Congenital anomaly requiring medical care constitutes approximate three percent of newborns and congenital cardiovascular defect is the leading cause of death [1-3]. In many countries including China, the contribution of birth defects to infant mortality has been increasing [4, 5]. Congenital heart disease is one of the most common birth defects with an incidence of 4 to 8 per 1,000 live births [4, 6, 7] and in China, the prevalence of congenital heart disease is 7 to 8 per 1,000 live births [8]. However, with improvements in prenatal diagnostic techniques and awareness of the importance of periconceptional folic acid supplementation, the incidence of neural tube defects has decreased significantly [9].

Investigation of modifiable risk factors attributable to heart defects has long been the focus of epidemiological research [10-12] and approximate 30% of the potentially modifiable factors were deemed contributing to some type of defects [13]. Currently, there is data showing contributing factors for different types of heart disease varies and regional differences in environmental risk factors were observed [14-16]. However, population-based data on non-inherited risk factors is still lacking and there is no reliable information available for developing preventive strategies.

The purpose of this retrospective study was to explore the relationship between non-inherited risk factors and congenital heart diseases through reviewing the medical records and collecting relevant data between 2012 and 2014. The study mainly focused on investigating maternal and paternal risk factors, as well as other factors identified during pregnancy follow-up.

This is a retrospective study of all born newborn cases in Panyu Maternal and Child Care Service Centre of Guangzhou. All data were retrieved from the hospital maternal care database and no special exclusion criteria was included in this study. Each case was identified by the name, gender, date of birth and bodyweight of the newborn baby and the information obtained from pregnancy follow-up was identified by the mother’s name and date of follow-up. After excluding all duplicate records, each newborn case was matched with the maternal and paternal information, as well as the relevant clinical information obtained during pregnancy follow-up. Any case with missing data was not included for statistical analyses.

Maternal baseline data collected included maternal age, Chinese ethnic group, census data, mother’s height, pre-pregnancy weight, mother’s education level, and family income. Maternal and delivery histories were also collected, which included gravidity, parity, and history of induced abortion, spontaneous abortion, threatened abortion, preterm birth, and still birth. Mother’s work environment classified as high risk including the workplace with high temperature, loud noise, dusty environment, peculiar smell, and those with exposure to chemicals, heavy metal and ionizing radiation, as well as contact with organic solvents were compared; however details of the distribution were not displayed. Other maternal factors identified during pregnancy were also studied, which included exposure to drugs/substances (e.g. antibiotics, antipyretic analgesics, NSAID, tocolytic, anticancer drugs, vitamin A, birth control pills, etc.), supplementation (e.g. folic acid and multivitamins), and other environmental exposure (e.g. toxic or harmful substances, X-ray irradiation, B-mode USG examination, electromagnetic radiation, ionizing radiation and high-intensity noise). Other routine assessments of vital signs and laboratory tests during pregnancy were also included in this retrospective review. Paternal data were also collected, which included the followings: paternal age, history of congenital heart disease or others, father’s education level and occupation, previous risk of work environment, smoking and drinking habit, history of drug abuse, and previous exposure to X-ray irradiation, electromagnetic radiation, ionizing radiation, high-intensity noise, heavy metals, organic solvents, and pesticides.

All numerical data were expressed as mean (standard deviation) and were compared between groups using t-test for independent samples. If normality was not met, the medians were described and compared using non-parametric test. Count data were compared using Pearson χ²-test or Fisher’s exact test. Non-parametric Mann-Whitney U test was used for data not normally distributed. Statistically significant individual factors identified from univariate logistic regression analysis were included in the subsequent multivariate conditional logistic stepwise regression analysis to confirm the independent risk or preventive factors for congenital abnormality. Statistical analyses were performed using SAS9.4. P-values of less than 0.05 were considered statistically significant in this study.

This study was approved by Medical Ethics Committee of China Panyu He Xian Memorial Hospital, No. Panhetong 201224. This study also obtained signed informed consent from all participants.

A total of 5,636 newborn cases in Panyu Maternal and Child Care Service Centre of Guangzhou were retrospectively reviewed. Among which, 145 cases were diagnosed congenital cardiac diseases. The three major cardiac diseases consisted of patent ductus arteriosus (n=39), atrial septal defect (n=15), and perimembranous ventricular septal defect (n=13). Compared to the normal newborns (n=5482), the gestational age of abnormal newborns was significantly lower than that of normal newborns (35.01 ± 5.83 weeks vs. 38.60 ± 1.72 weeks, p < 0.001) and the average weight was also significantly lower (2525.47 ± 1050.56 g vs. 3152.08 ± 457.83 grams, p < 0.001).

According to univariate logistic regression analysis of 5,381 evaluable cases, the following maternal factors were statistically associated with the congenital abnormality: maternal age, education level, occupation, previous work at high-risk factories, passive smoking, alcohol drinking, gravidity, history of induced abortion, history of spontaneous abortion, history of threatened abortion, folic acid supplementation before or during pregnancy, exposure to toxic or harmful substances before or during pregnancy, previous use of birth control pills, tocolytic drugs, vitamin A and other drugs (including traditional Chinese medicine / anticonvulsant / lithium agent), maternal ultrasound examination, high-intensity noise (> 80 db, ≥ 8 hours), extreme working conditions, weight, and thrombin time. Details of comparison of all maternal parameters between normal and abnormal fetal groups are displayed in Tables 1-5.

