Association between maternal obesity and metabolic disorders and congenital heart defects in the offspring: a literature review

Congenital heart defects (CHDs) are the most common congenital malformation and will, in severe cases, have a serious impact on neonatal mortality and morbidity. The aetiology of CHDs is complex. Large cohort studies have reported an association between increased risk of CHDs in the offspring and individual maternal metabolic disorders such as diabetes, hypertension, preeclampsia, and obesity. All conditions that can be related to insulin resistance and possibly metabolic syndrome (MetS). The aim of this review is to evaluate the existing evidence on the association between maternal metabolic disorders, defined as obesity, diabetes, hypertension, preeclampsia, dyslipidaemia, and MetS, or combinations thereof and CHDs in the offspring. A literature search was performed using PubMed and Embase databases. Of the 2,076 potentially relevant identified studies, 30 qualified for inclusion. Only one study dealt with the combination of more than one maternal metabolic condition as risk factor for CHDs in the offspring. All other studies investigated the individual metabolic disorders and their association with CHDs. Some disorders (chronic hypertension, gestational diabetes, and obesity) increased risk of CHDs marginally whereas pregestational diabetes and early-onset preeclampsia were highly associated with CHDs. Future studies on the combination of several metabolic disorders in the same pregnancy and their association with CHDs are needed.


64
Congenital heart defects (CHDs) are structural malformations of the heart and/or the great intrathoracic vessels. They 65 are the most frequent congenital malformation (1) and birth prevalence is estimated to be around 8-10 per 1,000 live 66 births worldwide (1)(2)(3). The most critical and severe defects, major CHDs, (prevalence of 0.14% in Denmark) have a 67 serious impact on neonatal mortality and morbidity, and frequently result in neonatal heart failure or circulatory 68 collapse requiring acute surgery (3). Whereas, other types of CHDs like bicuspid aortic valve and transient septal heart 69 defects are far more frequent but more often with limited clinical significance.

70
Population-based prevalence of CHDs are difficult to determine, but Denmark's electronic registers are a useful 71 resource for ascertaining the number of diagnosed cases as well as treatment complications and outcomes. Using 72 such data, Lytzen et al. found that 75% of children with major CHDs, born in Denmark 1996Denmark -2013 73 Furthermore, the 1-year mortality rate for this group was 19.6%, although the rate decreased during the study period 74 (4). Improved infant treatment has led to a decrease in infant mortality and an increase in children and adults living 75 with CHDs (5-7). Indeed, a Canadian study estimated that adults account for two thirds of patients with CHDs (8). The 76 mortality rate is increased in adults with CHDs compared to the general population with the majority of patients with 77 CHDs succumbing to cardiovascular causes such as arrhythmia and heart failure (9).

84
The normal pregnancy is an adaptation between the maternal metabolism and foetal development. Differentiation of 85 cardiac tissues begins in the third week of gestation, and by week 8, the foetal heart has undergone major changes, 86 and will resemble the postnatal heart in structure and function (14). Thus, the maternal-foetal interaction in first 87 trimester is likely to be the most relevant for an association between maternal metabolism and CHDs. In first 88 trimester, the foetus does not have the ability to secrete insulin which may result in foetal hyperglycaemia in the 89 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. . https://doi.org/10.1101/2020 event of relative maternal insulin resistance (15). The significance of foetal hyperglycaemia has not been 90 demonstrated in humans, however, animal models have shown that in embryos of chicks and rodents, exogenous 91 glucose may cause a variety of malformations (16,17). Furthermore, increasingly worse glycaemic control around 92 conception in women with diabetes type 1 (DM1) was associated with a progressively increased risk of CHDs in the 93 offspring (18).

98
MetS comprises a cluster of metabolic disorders in combination, usually any three of the following: abdominal obesity, 99 insulin resistance, dyslipidaemia and hypertension (28,30). It is commonly described in a non-pregnant population as a 100 risk factor for diabetes type 2 (DM2) and cardiovascular disease. Only one study has assessed MetS in pregnant 101 women and reported an incidence of 12.3% at 15 weeks' gestation (31) using the International Diabetes Federation 102 criteria (30). Since MetS is more prevalent with increasing age (28), it seems plausible that obesity, diabetes, 103 hypertension, PE and dyslipidaemia as individual conditions, or more likely in combination, during pregnancy are 104 related to MetS (28,29). Furthermore, the same conditions have been associated with increased risk of CHDs in the 105 offspring (19)(20)(21)(22)(23)(24)(25)(26)(27). We hypothesize that the combination of several maternal metabolic disorders (defined as obesity, 106 diabetes, hypertension, PE and dyslipidaemia), all related to MetS, could be associated with cumulative risk of CHDs in 107 the offspring.

108
The aim of this review is to evaluate the literature of associations between maternal metabolic disorders (obesity, 109 diabetes, hypertension, PE, dyslipidaemia, or MetS) or combinations thereof and CHDs in the offspring. Furthermore, 110 to point out gaps in current knowledge and make recommendations for future research. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. divided by the square of the height in meters (kg/m 2 )) was defined as pre-pregnancy or early-pregnancy BMI 125 measured in first trimester. The World Health Organization has defined BMI groups  126 BMI 25.0-29.9: overweight; BMI 30.0-34.9: obesity class I; BMI 35.0-39.9: obesity class II; BMI ≥40: obesity class III) 127 (33). The outcome of interest was CHDs defined as structural malformations of the heart chambers, heart valves, great 128 arteries and septal defects, corresponding to DQ20-26 in the World Health Organization International Classification of 129 Diseases (ICD-10) or diagnoses referable to these. The CHD diagnoses are described as "any CHD" for the whole group 130 of diseases, "major CHDs" for a group of the most critical and severe CHD diagnoses (Table 1)  138 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. One author (GH) conducted the study selection and quality assessment based on a full-text review. Any doubts were 141 resolved by discussion with at least one co-author. The Newcastle-Ottawa Scale (NOS) was used to assess the quality 142 of eligible studies (34). Using this tool, each study was judged on eight items, arranged into three categories: selection 143 of cohorts or cases and controls; comparability of cohorts or cases and controls; and ascertainment of outcome or 144 exposure. Highest quality studies could be awarded nine stars. If a study received seven or more stars, it was 145 considered of high methodological quality and included in the review (34).

