Impact of exposure to extreme heat events during pregnancy on the incidence of congenital heart disease in offspring: a meta-analysis Authors: Malavika Krishnakumar, Aparna Hari, Georg Gutjahr, Perraju Bendapudi, Hisham Ahamed, Poornima Prabhakaran, Raman Krishna Kumar, Manu Raj Published: 2024 ABSTRACT AbstractBackgroundThe aetiology of congenital heart disease (CHD) is multifactorial. Environmental risk factors have emerged as an important modifiable determinant of several congenital cardiac conditions. Previous studies have shown a strong relationship between CHD and air pollution. Much less is known about the influence of ambient temperature on CHD. The primary objective of this meta-analysis was to examine the association between exposure to extreme heat events (EHE) during pregnancy and CHD in the offspring.Methodscomprehensive literature search was conducted using electronic databases such as Scopus and PubMed, along with a review of secondary references. This process yielded eight studies that met the inclusion criteria: four from the USA, two from China, one from Israel and one from Canada. The risk of bias was assessed using the Newcastle Ottawa scale. Mixed-effect regression was used to pool the estimates of individual studies. Heterogeneity was measured by I2. Results were visualised by a forest plot, a Baujat plot and an albatross plot. Subgroup analyses were performed for climate zones, exposure definitions and the study region. As a sensitivity analysis, a leave-one-out meta-analysis was performed to assess the robustness of the estimates.ResultsExposure to EHE during pregnancy increased the risk for CHD in the offspring (OR=1.12, 95% CI 1.04 to 1.34). Studies from the USA observed the smallest effect (OR=1.01, 95% CI 0.91 to 1.13), studies in the continental climate zone observed a slightly larger effect (OR=1.07, 95% CI 0.97 to 1.19), and studies from the temperate climate zone observed the largest effect (OR=1.35, 95% CI 1.23 to 1.48). Subgroup analysis was conducted with respect to exposure, effect definition and region. The residual heterogeneity measures were 88% (exposure), 61% (effect definition) and 38% (Region).ConclusionThe association between prenatal EHE exposure and future risk of CHD has important implications for pregnant mothers, infant health and health policy. Future studies should explore the additional burden of CHD contributed by current trends in global ambient temperature in other parts of the world.PROSPERO registration numberCRD42023455934Keywords: Cardiology, Epidemiology, Neonatology WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS Environmental factors like exposure to extreme heat events during early pregnancy can increase risk of congenital heart disease in offspring. Environmental factors like exposure to extreme heat events during early pregnancy can increase risk of congenital heart disease in offspring. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Future research should focus on the role of environmental factors like rising ambient temperature in modifying the risk of congenital heart disease on a global scale. Future research should focus on the role of environmental factors like rising ambient temperature in modifying the risk of congenital heart disease on a global scale. Introduction Congenital heart defect (CHD) is defined as structural abnormalities in the heart or major blood vessels present at birth.1 CHD is the most prevalent birth defect, with recent statistics indicating approximately nine cases per 1000 live births globally.2 While most cases of CHD are noncritical in nature, inadequate management of critical defects can lead to substantial morbidity and mortality.3 Despite significant progress in the diagnosis and management of CHD, our comprehension of the underlying causes remains incomplete.4 Nearly all congenital heart anomalies manifest during the first trimester, the crucial phase of heart development. Approximately 20% of CHD cases can be attributed to factors such as chromosomal abnormalities, single gene disorders and exposure to teratogens.2 Limited evidence points to the potential impact of environmental factors such as air pollution and elevated temperature or heat waves possibly playing a role in CHD development.3 In the past decade, several epidemiological studies have investigated the potential link between heat exposure during pregnancy and CHD. However, the results of the studies are inconsistent, and the association is not well understood. The research questions are defined by the following population, exposure, comparator and outcome terms5: population, all newborns; exposure, extreme heat events (EHE) during the critical period of heart development; comparator, pregnancy during normal temperature conditions; and outcome, the presence or absence of CHD in the foetus. The primary objective of this meta-analysis is to quantify the extent of the association between maternal exposure to EHE during pregnancy and the likelihood of CHD in the foetus. The precise mechanisms linking temperature to heart defects remain unclear but may involve a direct impact of heat or a heat shock response. Maternal exposure to elevated temperatures could result in foetal cell death, potentially causing congenital malformations, as observed in animal models.2 Additionally, increased temperatures trigger a heat-shock response, prompting the production of heat-shock proteins. Although this response is typically short-lived, disruptions in crucial developmental processes due to reduced transcription and translation of normal proteins could lead to embryonic death, retarded growth and developmental abnormalities.6 While severe