Multivariate regression analysis revealed that maternal age (OR = 1.039, 95% CI: 1.014-1.065, p = 0.002), gravidity (OR = 1.221, 95%: 1.107-1.348, p < 0.0001), passive smoking (OR = 1.763, 95% CI: 1.149-2.707, p = 0.01), maternal ultrasound examination (OR = 2.008, 95% CI: 1.210-3.333, p = 0.007), high-risk working condition (OR = 2.412, 95% CI: 1.278-4.552, p = 0.007), and thrombin time (OR = 0.895, 95% CI: 0.817-0.982, p = 0.019) were significantly associated with increased risk of congenital abnormality.According to another univariate logistic regression analysis of 5,436 evaluable cases on paternal factors, the following factors were statistically associated with the congenital abnormality: paternal age, education level, highrisk workplace (printing industry, the footwear industry), alcohol consumption, high intensity noise (> 80 db, ≥ 8 hours), and heavy metal exposure. Details of comparison of all paternal parameters between normal and abnormal fetal groups are displayed in Tables 6 and the result of univariate analysis is presented in Table 7.

Multivariate regression analysis revealed that paternal age (OR= 1.022, 95% CI: 1.007-1.038, p = 0.005), higher education levels (junior high school: OR = 0.258, 95% CI: 0.124-0.537, p = 0.000, high school: OR = 0.231, 95% CI: 0.112-0.476, p < 0.0001, college: OR = 0.192, 95% CI: 0.088-0.415, p < 0.0001, and undergraduate: 0.152, 95% CI: 0.066-0.350, p < 0.0001), except for postgraduate education levels when compared to those with less than primary education, alcohol consumption (OR = 1.617, 95% CI: 1.072-2.440, p = 0.022) and heavy metal exposure (OR = 4.466, 95% CI: 1.556-12.818, p = 0.005) were significantly associated with increased risk of congenital abnormality.

According to another univariate logistic regression analysis of 5,381 evaluable cases for investigating the risk factors during pregnancy, extreme work environments such as high temperature workplace, solvent exposure, and RBC count were significantly associated with the occurrence of congenital abnormality. Details of comparison of other maternal factors during follow-up between normal and abnormal fetal groups are displayed in Tables 3 and 5 and the result of univariate analysis is presented in Table 7.

Multivariate regression analysis did not reveal any risk factor during pregnancy associated with congenital abnormality, but RBC count was revealed as a protective factor (OR = 0.349, 95% CI: 0.180-0.677, p = 0.002).

When maternal factors, paternal factors, and other factors identified from regular follow-up during pregnancy, a total of 5,261 evaluable cases were analyzed. Multivariate regression analysis revealed that maternal age (OR = 1.076, 95% CI: 1.007-1.148, p = 0.029), gravidity (OR = 1.153, 95% CI: 1.024-1.298, p = 0.018), and RBC count (OR = 0.374, 95% CI: 0.187-0.745, p = 0.005) measured during pregnancy follow-up were significantly associated with the occurrence of congenital abnormality after controlling other factors (Table 8).

There are conflicting results on the association between mother’s education level and congenital heart disease [17-19]. In the present study, there was a significant association between mother’s education level and the occurrence of congenital anomaly (p = 0.043), and a trend of inverse correlation was observed. When compared to mothers who only received primary education using univariate logistic regression, those who had received undergraduate education were less likely to have a baby with congenital abnormality (OR= 0.248, 95% CI: 0.077-0.802), however, other levels of education were not considered relevant to the abnormal outcome. It was also confirmed from the multivariate regression analysis that mother’s higher education levels was not a protective factor.

Other maternal factors such as mother’s age, gravidity, passive smoking, maternal ultrasound examination, and high-risk working condition were identified as independent risk factors of congenital heart disease. Similar to other studies [20-23], mothers with older age, increased gravidity, passive smoking, and those working in high-risk condition were probably associated with the congenital defects. However, maternal ultrasound examination was considered as a risk factor of the congenital problem, whereas there is lack of evidence showing the association with maternal exposure to ionizing radiation or display terminals [24-26]. Evidence for the association of thrombin time and the abnormal outcome was not found in the literature as well. In agreement with suggestions by Liu et al. [27], the results might be biased and further study is warranted.

In this study, the relationship between paternal factors and the outcome of congenital heart defects were also explored. Similar to other large-scale studies [28-31], higher paternal age, lower education level, alcohol consumption, and heavy metal exposure were considered as possible risk factors of congenital anomaly. It further testifies that as for the Chinese population, better health education and prevention of exposure to alcohol and heavy metal might be able to reduce the chance of congenital heart defects.

Together with maternal and paternal factors, as well as other factors identified during pregnancy follow-up, only two independent risk factors, higher maternal age, and increased gravidity, and one protective factor of higher RBC count measured during pregnancy follow-up were identified. Both maternal age and gravidity were discussed above and considered as known risk factors of congenital anomaly. Interestingly, in the present study, a significantly higher maternal RBC count observed in normal fetal group than abnormal fetal group (4.44 ± 3.86×10¹²/L vs. 4.14 ± 0.90×10¹²/L, p = 0.002), was considered an independent protective factor (OR= 0.374, 95% CI: 0.187-0.745, p = 0.005). There is however not much evidence showing the relationship between RBC count and the abnormal outcome. Indeed, RBC itself may provide a storage reservoir of essential elements for fetal development such as folate, which was found closely related with the incidence of main birth defects including congenital heart defects [32]. Therefore, the maternal level of RBC-folate and its relationship with birth defects should be investigated in future studies.

One of the major limitation of the retrospective study was that the sample size was limited by the number of available medical records and unbalanced sample size included for analysis. It may bias the study results as discussed previously. A large-scale prospective study is recommended.

The retrospective analysis of maternal database in a single institution revealed that higher mother’s age and increased gravidity might increase the risk, whereas the higher RBC count might reduce the chance of congenital heart diseases.