Parental risk factors for congenital diaphragmatic hernia – a large German case-control study

Congenital diaphragmatic hernia (CDH) represents a severe birth-defect with an estimated birth prevalence of 1 in 2500 newborns [1]. It is characterized by a defect in the embryonic development of the diaphragm which results in an incomplete separation of the abdominal cavity and the thoracic cavity. This incomplete separation leaves to variable degrees stomach, intestine, liver or spleen displaced in the thoracic cavity compromising pulmonary development with concomitant pulmonary hypoplasia and secondary pulmonary hypertension at birth [2]. Normally, the diaphragm develops during the fourth week of gestation and closes until twelfth week [3]. Different types of CDH are classified according to their location. The most common type is the posterolateral Bochdalek hernia, a defect in the lumbocostal trigonum with herniation mostly resulting left-sided [4]. Current knowledge on the underlying cause for CDH is still scarce. Genetic factors might play an important role since about 20% occur as part of a genetic defect or syndrome [5]. In the majority of cases CDH presents as an isolated defect or as one of many malformations without any genetic abnormality detected [6]. The most common co-occuring malformations compromise the cardiovascular, genitourinary, gastrointestinal or the central nervous system [7]. Besides genetic factors the following environmental risk factors have been previously associated with the development of CDH: maternal body mass index (BMI) [8], maternal age [9, 10], pre-gestational diabetes [11], maternal pre-gestational hypertension [12], pre-conceptional diet, especially riboflavin supplementation [13].

In the present study, we hypothesized that periconceptional and prenatal environmental factors have a greater impact on the development of CDH than previously assumed [14]. Here we compared the impact of parental periconceptional smoking and alcohol intake, maternal periconceptional exposure to coffee, sweeteners and hairspray as well as maternal body mass index (BMI). Furthermore, we analyzed if the development of CDH might be influenced by chronic diseases like maternal hypothyroidism or maternal periconceptional folic acid intake, previous miscarriages or infertility treatment. We compared data of environmental risk factors obtained from 199 CDH patients born at the University Hospital of Bonn and the University Hospital of Mannheim between 2001 and 2019 to data of environmental risk factors obtained from 597 healthy control newborns born all over Germany in 2011.

Data of 199 CDH patients consisting of 85 female and 114 male patients born between 2001 and 2019 were compared with 597 healthy control infants (289 females and 305 males) born in 2011. Demographics of CDH and controls are shown in Table 1. The male-to-female ratio is hence 1.341 compared to 1.055 in the normal population, which corresponds to an odds ratio (OR) of 1.271 (95%-CI: 0.919, 1.757) for male gender (P = 0.163). 42 cases were treated at the University Hospital of Bonn and 157 of CDH cases were treated at the University Hospital of Mannheim. The majority of CDH-patients was born with a left-sided CDH (82.9%), 14.6% had a right-sided CDH and one patient presented with bilateral CDH. 66 cases (33.2%) required ECMO-assistance. 76.4% of the CDH-patients presented with isolated CDH defects and 21.6% were so-called “complex CDH-cases” with additional anomalies. Most of these associated defects were related to the cardiovascular system (62.8%) followed by malformations of the respiratory, genitourinary, central nervous or gastrointestinal system (Table 2). Both gestational age and birth weight of CDH cases were significantly lower compared with controls (P < 0.0001). There was no statistically significant difference between both groups according to the parent’s age at the time of child’s birth (P = 0.147 for maternal age and P = 0.502 for paternal age). Parent’s baseline characteristics according to both groups are shown in Table 3. Altogether we analyzed more than 25 different variables as potential risk factors. The impact of known risk factors for the development of CDH are summarized in Table 4 (univariate analyses) and Table 5 (multivariate analysis).

Here, in a case-control study we investigated potential environmental risk factors in 199 CDH patients compared to 597 healthy control newborns. Our analyses did not find association between parental periconceptional smoking and CDH. However, we observed significant association for maternal periconceptional alcohol intake. Accordingly, previous studies by Felix et al. in 2008 and McAteer et al. in 2014 described association between the occurrence of CDH and maternal periconceptional alcohol intake [11, 16]. In general, CDH is not part of the foetal alcohol related anomaly spectrum [17]. Nevertheless, previous studies report an association of maternal alcohol consumption and the expression of birth defects [18]. Yet, no pathophysiological concept exists that could causally link periconceptional or pregnancy related maternal alcohol consumption and the expression of structural birth defects. Likely, the observational association of maternal alcohol consumption with structural birth defects is more of an indirect association, warranting future studies, observational or experimental, to investigate this. The here described association between maternal periconceptional use of hairspray, the development of CDH, and the fact that hairsprays have already been discussed to possibly bear an impact on pregnancy and embryogenesis warrants further studies [19]. Especially, since a recent study by Karzi et al. in 2019 showed that the frequent use of hairspray and sunscreens is correlated with higher levels of parabens in hair of the pregnant women [20]. Recently, several experimental studies have investigated the embryonic toxicity of parabens using zebrafish larvae providing several lines of evidence, for the detrimental effects of parabens on zebrafish early-life stages with attention to its developmental toxicity [21, 22]. Observed features among zebrafish larvae where pericardial edema, yolk edema, blood stasis, reduction in blood circulation, reduced heartbeat and notochord curvature [23]. While parabens are commonly found in personal care products such as hair spray, it is unclear, whether the maternal use during the periconceptional period might exert toxic effects on human embryonic development. Hence, the current observation should warrant measurements of parabens in women using on a regular level hair spray or paraben containing care products and correlate the measured quantities in women to the toxic quantities used in the above mentioned zebrafish studies.