146
Results are presented as odds ratios (OR), relative risks (RR) or prevalence ratios (PR) with 95% confidence interval (CI) 147 unless other is stated.

174
In an American multi-site, population-based case-control study with 21,762 participants (National Birth Defects 175 Prevention Study, NBDPS), Fisher et el. found use of antihypertensive medication (centrally acting antiadrenergics, β-176 blockers, renin-angiotensin system blockers, calcium channel blockers, and diuretics) from one month before 177 conception through the third month of pregnancy to be associated with an increased risk of both any CHD (OR 1.59; CI 178 1.23-2.05) and five subtypes of CHDs (23). Although, they observed statistically significant estimates for coarctation of 179 the aorta (CoA), pulmonary valve stenosis, Ebstein's anomaly, perimembranous VSD, and ASD secundum, ranging from 180 OR 1.90 (CI 1.09-3.31) for VSD to OR 3.89 (CI 1.51-10.06) for Ebstein's anomaly, they did not find an association for a 181 multiple of other CHD subtypes (23). The study also found a significant association between untreated hypertension 182 or late pregnancy antihypertensive medication initiation and the same subtypes of CHDs (with the exception of 183 Ebstein's anomaly) (23). Information on maternal use of antihypertensive medication in the NBDPS was from 184 telephone interviews with the mother made between six weeks and 24 months after estimated due date (23). (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  Several studies have investigated the association between maternal overweight or obesity and any CHD, but only 210 some have specified how they defined "any CHD" (22,26,27,40,42,46 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. . https://doi.org/10. 1101/2020 Four studies evaluated subtypes of CHDs according to Botto et al.'s classification (56). All studies found positive 240 associations for PGDM and selected subtypes listed in Table 1 (19)(20)(21)52), except for anomalous pulmonary venous 241 return in two studies (19,21), RVOTO in one study (21), and heterotaxia in one study (52). Two studies presented 242 associations between PGDM and major CHDs, but they defined major CHDs differently (  (52)). These studies also reported on GDM and subtypes of CHDs (Table 4) (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. . https://doi.org/10. 1101/2020 No studies assessing the association between CHDs in the offspring and maternal MetS were identified. Only one 265 study dealt with the combination of more than one maternal metabolic condition as risk factor for CHDs in the 266 offspring (40)

288
PE is considered a disease of multiple aetiologies and may present with signs ranging from mild hypertension with a 289 normal foetus to a condition with severe intrauterine growth retardation and a maternal hypertensive syndrome (60). 290 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. . https://doi.org/10. 1101/2020 As the mild hypertension may reflect a physiological response to the placental hypo-perfusion whereas the maternal 291 syndrome is caused by widespread endothelial damage, it is important to study the association between CHDs and PE 292 in cohorts of well-characterized patients. In two large studies, the association between CHDs and PE is much stronger 293 in PE with early-onset (24,25). This is compatible with an interference with foetal glucose regulation in first trimester.

294
Furthermore, one study also report a strong relation between CHD and PE with intrauterine growth retardation (25).

295
Women with chronic hypertension will most likely be undergoing treatment with antihypertensive drugs. It is thus 296 important to distinguish between the effects of antihypertensive drugs and the hypertension proper. A meta-analysis 297 by Ramakrishnan et al. observed an association between untreated hypertension and CHDs, which suggests that the 298 association between hypertension and CHDs is not simply due to teratogenic effects of medication alone (61). But the 299 effect is larger for treated hypertension. Thus, antihypertensive medications may lead to an additional increase in risk.

300
Treated hypertension might also indicate more severe disease in the mother.  (63,64). It is very well-established that maternal obesity is associated with marginally increased risk of CHDs. However, 306 an increase of 20-40% of a very small risk may not be clinically significant. Particularly as it has not been shown that 307 weight reduction normalizes the risk.

308
No articles on dyslipidaemia fulfilled the inclusion criteria for the review. However, one small case-control study from 309 China found significantly increased levels of low-density lipoprotein cholesterol and apolipoprotein B at 24 to 28 310 weeks' gestation in the pregnancies with a foetus with CHD compared to uncomplicated pregnancies (65). The mean 311 BMI was normal in both groups (p=0.83) (65).

312
Our analysis shows a remarkable lack of knowledge on the link between MetS, a major global health problem, and 313 CHDs in the offspring. There is a need for large population-based studies of the risk for foetal CHDs and combinations 314 of various maternal metabolic disorders. Furthermore, such studies should be ethnically inclusive and include women 315 of different socio-economic status. There is also a need for more detailed studies of the molecular mechanisms of 316 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020.

336
The shortage, found in this study, of high-quality studies on the combination of metabolic disorders as a risk factor for 337 CHDs in the child makes it difficult to define treatment and preventive strategies. 338 All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted June 26, 2020. . https://doi.org/10.1101/2020.06.25.20140186 doi: medRxiv preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.