & Subramanian, S. V. Child wasting before and after age two years: A cross-sectional study of 94 countries. eClinicalMedicine 46, doi:10.1016/j.eclinm.2022.101353 (2022).
UNICEF. Levels and trends in child malnutrition. (WHO, 2021).
Kinyoki, D. K. et al. Mapping local patterns of childhood overweight and wasting in low- and middle-income countries between 2000 and 2017. Nat Med 26, 750–759, doi:10.1038/s41591-020-0807-6 (2020).
Yitshak-Sade, M. et al. The effect of prenatal temperature and PM(2.5) exposure on birthweight: Weekly windows of exposure throughout the pregnancy. Environment international 155, 106588, doi:10.1016/j.envint.2021.106588 (2021).
Tapia, V. L. et al. Association between maximum temperature and PM(2.5) with pregnancy outcomes in Lima, Peru. Environ Epidemiol 5, e179, doi:10.1097/ee9.0000000000000179 (2021).
Bonell, A. et al. Environmental heat stress on maternal physiology and fetal blood flow in pregnant subsistence farmers in The Gambia, west Africa: an observational cohort study. The Lancet Planetary Health 6, e968-e976, doi:https://doi.org/10.1016/S2542-5196(22)00242-X (2022).
Samuels, L. et al. Physiological mechanisms of the impact of heat during pregnancy and the clinical implications: review of the evidence from an expert group meeting. Int J Biometeorol, doi:10.1007/s00484-022-02301-6 (2022).
Serrat, M. A. Environmental temperature impact on bone and cartilage growth. Comprehensive Physiology 4, 621–655, doi:10.1002/cphy.c130023 (2014).
Foster, F. & Collard, M. A reassessment of Bergmann's rule in modern humans. PLoS One 8, e72269, doi:10.1371/journal.pone.0072269 (2013).
Austin, D. M. & Lansing, M. W. Body Size and Heat Tolerance: A Computer Simulation. Human Biology 58, 153–169 (1986).
Tilkens, M. J., Wall-Scheffler, C., Weaver, T. D. & Steudel-Numbers, K. The effects of body proportions on thermoregulation: an experimental assessment of Allen's rule. J Hum Evol 53, 286–291, doi:10.1016/j.jhevol.2007.04.005 (2007).
Rowland, M. G., Cole, T. J. & Whitehead, R. G. A quantitative study into the role of infection in determining nutritional status in Gambian village children. Br J Nutr 37, 441–450, doi:10.1079/bjn19770047 (1977).
Poskitt, E. M., Cole, T. J. & Whitehead, R. G. Less diarrhoea but no change in growth: 15 years' data from three Gambian villages. Arch Dis Child 80, 115–119; discussion 119–120, doi:10.1136/adc.80.2.115 (1999).
Baker, R. E. & Anttila-Hughes, J. Characterizing the contribution of high temperatures to child undernourishment in Sub-Saharan Africa. Scientific reports 10, 18796, doi:10.1038/s41598-020-74942-9 (2020).
USAID. The DHS Program, <https://dhsprogram.com/topics/Nutrition/Index.cfm> (
Ceesay, S. M. et al. Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5 year randomised controlled trial. Bmj 315, 786–790, doi:10.1136/bmj.315.7111.786 (1997).
Fulford, A. J. C., Rayco-solon, P. & Prentice, A. M. Statistical modelling of the seasonality of preterm delivery and intrauterine growth restriction in rural Gambia. Paediatr Perinat Epidemiol 20, 251–259 (2006).
Moore, S. E. et al. A randomized trial to investigate the effects of pre-natal and infant nutritional supplementation on infant immune development in rural Gambia: the ENID trial: Early Nutrition and Immune Development. BMC pregnancy and childbirth 12, 107, doi:10.1186/1471-2393-12-107 (2012).
Stirnemann, J. et al. International estimated fetal weight standards of the INTERGROWTH-21(st) Project. Ultrasound Obstet Gynecol 49, 478–486, doi:10.1002/uog.17347 (2017).
Villar, J. et al. International standards for newborn weight, length, and head circumference by gestational age and sex: the Newborn Cross-Sectional Study of the INTERGROWTH-21st Project. Lancet 384, 857–868, doi:10.1016/s0140-6736(14)60932-6 (2014).
Eriksen, K. G., Andersson, M., Hunziker, S., Zimmermann, M. B. & Moore, S. E. Effects of an Iodine-Containing Prenatal Multiple Micronutrient on Maternal and Infant Iodine Status and Thyroid Function: A Randomized Trial in The Gambia. Thyroid: official journal of the American Thyroid Association 30, 1355–1365, doi:10.1089/thy.2019.0789 (2020).
de Onis, M. et al. The WHO Multicentre Growth Reference Study: planning, study design, and methodology. Food Nutr Bull 25, S15–26, doi:10.1177/15648265040251s103 (2004).
Di Napoli, C., Barnard, C., Prudhomme, C., Cloke, H. L. & Pappenberger, F. ERA5-HEAT: A global gridded historical dataset of human thermal comfort indices from climate reanalysis. Geoscience Data Journal 8, 2–10, doi:https://doi.org/10.1002/gdj3.102 (2021).
Services, C. C. C. (Copernicus Climate Change Service Climate Data Store).
Jendritzky, G., de Dear, R. & Havenith, G. UTCI—Why another thermal index? Int J Biometeorol 56, 421–428, doi:10.1007/s00484-011-0513-7 (2012).
Havenith, G. & Fiala, D. Thermal indices and thermophysiological modeling for heat stress. Comprehensive Physiology 6, 255–302, doi:10.1002/cphy.c140051 (2016).
Romaszko, J., Dragańska, E., Jalali, R., Cymes, I. & Glińska-Lewczuk, K. Universal Climate Thermal Index as a prognostic tool in medical science in the context of climate change: A systematic review. Sci Total Environ 828, 154492, doi:https://doi.org/10.1016/j.scitotenv.2022.154492 (2022).
Silver, M. J. et al. Environmentally sensitive hotspots in the methylome of the early human embryo. elife 11, doi:10.7554/eLife.72031 (2022).
Dominguez-Salas, P. et al. Maternal nutrition at conception modulates DNA methylation of human metastable epialleles. Nat Commun 5, 3746, doi:10.1038/ncomms4746 (2014).
Johnson, W., Elmrayed, S. A., Sosseh, F., Prentice, A. M. & Moore, S. E. Preconceptional and gestational weight trajectories and risk of delivering a small-for-gestational-age baby in rural Gambia. The American journal of clinical nutrition 105, 1474–1482, doi:10.3945/ajcn.116.144196 (2017).
Gasparrini, A. Distributed Lag Linear and Non-Linear Models in R: The Package dlnm. J Stat Softw 43, 1–20 (2011).