heat-shock responses in pregnant women due to elevated temperatures are unlikely, extreme heatwaves may trigger milder responses that could plausibly result in noncritical conditions like atrial septal defect (ASD).3 Methods Protocol and registration The study protocol was designed based on the Preferred Reporting Items for Systematic Review and Meta-analysis Protocols and has been registered at the PROSPERO (CRD42023455934). Eligibility criteria Inclusion criteria were defined for cohort or case-control studies that investigate exposure to EHE (definitions given below) during the critical period of pregnancy and CHD. The critical period for exposure is defined as weeks 2–8 of the gestation period, as this is the time period of organogenesis of the heart.7 Animal studies and methodological reviews were excluded. All studies that did not report an estimate of the association between the ambient temperature and CHD or provided sufficient data to get such an estimate were excluded. Information sources A comprehensive literature review search was conducted on 4 July 2023 in electronic databases Scopus and PubMed. We also conducted a snowball search to identify additional studies by searching the reference lists of publications eligible for full-text review and using Google Scholar to identify and screen studies citing them. Full details on the search strategies are presented in online supplemental appendix table 1. Study selection We extracted studies based on the screening strategy. All duplicates were removed through Mendeley. The selected titles and abstracts were screened by two researchers (AH, KM) independently. In case of a disagreement, a third researcher (PB) provided the final decision. Two investigators (AH and PB) independently screened full-text articles for final selection based on inclusion criteria. In case of disagreement, consensus was reached on inclusion or exclusion by discussion, and if necessary, the third investigator, (HA) was consulted. Data collection process All the selected full-text studies were independently reviewed for final data extraction by the two investigators (KM, GG). Any disagreements between authors were resolved through discussion. The final extracted data include the name of the first authors, year, title, journal, exposure definition, time period of exposure, the place and climate of the study, total sample size, total number of cases, estimates of the association between exposure and outcome, and the variability of the estimates. Definitions The outcome was defined as the diagnosis of CHD, which includes both critical and non-critical conditions. The following definitions of the exposure variables were used: (1) temperature exceeding a predefined cut-off for a certain time period as defined in the individual studies;3 8 (2) EHE 95, defined as the two consecutive days above the 95th percentile of the temperature in the study region; and (3) EHE 90, defined as the three consecutive days above 90th percentile of the temperature in the study region. If studies reported estimates for multiple definitions of EHE, we preferred EHE 95. All effects were extracted from the source papers and converted to log ORs independently by two authors (KM, GG). In case of disagreement, the two authors discussed the paper and decided how the extraction should be handled. Studies were classified by climate zones according to the Köppen climate classification, as shown in figure 1.9 Koppen climate classification Koppen climate classification and the location of the selected studies. Figure adapted from Beck et al.13 Handling of missing data In studies where relevant data were missing, the reviewing authors searched for the original sources from the institution’s library or contacted the primary authors to obtain any missing information. If the data could not be obtained, then the particular study was excluded from the analysis. Risk of bias The risk of bias was assessed using the Newcastle Ottawa scale (NOS),10 which is used to assess the quality of the included studies. The details of the assessed categories can be found in table 1. Two researchers (AH, PB) independently assessed the risk of bias for each study. In case of disagreement, a third author (GG) made the final decision. The risk of publication bias was assessed using a funnel plot. Table 1Study characteristics of the selected studiesStudyStudy designStudy locationClimate zoneSample sizeExposure periodHeat exposureRisk of bias*SelectionComparabilityOutcomeTotalJudge et al8Case controlUSAContinental502 cases, 1066 controls3–8 weeks1 day >38°4228Van Zutphen et al14Case controlUSAContinental6422 cases, 59 328 controls3–8 weeksEHE903216Agay-Shay et al7CohortIsraelContinental135 5273–8 weeksEHE904228Auger et al3CohortCanadaContinental704 2092–8 weeks3 days >30°3216Lin et al15Case controlUSAMixed US†5848 cases, 5742 controls3–8 weeksEHE90 and EHE954217Stingone et al4Case controlUSAMixed US†2632 cases, 4033 controls3–8 weeksEHE90 and EHE954217Yu et al2CohortChinaTemperate1 918 1052–8 weeksEHE90 and EHE954116Xu et al6CohortChinaTemperate983 523Individual weeksEHE 952114*Risk of bias was assessed using the Newcastle Ottawa scale.