The non‐genetic paternal factors for congenital heart defects: A systematic review and meta‐analysis

Abstract Background Advances have been made in identifying genetic etiologies and maternal risk factors of congenital heart defects (CHDs), while few literatures are available regarding paternal risk factors for CHDs. Thus, we aim to conduct a meta‐analysis and systematic review about the non‐genetic paternal risk factors for CHDs. Methods We searched the PubMed, MEDLINE, and Cochrane Library online databases and identified 31 studies published between 1990 and 2018 according to the inclusion criteria. Paternal risk factors were divided into subgroups, and summarized odd ratios (OR) were calculated. Results Paternal age between 24 and 29 years decreased the risk of CHDs in the offspring (OR = 0.90 [0.82, 0.98]), while paternal age ≥ 35 years old increased the risk of CHDs (35‐39 years old: OR = 1.14 [1.09, 1.19], and ≥ 40 years: OR = 1.27 [1.14, 1.42]). Paternal cigarette smoking increased the risk of CHDs in a dose‐dependent way. Paternal wine drinking (OR = 1.47 [1.05, 2.07]) and exposure to chemical agents or drugs (OR = 2.15 [1.53, 3.02]) also increased the risk of CHDs. Some specific paternal occupations were also associated with increased risk for CHDs or CHD subtypes including factory workers, janitors, painters, and plywood mill workers. Conclusions This meta‐analysis and systematic review suggested that advanced paternal age, cigarette smoking, wine drinking, exposure to chemical agents or drugs and some specific occupations were associated with an increased risk of CHDs. More measures should be taken to reduce occupational and environment exposures. At the same time, fertility at certain age and establishment of healthy life habits are strongly recommended.

worldwide, 1 which were also the leading cause of infant deaths. 2 Over the past decades, there have been advances in the understanding of the risk factors for CHDs, that both genetic and non-genetic risk factors are associated with the prevalence of CHDs. In the past, most investigations focused on maternal and genetic factors, while paternal factors attracted less attention. However, evidences suggested that paternal age, cigarette smoking, wine drinking, and occupational/environment exposures might have associations with various birth defects including CHDs. [3][4][5][6][7][8][9][10][11][12][13][14][15] Therefore, we aimed to provide a current review of paternal factors for CHDs.

| MATERIALS AND METHODS
This report of systematic review and meta-analysis followed the instructions of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). 16

| Search strategy
We searched the PubMed, MEDLINE, and Cochrane Library online databases. We used the selected search terms and the Medical Subject Headings (MeSH) that were related to "congenital heart defect," "risk factor," "exposure," and "paternal". In addition to these search terms, individual risk factors also were included in the search terms (eg, "age," "smoking," and "drinking"). Reference lists of articles were reviewed to get more potentially eligible articles.

| Inclusion criteria and exclusion criteria
We selected articles that (a) were observational epidemiologic study (case-control and cohort study), (b) examined the association between any paternal exposures (eg, paternal age, paternal smoking, paternal drinking, paternal occupation, and paternal exposure to chemical agents) and CHDs overall or any one of the CHD subtypes in infants, (c) were written either in English, Chinese, or French, and (d) reported ORs (ie, risk ratios [RR] or odds ratios [OR]) and associated 95% confidence intervals (CIs) or had raw data available.
The exclusion criteria were: articles that (a) did not examine the association between any paternal exposures and any CHD subtypes in infants, (b) did not reported ORs and associated 95% CIs or had no raw data available, and (c) we could get the full text.
In the case of multiple publications using the same database, we selected the study that contained the most comprehensive information (eg, longest study periods or most CHD subtypes analyzed).

| Data extraction
The studies meeting the inclusion criteria were independently reviewed by two authors (JP, JW) to extract study characteristics (eg, authors, year of publication, geographic region, periods of data collection, study design, sample size, exposure data, exposure period around pregnancy) and measures of association (eg, OR, RR). Measures of association not available in the original article were calculated based on raw data. Discrepancies between the authors were resolved by discussion.