Table 1: Model parameters included in both the in utero and 0-2 years multilevel model of heat stress on WAZ, WHZ and HAZ.

In utero to birth model variables	0-2 years model variables
*Maternal factors*	*Maternal factors*
Maternal age	Maternal age
Enrolment BMI	Enrolment BMI
Maternal MUAC*	Parity
Parity
*Child-level factors*	*Child-level factors*
Age*	GA at birth
Sex	Age*
	Sex
	Acute illness in last 4 weeks*
*Environmental factors*	*Environmental factors*
Heat stress (UTCI)*	Heat stress (UTCI)*
Season of conception (natural spline)	Season of conception (natural spline)
Season of exposure*	Season of exposure*
	Year*

* = time varying covariates. BMI = body mass index; MUAC = mid-upper arm circumference; GA = gestational age

Table 2: Model estimates for in utero heat stress exposure on weight-for-age z-score and 0-2 years heat stress exposure on weight-for-age, weight-for-height and height-for-age z-scores. Heat stress defined as 95^th percentile UTCI at 30, 60 and 90 days exposure compared to 26°C.

30 days exposure	60 days exposure	90 days exposure
In Utero to birth
0.02 (-0.17;0.21)	0.10 (-0.12;0.31)	-0.057 (-0.29; 0.17)
0-2 years WAZ
-0.06 (-0.11;-0.0050)	-0.11 (-0.17;-0.04)	-0.18 (-0.27;-0.095)
0-2 years WHZ
-0.14 (-0.21;-0.076)	-0.22 (-0.31;-0.14)	-0.29 (-0.40; -0.18)
0-2 years HAZ
0.026 (-0.027;0.079)	0.019 (-0.045;0.083)	-0.085 (-0.17; -0.001)

In utero multivariable model adjusted for maternal age, maternal BMI at enrolment, maternal MUAC, parity, sex of infant, age, season of conception, season of measurement. Infant model adjusted for maternal age, maternal BMI at enrolment, parity, gestational age at birth, age, sex, acute illness in prior 4 weeks, season of conception, season of exposure, year. BMI = body mass index; MUAC = mid-upper arm circumference

There is NO Competing Interest.

Supplement.docx
Supplementary material

Download PDF

Version 1

posted

You are reading this latest preprint version

Assessing the impact of heat stress on growth faltering in the first 1000 days of life in rural Gambia

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Online Methods

Inclusion and Ethics

Birth cohort data

Environmental exposure data

Statistical analysis

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1