Low and High Ambient Temperatures during Pregnancy and Birth Weight among 624,940 Singleton Term Births in Israel (2010–2014): An Investigation of Potential Windows of Susceptibility

Background: Exposure to heat during pregnancy has been associated with reduced fetal growth. Less is known about associations with cold and the potential for critical time windows of exposure. Objectives: We aimed to evaluate, in a national retrospective cohort, critical windows of susceptibility during pregnancy to extreme temperatures (low and high) and fetal growth, among 624,940 singleton term births in Israel during the period 2010–2014. Methods: Temperature exposures were estimated using a spatially refined gridded climate data set with a 1-h and 1-km2 resolution. Percentiles of temperature were categorized by climatic zone for the entire pregnancy and by trimesters and weeks. Generalized additive models with the distributed lag nonlinear model framework were used to estimate unadjusted and adjusted associations between percentiles and categories of temperature and fetal growth markers: term [births after 36 weeks of gestational age (GA)] mean birth weight and term low birth weight (tLBW, term infants with birth weight below 2,500g). Results: After adjustment, extreme temperatures (percentiles) during the entire pregnancy were associated with a lower mean birth weight {≤10th vs. 41st–50th percentile: –56g [95% confidence interval (CI): –63g, –50g)]; >90th vs. 41st–50th percentile: –65g; 95% CI: –72g, –58g}. Similar inverse U-shaped patterns were observed for all trimesters, with stronger associations for heat than for cold and for exposures during the third trimester. For heat, results suggest critical windows between 3–9 and 19–34 GA-weeks, with the strongest association estimated at 3 GA-weeks (temperature >90th vs. 41st–50th percentiles: –3.8g; 95% CI: –7.1g, –0.4g). For cold, there was a consistent trend of null associations early in pregnancy and stronger inverse associations over time, with the strongest association at 36 GA-week (≤10th vs. 41st–50th percentiles: –2.9g; 95% CI: –6.5g, 0.7g). For tLBW, U-shape patterns were estimated for the entire pregnancy and third trimester exposures, as well as nonsignificant associations with heat for 29–36 GA-weeks. Generally, the patterns of associations with temperatures during the entire pregnancy were consistent when stratified by urbanicity and geocoding hierarchy, when estimated for daily minimum and maximum temperatures, when exposures were classified based on temperature distributions in 49 natural regions, and when estimated for all live births. Discussion: Findings from our study of term live births in Israel (2010–2014) suggest that exposure to extreme temperatures, especially heat, during specific time windows may result in reduced fetal growth. https://doi.org/10.1289/EHP8117

. Annual mean results for cross validation of the INCA system for five stations, for 2012 and 2014.   Table S2. Summary statistics of the mean temperatures in every climatic zone during research period (2010)(2011)(2012)(2013)(2014) and the distribution within the singleton term livebirths born in that climatic zone (N=624,940) for the entire pregnancy and trimester-specific average estimates. Table S3. Summary statistics of the daily maximum and minimum temperatures in every climatic zone during research period (2010)(2011)(2012)(2013)(2014) and the distribution within the singleton term livebirths born in that climatic zone (N=624,940) for the entire pregnancy and trimester-specific average estimates. Figure S3. Unadjusted generalized additive model (GAM) plots showing the predicted estimated mean term birthweight (tBW) as smooth function of the entire pregnancy average of the daily mean temperature (°C), for every climatic zone (A-C) and (D) combined with percentiles calculated by the climatic zone. The tick marks on the x-axis are observed data points. The y-axis represents the smooth predicted estimated tBW. The shaded areas indicate the 95% confidence intervals. Figure S4. Illustration of potential confounding variables and possible selection bias in the association between temperature exposure and birthweight and the minimal adjustment set to block all biasing paths. Directed Acyclic Graph (DAG) drawn and analyzed with DAGitty (Textor et al., 2017). Table S4. Summary of the loading and the Pearson correlations between the nine socioeconomic variables and the first area-level socio-economic factor, using Principal Component Analysis (PCA). Description of the Imputation Procedure Table S5. Unadjusted and adjusted associations (differences in mean and ORs (95% CI)) between entire pregnancy average of the