†Continental-temperate-arid climate zones in the USA.EHEextreme heat events Study characteristics of the selected studies Risk of bias was assessed using the Newcastle Ottawa scale. Continental-temperate-arid climate zones in the USA. EHEextreme heat events Synthesis We categorised all studies that satisfy the inclusion and exclusion criteria according to the definition of EHE, the study region and the climate zone. For one study (Xu et al), the estimates of the OR were estimated from figures. For two studies (Stingone et al and Auger et al), the OR was calculated from the reported events in the case and control groups. For the remaining studies, an estimate of the OR was extracted from the papers. A mixed-effect regression model was used to pool the estimates of individual studies. Heterogeneity was measured by I2. We used a forest plot to display the effect estimates, CIs and supplementary statistics for the included studies. We used an albatross plot to visualise the effect of p values on the study size and a Baujat plot to visualise the influence of different studies. Subgroup analyses were performed for climate zones, exposure definitions and the study region. As a sensitivity analysis, a leave-one-out meta-analysis was performed to assess the robustness of the estimates. All analyses were done in R V.4.3.1 using the packages meta, dmetar, and metap.11 Patient and public involvement statement Patients and public were not involved in the design, conduct, reporting, or dissemination plans of this study. Results Study selection Figure 2 shows the selection process. We initially screened 463 studies for eligibility. We removed duplicates, and 336 unique records remained after this process. We proceeded to abstract screening process and selected 14 studies for full-text review. We did a comprehensive evaluation of these full-text articles and identified eight studies that met the predefined inclusion and exclusion criteria. Among the eight studies that were selected, four were from the USA, two from China, one from Israel and one from Canada. Online supplemental tables 1,2 list studies that passed the initial screening but were excluded after thoroughly assessing the full texts. Online supplemental table 3 offers a comprehensive overview of the characteristics of the selected studies. Study selection process. Flow diagram showing the results of the search and selection process from the number of records identified in the search to the number of studies included in the review. Study characteristics The study characteristics, including the study design, study location, climate zone, sample size, length of exposure, type of exposure and the risk of bias assessment, are presented in table 1. The NOS risk of bias tool was used to grade the selected studies.10 According to the Agency Healthcare Research and Quality standards,12 studies with a total point score of at least six are considered ‘good’, and 3–5 points are considered ‘fair’. Additional study details are defined in the online supplemental appendix. Result of synthesis Figure 3 shows a forest plot with the effect estimates, the confidence intervals of each study, the overall summary estimate as well as the estimates in each climate zone. Effect estimates in the studies differed significantly between the climate zones (p<0.01). Stratifying by climate zones reduced the heterogeneity from 91% (overall heterogeneity) to 36% (residual heterogeneity). Hence, adjusting for climate zones reduced the heterogeneity by 55%. Forest plot displaying individual study effects. Forest plot with the effects of the individual studies, the overall pooled estimates, the estimates in each of the subgroups and the estimates of the heterogeneity overall or after adjusting for the respective climate zones. The size of each box represents the study weight. The overall effect of EHE on the risk for CHD was estimated by an OR of 1.12 with a CI of 1.04–1.34, which was highly significant (p<0.01). Studies with a mix of different locations in the USA observed the smallest effect (OR=1.01, 95% CI 0.91 to 1.13), studies in the continental climate zone observed a slightly larger effect (OR=1.07, 95% CI 0.97 to 1.19), and studies from the temperate climate zone observed the largest effect (OR=1.35, 95% CI=1.23 to 1.48). Subgroup analysis was also conducted with respect to the exposure, effect definition and region. The residual heterogeneity measures obtained after classifying according to the respective subgroups were 88% (exposure), 61% (effect definition) and 38% (Region). The risk of publication bias was assessed by the contour-enhanced funnel plot in online supplemental figure 1. Studies with the least variable effect sizes appear at the top of the funnel, and the smaller, less precise studies appear at the bottom. Contours were added to the plot, which corresponds to the lines of statistical significance. The studies included in the analysis belong to the area of statistical significance, which can be an indication that asymmetry is due to publication bias. ASD was the most common diagnosis of CHD in most of the studies. Figure 4 shows the difference in the effect of EHE on ASD and other types of CHD (both critical and non-critical CHDs). The effect of exposure to EHE on the diagnosis of ASD showed considerable variation across studies, while its effect on the diagnosis of other types of CHD was similar, except for Zutphen et al, who found a very large effect on other types but little effect on ASD. Scatter plot of OR for atrial septal defect (ASD) and other congenital heart diseases. The solid line indicates equal ORs, while points below the line indicate a larger OR for ASD. The Baujat plot in online supplemental figure 2 shows that the study by Yu et al has the largest effect on the pooled result due to the high sample size. As an additional way to test the robustness of the estimates, the leave-one-out meta-analysis in online supplemental figure 3 again shows that Yu et al has the largest influence on the heterogeneity measure and that leaving out any of the studies does not change the overall pooled estimate much. Furthermore, the Albatross plot in online supplemental