While previous studies suggested an association of periconceptional maternal overweight and CDH [24, 25] our study could not replicate this association (P = 0.526). The studies of García et al. published in 2016 [9] and Paoletti et al. published in 2020 [10] revealed a significant increased risk of bearing a child with CDH for mothers with an age ≥ 35 years (P < 0.001 and P = 0.0004). While our study did not replicate this association, median maternal age of both, case and control mothers in our study, was below 35 years of age. In 2014, McAteer et al. [11] described an association between maternal pregestational diabetes and CDH (P = 0.003). Our study could not replicate this finding (P = 0.827). Furthermore, we did not observe the previously described association between periconceptional maternal hypertension and CDH by Mesas Burgos et al. in 2019 [12]. In fact, we did not find any association between maternal chronic diseases of the cardiovascular or respiratory spectrum and CDH in their offspring. However, our study is suggestive (P = 0.051) for an association between maternal periconceptional hypothyreosis and CDH which should warrant further studies in the future.

Our study did not reveal any potential associations between maternal periconceptional intake of medical drugs and CDH. Here, Crider et al. had previously observed an association between periconceptional maternal intake of sulfonamides and CDH [26]. Furthermore, our results do not support maternal periconceptional folic acid supplementation to be protective in the prevention of CDH. However, prevention of CDH has not been an indication for maternal periconceptional folic acid supplementation.

Regarding periconceptional maternal vitamin supplementation, the respective questionnaire items were barely answered except for folic acid intake. This is most likely due to the fact, that in Germany, no monitoring or systematic information exists on what and how many vitamins should be supplemented during pregnancy, except for folic acid. Nevertheless, previous reports by Beurskens et al. in 2013 [27] describe an association between ‘dietary vitamin A intake during pregnancy below the recommended daily intake (800 μg per day) in normal-weight mothers’ and the risk of bearing a child with CDH (P = 0.01). Recently Carmichael et al. described low maternal qualitative diets and lack of riboflavin to be associated with CDH formation in the offspring [13]. Moreover, Turkmen et al. described lower maternal vitamin D (P = 0.019) and calcium levels (P < 0.001) to be associated with ‘pregnancies complicated by CDH’ and hypothesized that hypovitaminosis D may ‘play a vital role in the pathogenesis of CDH’ [28]. The exact mechanisms of different hypo- or hypervitamonsis and the expression of birth defects are vastly unknown. Yet, the observations made for transcription factor GATA-4, encoded by the GATA4 gene (MIM #600576) and its function in a vitamin A-deficient embryonic environment could be of model character. GATA4 belongs to the few genes identified to play a role in human CDH formation [29]. On the other hand, Feng et al. were able to show in an experimental mouse study, that the mouse embryos of vitamin A-deficient mothers, presented with a high incidence of cardiac defects [30]. Interestingly, they also found high methylation status in the CpG loci of GATA4 gene with a low expression of GATA-4 mRNA from vitamin A-deficient group embryos. Hence, embryonic transcriptome maps of the developing diaphragm in vertebrates could yield further genetic drivers, that may play a role in (human) diaphragm development allowing for the identification of genetic drivers, that require certain vitamins or track elements. Here, hypomorphic alleles, possibly discovered by genome-wide association studies and a deficient intrauterine environment might lead to CDH or other common birth defects.

To the best of our knowledge, the present study is the first German study that assessed the potential association between CDH and various parental environmental risk factors. Here, we were able to recruit a large sample of CDH families and respective healthy controls newborns in a match of 1:3 (cases:controls) to yield maximal statistical power and to facilitate the detection of associations between potential risk factors and CDH.

Our study is limited by a potential recall bias among CDH case mothers or fathers due to the retrospective collection of data, which in some cases was carried out up to 17 years after the child’s birth whereas the control group received the questionnaire almost immediately after delivery. Furthermore, we cannot exclude, a potential bias in both groups, cases and controls, especially among case parents in that parents with an unhealthy lifestyle habit might not have been willing to participate due to guilt feelings. Nevertheless, our results suggest an increased direct or indirect risk for CDH formation to be associated with maternal periconceptional alcohol consumption regardless of the beverage chosen or the frequency of consumption. Furthermore, we observed an association between maternal periconceptional use of hairspray, which might point to an embryonic impact of parabens on embryonic development.