daily mean temperature, categorized to centiles by climatic zones and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014. Table S6. Unadjusted and adjusted associations (differences in mean (g) and ORs (95% CI)) between trimester specific average of the daily mean temperature, categorized to centiles by climatic zones and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014. Figure S6. Adjusted exposure-lag-response surfaces for the associations (differences in mean (g) and ORs (95% CI)) between temperature percentiles, relative to the 50 th percentile, and tBW and tLBW, for singleton live births born in Israel during 2010-2014. Table S7. Unadjusted and adjusted associations (differences in mean (g) (95% CI)) between extreme week specific temperature (cold (<=10 th ) and heat (>90th)), compared to the 41-50 th reference category) and tBW, for singleton live births born in Israel during 2010-2014. Table S8. Unadjusted and adjusted associations (OR (95% CI)) between extreme week specific temperature (cold (<=10 th ) and heat (>90 th ), compared to the 41-50 th reference category) and tLBW, for singleton live births born in Israel during 2010-2014. Table S9. Unadjusted and adjusted associations (differences in mean (g) and ORs (95% CI)) between entire pregnancy average of the daily mean temperature, categorized to centiles by climatic zones and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014, for those living in urban areas or rural and semi-rural areas. Table S10. Unadjusted and adjusted associations (differences in mean (g) and ORs (95% CI)) between entire pregnancy average of the daily mean temperature, categorized to centiles by climatic zones, and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014, for those whose addresses were geocoded at the settlement level or at the home or street level. Table S11. Summary statistics of the entire pregnancy and trimester specific averages of the daily mean temperature, the number of births and the area by the natural region. Table S12. Unadjusted and adjusted associations (differences in mean (g) and ORs (95% CI)) between entire pregnancy and trimester specific average of the daily mean temperature categorized to centiles by natural regions and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014. Table S13. Unadjusted and adjusted associations (differences in mean (g) and ORs (95% CI)) between entire pregnancy average of the daily maximum and minimum temperature, categorized to centiles by climatic zones and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014.   Figure S1). Figure S1. Scatterplots for the cross-validation analysis, the observed vs. predicted hourly temperature for each station, for 2012 and 2014, based on approximately 8,750 (hourly) data points. Table S2. Summary statistics of the mean temperatures in every climatic zone a during research period (2010)(2011)(2012)(2013)(2014) and the distribution within the singleton term livebirths born in that climatic zone (N=624,940) for the entire pregnancy and trimesterspecific average estimates.   Table S4. Summary of the loading a and the Pearson correlations between the nine socioeconomic variables and the first arealevel socio-economic factor, using Principal Component Analysis (PCA). a the amount of variance in each variable, that is accounted for the first area-level socio-economic factor in a principal component analysis (PCA). Figure S5. The Scree plot a of the Principal Component Analysis (PCA). a a scree plot is a plot of the eigenvalues that is used to determine the number of factors to retain in a principal component analysis (PCA).

•
Iterative method: the fully conditional specification (FCS ; also known as sequential regression and MI using chained equations (MICE)) (with 5 iterations).
-Parental variables: maternal education, paternal education, maternal age, paternal age, maternal family status, maternal origin, paternal origin, maternal working status, paternal working status.
-Variables determined based on address: geocoding level, urbanicity level.
Area-level socio-economic variables: percentage of academic graduates aged 25-54, percentage of those with any income from work aged 15+, percentage of women aged 25-54 who are not in the civilian labor force, percentage of workers with income above twice the average wage, percentage of recipients of income assurance, percentage of those with income from work below the minimum wage, monthly average income per person (standardized), average years of education for those aged 25-54, the ratio of the number of "dependent"(0-19 year old and 65+ ) residents to the number of "providers" (20-64 year old), area-level socio-economic variable (factor).