figure 4 shows that most studies point towards a single standardised effect size. Ethical clearance This study is a systematic review and meta-analysis by design and an ethical clearance is not required due to the nature of the study design. Discussion Using population-based data for a large number of neonates (3 833 794), we illustrate that maternal exposure to ambient temperature or extreme heat waves during 2–8 weeks postconception was associated with an increased risk of congenital heart defects in the offspring. The risk for CHD development from EHE had an OR of 1.13. To the best of our knowledge, this study is the first meta‐analysis assessing the association between heat exposure and risks of developing congenital heart defects. Identification of such modifiable risk factors aids public health practitioners in devising preventive programmes to promote behaviours that reduce the burden of CHD. Over the past decade, various epidemiological studies have assessed the impact of temperature exposure on congenital heart anomalies.3 4 7 However, the findings from these studies have been inconsistent across different populations. Possible reasons may include the difference in study design, population characteristics, and geographical variation. Notably, a majority of these studies were conducted in regions with moderate climate zones, where the average daily temperatures tend to be relatively low. Consequently, it is inappropriate to extrapolate these findings to tropical countries. In this meta-analysis, the subgroup analyses found variable results in the different climate zones, and adjusting for these zones reduced heterogeneity from 91% to 36%. Apart from one study (Xu et al), there was a consensus in the definition of the critical period (week 2–8 postconception) of exposure in all the studies. However, Xu et al speculated that the impact of EHE on the development of CHD might extend beyond the period of embryogenesis. In the majority of the studies, ASD was the most common CHD. The ambiguity in the definition of ASD might have contributed to the considerable variation of effects of EHE on the diagnosis of ASD in the various studies, whereas the effects on other types of CHD have been relatively uniform. As global temperatures continue to rise, the frequency, severity and duration of EHE are expected to increase, making it imperative for the public to be aware of the risks of exposure to heat during pregnancy. Furthermore, CHD may impair the individuals’ quality of life and impose substantial financial burdens on families.2 Therefore, identifying modifiable risk factors and addressing these upstream determinants early in life is of paramount importance. In light of the lack of any literature on the tropical and arid zones, there is scope for investigation of the influence of ambient temperature on pregnancy outcomes in these climatic zones. Future studies should address some of the limitations of the current studies, taking into account stillbirths, abortions, and standardised definitions for CHD. Study limitations We would like to acknowledge two study limitations related to incomplete information and residual confounding. The included studies failed to report stillbirths and abortions. A subset of these two could probably have been the result of underlying CHD. This lack of data could probably result in under-reporting of the actual association between study exposure and CHD. There is a possibility of residual confounding due to the lack of data related to presence or absence of air conditioning in the residence or at the workplace during the mothers’ exposure period. We acknowledge that the generalisability of the study results is sub-optimal given the localised geographical distribution of the study sites. In addition, half of the included studies (four) were from USA. These factors limit the degree to which the study results can be generalised across the globe. Conclusions The current study establishes a positive association between maternal exposure to EHE during early pregnancy and the risk of CHD in the foetus. There seems to be significant heterogeneity for the effect of EHE across various time zones with temperate zone showing the maximum impact. The study findings underline the importance of the impact of climate change while formulating public health policies. The link between rising ambient temperature and the risk of CHD needs to be explored in detail by future research studies from other part of the world. This is essential for a better understanding of this phenomenon and appropriate action towards mitigation of the after effects of the same. supplementary material Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Data availability free text: Data are available upon reasonable request after appropriate clearance from the IRB of the study institution. Map disclaimer: The inclusion of any map (including the depiction of any boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied. Patient and public involvement: Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research. Contributor Information Malavika Krishnakumar, Email: malavikakrishnakumar2017@gmail.com. Aparna Hari, Email: aparnahari018@gmail.com. Georg Gutjahr, Email: georg.gutjahr@gmail.com. Perraju Bendapudi, Email: bendapudip@googlemail.com. Hisham Ahamed, Email: ahamed.hisham@gmail.com. Poornima Prabhakaran, Email: Poornima.prabhakaran@ccdcindia.org. Raman Krishna Kumar, Email: kumar_rk@yahoo.com. Manu Raj, Email: drmanuraj@gmail.com. Data availability statement Data are available upon reasonable request. References Supplementary Materials Data Availability Statement Data are available upon reasonable request.