| Statistical analysis
We tested for heterogeneity across studies using Cochran's Q-test. If there was an evidence of heterogeneity (P < .1), we used a random- -no report about OR and 95%CI and no available data to calculate OR and 95%CI(N = 5) -not found (N = 4) -Multiple publications of same data (N = 1) -language in Lithuanian (N = 1) F I G U R E 1 Flow chart of study selection process. CI, confidence intervals; OR, odd ratios Subgroup analysis was performed based on the different paternal factor, and sensitivity analysis was conducted. Publication bias was evaluated visually by funnel plots.
Four studies were excluded since we had no access to full text.
One study was only available in Lithuanian and was excluded.
Finally, we included 31 studies for the meta-analysis and systematic review.

| Study characteristics
The characteristics of the included studies were summarized in
In addition, eight studies categorized paternal age into different age groups and we summarized the same age group, namely, <20, 20 to 24, 25 to 29, 30 to 34, 35 to 39, and ≥40 years of age. As shown in Table 2

| Paternal wine drinking
Seven studies 6,11,14,15,17,20,21 evaluated the effect of paternal alcohol consumption on CHDs. The summarized OR was 1.47 (1.05, 2.07) ( Table 2, TABLE S3), indicating that paternal alcohol intake was a risk factor for CHDs in the offspring. However, the definition of paternal wine drinking was various from studies. The most common definition was defined by drinking capacity, that is, wine drinking mean a reported alcohol intake of on average at least 50 mL per day or per time without specifying wine. 6,11,15,20,21 Others defined wine drinking by the amount of wine categories. 11,14 Only one study did not specify the definition of wine drinking. 17 3.6 | Paternal exposure to chemical agents or drugs Seven studies 3,11,14,15,20,22,23 evaluated the effect of paternal exposure to chemical agents or drugs on CHDs. These toxic chemical agents including pesticides, polychlorinated compounds, phthalates, alkyl phenolic compounds, bisphenol A, heavy metals, 15 hydrocarbons, 3 marijuana, and cocaine. 11 After meta-analysis, we found that paternal exposure to chemical agents or drugs had a strong association with increased risk of CHDs (OR = 2.15 [1.53, 3.02]) ( Table 2, TABLE S4).

| DISCUSSION
More and more evidence showed that not only maternal factors but also some paternal factors were associated with increased risk of CHDs. Nevertheless, there was little review or meta-analysis focused on the non-genetic paternal factors for CHDs, and our study made up this blank. We analyzed almost all the current literature and made a relatively comprehensive summary about the non-genetic paternal factors for CHDs. After subgroup analysis, we found that advanced paternal age, cigarette smoking, wine drinking, some occupations, and exposure to chemical agents and drugs were still associated with the increased risk of CHDs.
Advanced paternal age was previously found to be associated with increased DNA mutations and chromosomal aberrations in sperm. 37 Genetic changes in sperm associated with advanced paternal age could lead to an increased risk for birth defects in offspring. 10 Consistent with these findings, we found that advanced paternal age (≥35 years) was associated with increased risk of CHDs. On the contrary, paternal age between 25 to 29 years decreased the risk of CHDs. This suggested that a certain reproductive age might be helpful to reduce the prevalence of CHDs, which could help to provide evidence for governmental health policy. In addition, these conclusions still need further cohort studies with larger sample to confirm.
Cigarette smoking is a well-known teratogenic risk factor for birth defects and it can affect a number of developing structures. 35 Nicotine, the main toxic agent during smoking, could affect sperm activity greatly and lead to chromosome aberration, which might affect the fetal development, and result in the occurrence of cardiac malformations. 38 Besides, paternal smoking could induce maternal passive smoking, which could also increase the risk of CHDs. 39 Consistently, Deng  performed. The probable reason might be the distinct subtypes of CHDs, which could obscure findings when subtypes were "lumped" into a common phenotype to increase study power. For publication bias, the funnel plot of smoking subgroup showed asymmetry, which indicated publication bias. However, for the rest of subgroups, the funnel plots were basically symmetric.

| CONCLUSIONS
In conclusion, we summarized all the articles about non-genetic paternal risk factors for CHDs and found that advanced paternal age, cigarette smoking, wine drinking, some occupations, and exposure to chemical agents and drugs would increase the risk of CHDs. It is important and urgent to encourage fertility at certain age, building a healthy life habit, beginning with quitting smoking and drinking, and trying to avoid occupational and environment exposures.