-Exposure variables: Entire pregnancy average of the daily mean temperature, first trimester average of the daily mean temperature, second trimester average of the daily mean temperature, third trimester average of the daily mean temperature, entire pregnancy average of the daily maximum temperature, first trimester average of the daily maximum temperature, second trimester average of the daily maximum temperature, third trimester average of the daily maximum temperature, entire pregnancy average of the daily minimum temperature, first trimester average of the daily minimum temperature, second trimester average of the daily minimum temperature, third trimester average of the daily minimum temperature, centiles by climatic zones of the entire pregnancy average of the daily mean temperature, centiles by climatic zones of the first trimester average of the daily mean temperature, centiles by climatic zones of the second trimester average of the daily mean temperature, centiles by climatic zones of the third trimester average of the daily mean temperature; centiles by climatic zones of the entire pregnancy average of the daily maximum temperature, centiles by climatic zones of the first trimester average of the daily maximum temperature, centiles by climatic zones of the second trimester average of the daily maximum temperature, centiles by climatic zones of the third trimester average of the daily maximum temperature; centiles by climatic zones of the entire pregnancy average of the daily minimum temperature, centiles by climatic zones of the first trimester average of the daily 14 minimum temperature, centiles by climatic zones of the second trimester average of the daily minimum temperature, centiles by climatic zones of the third trimester average of the daily minimum temperature; climatic zones, natural regions, All variables were included as continuous variables except the following: -Dichotomous variables: LBW, PTD, child sex, congenital malformation, maternal working status, paternal working status, maternal family status, geocoding level, urbanicity level.
-Maternal and paternal education, maternal and paternal age were included as continuous and as categorical. Table S5. Unadjusted and adjusted associations a (differences in mean and ORs (95% CI)) between entire pregnancy average of the daily mean temperature, categorized to centiles by climatic zones b and tBW and tLBW (compared to 41-50 th reference centile), for singleton live births born in Israel during 2010-2014. a These were estimated by generalized linear models with normal distribution and identity link function (for tBW) or with binomial family and logit link function (for tLBW). Models were adjusted for: newborn ethno-religious group, parity, maternal age, maternal marital status, maternal origin, maternal occupation, maternal education, year and month of LMP, area-level socio-economic variable; for complete case (N= 370,691), unadjusted and imputed data (N=624,940). LMP-last menstrual period; tBW-term mean birthweight (mean birthweight among births>36 gestational age weeks); tLBW-term low birthweight (births>36 gestational age weeks with birthweight <2500 grams); OR-odds ratio. b See Table S2 for corresponding numeric data of the temperature categories. Complete case adjusted results are presented graphically in Figure 1 and Figure 3. tBW (differences in mean (g) 1.50, 1.80) a These were estimated by generalized linear models with normal distribution and identity link function (for tBW) or with binomial family and logit link function (for tLBW). Models were adjusted for: newborn ethno-religious group, parity, maternal age, maternal marital status, maternal origin, maternal occupation, maternal education, year and month of LMP, area-level socio-economic variable; for complete case (N= 370,691), unadjusted and imputed data (N=624,940). T1-first trimester; T2-second trimester; T3-third trimester. LMP-last menstrual period; tBW-term mean birthweight (mean birthweight among births>36 gestational age weeks); tLBW-term low birthweight (births>36 gestational age weeks with birthweight <2500 grams); OR-odds ratio. b See Table S2 for corresponding numeric data of the temperature categories. Complete case adjusted results are presented graphically in Figure 2 and Figure 4. Figure S6. Adjusted a exposure-lag-response surfaces for the associations (differences in mean (g) and ORs (95% CI)) between temperature percentiles, relative to the 50 th percentile, and tBW and tLBW, for singleton live births born in Israel during 2010-2014. a These were estimated by nonlinear distributed lag models (DLNM). Models were adjusted for: newborn ethno-religious group, parity, maternal age, maternal marital status, maternal origin, maternal occupation, maternal education, year and month of LMP, area-level socio-economic variable; for complete case (N= 370,691). LMP-last menstrual period; tBW-term mean birthweight (mean birthweight among births>36 gestational age weeks); tLBW-term low birthweight (births>36 gestational age weeks with birthweight <2500 grams); OR-odds ratio.