Introduction

Congenital heart diseases or defects (CHD) which share a significant proportion in CVD burden arises due to incomplete development of heart during the first 6-weeks of gestation [1]. The origin of CHD is diverse which can be associated with a syndrome or be isolated (non-syndromic). It is hypothesized that susceptibility of cardiac defects increases with dual interaction of key gene(s)/SNP-environmental factors which perturb normal cardiac developmental process during embryonic life. The risk of CHD is greatly influenced by variants within the genes involved in folate-homocysteine metabolism [2,3,4]. Many studies have revealed that the risk of CHD in new-borns of females carrying mutations in genes involved in folate metabolism can be reduced by maternal periconceptional use of multivitamins or folic acid [5], however, the mechanism underlying this effect is still under investigation. Folate and vitamin B12 are known to influence homocysteine concentration. Folates taken in diet are usually polyglutamates which are converted to simpler forms, particularly monoglutamates, dihydrofolate, tetrahydrofolate and finally to methylated form of folate i.e. 5, 10-methylenetetrahydrofolate (5,10-MTHF) and 5-methyltetrahydrofolate (5-MTHF) by a specialised enzyme of the pathway. Homocysteine and folate metabolism is dependent on a couple of genes performing their specific role but two genes namely MTHFR and MS are considered critical genes for development of diseased cardiovascular phenotypes. A common mutation, C677T (rs1801133), in exon 4 of the MTHFR gene results in decreased enzyme activity and contributes to increased plasma homocysteine, particularly in individuals with low folate status. Rady and co-workers reported a novel polymorphic site of the MTHFR gene at nucleotide position 1793 G to A transition in exon 11 (rs2274976) which results an arginine-to-glutamine change at codon 594 and modifies enzyme activity [6]. The A2756G mutation (rs1805087) in MS gene alters re-methylation process and is also associated with increased homocysteine levels and risk of CHD. Most of the research in relation to folate-homocysteine metabolising pathway with the risk of CHD is based on parent-of-origin effect. There are very few studies focussing on embryonic variation in candidate genes of folate-homocysteine metabolising pathway in association with the development of structural congenital heart malformations during early pregnancy. Consistent with this view, we attempted to perform a case–control pilot study involving evaluation of two important genes: MTHFR (C677T and G1793A) and MS (A2756G) gene variations with risk of CHD in Jammu region of UT of J&K, India. Further, we also performed an updated meta-analysis with trial sequential analysis to investigate the association between MTHFR (C677T and G1793A) and MS (A2756G) polymorphisms and risk of CHD with increased statistical power.

Methodology

Study population and area

The present study was ethically approved by Institutional Ethical Committee, University of Jammu. The present study was carried out on 150 children, out of whom 50 children (0–12 years) were confirmed cases of CHD and 100 children (below 18 years) were unrelated healthy controls belonging to Jammu region of Union Territory of Jammu and Kashmir. The CHD cases were enrolled from In-patient Department of Paediatrics whereas controls were recruited from Out-patient Department of Paediatrics, Shri Maharaja Gulab Singh (SMGS) hospital, Jammu. Data and blood collection was done after having an informed written consent from attendant or guardian of the children. The diagnosis and classification of CHD was based on the clinical and the echocardiography findings. The inclusion/exclusion criteria were followed wherein patients with any form of CHD were included whereas patients with syndromes and neural tube defects were excluded. Controls admitted to hospital for minor ailments with no history of CHD or other major abnormality and also children visiting for blood typing were recruited for the study under reference. Power of the study for sample size calculation was done by using online tool based on mean and standard deviation of two groups of study subjects, two tail test and with alpha value of 5% (https://www.sphanalytics.com/statistical-power-calculator-using-average-values/). The power of the study obtained was more than 80%.

Blood collection and DNA isolation

500 μl-1 ml of blood was collected in EDTA coated vacutainers from each child by trained paramedical staff of the Hospital. Isolation of DNA from whole blood was carried out using commercially available kits (DNeasy Blood and Tissue Kit, QIAGEN). The quantitative and qualitative analysis of isolated DNA was performed by spectrophotometry and 1.5% agarose gel electrophoresis respectively.

Genotyping

Genotyping was performed by polymerase chain reaction-restriction fragment length polymorphism (PCR–RFLP) technique. Briefly, PCR was carried out in a reaction volume of 25 μl each in thin walled tubes, consisting of 5.0 μl of PCR buffer (10X), 2.5 μl of MgCl2 (25 mM), 0.5 μl of dNTPs (10 mM), 0.5 μl (100 pmol/µl) of each of the forward and reverse primers, 0.3 μl (5unit/μl) of Taq DNA polymerase enzyme and 2 μl (40 ng) of genomic DNA. PCR amplification was carried out using the Veriti, Applied Biosystems by life technology, Singapore and amplification and RFLP conditions for all the three polymorphisms are given in Table 1. The gel images of PCR–RFLP for MTHFR (C677T and G1793A) and MS (A2756G) polymorphisms with band sizes have been depicted in Fig. 1, 2 and 3 respectively.

Table 1 Details of Primer sequence, amplification conditions and restriction enzymes
Fig. 1
figure 1

RFLP Gel Image of MTHFR C677T Polymorphism. L: 100 bp Ladder; Lane 1,2,3,5,6,8: 198 bp (CC: Wild); Lane 4,7: 198 + 175 + 23 bp (CT: Hetero). *23 bp band will not be visible on agarose gel

Fig. 2
figure 2

RFLP Gel Image of MTHFR G1793A Polymorphism. L: 100 bp Ladder; Lane 3,4,6,7,8: 233 + 77 bp (GG: Wild); Lane 1,2,5: 310 + 233 + 77 bp (GA: Hetero)

Fig. 3
figure 3

RFLP Gel Image of MS A2756G Polymorphism. L: 100 bp Ladder; Lane 1,2,6,7: 211 bp (AA: Wild); Lane 2,4,5: 131 + 80 + 211 bp (AG: Hetero)

Statistical analyses

Genotypic frequency as well as allelic frequency was calculated by gene counting method. Hardy–Weinberg equilibrium (HWE) analysis and the differences in genotypic frequencies between two study groups were examined by using Pearson’s goodness of fit Chi-square test. To assess the association, odds ratios (OR) with 95% CI were calculated under different genetic models by using Statistical Package for Social Sciences (SPSS-version 20) software and also by another method provided by the Institute of Human Genetics accessed via the link: http://ihg.gsf.de/cgi-bin/hw/hwa1.pl. A p-value of < 0.05 was considered as statistically significant.

Meta-analysis

Literature search

Research papers (published up to February, 2021) examining the association between MTHFR C677T, MTHFR G1793A and MS A2756G polymorphisms and congenital heart defects were extracted from databases such as PubMed, Science direct, Proquest, Ovid and Google Scholar. Key words used for the database search were as follows: methylenetetrahydrofolate reductase; MTHFR gene polymorphisms; Methionine synthase; MS/MTR gene polymorphisms; Congenital heart defects; Congenital heart diseases; MTHFR C677T; MTHFR G1793A and MS/MTR A2756G. Reference records of studies included in our meta-analysis were manually searched for possible eligible articles.

Inclusion and exclusion criteria

The inclusion/exclusion criteria used for screening of eligible study are given in Table 2.

Table 2 Inclusion/Exclusion criteria for eligible studies

Data extraction and quality assessment

From each eligible study, the following data were extracted by the two investigators independently using a standardized form: first author, publication year, country of origin, ethnicity, number of cases and controls, genotype frequency, source of controls, genotyping method, and Hardy–Weinberg equilibrium (HWE). We investigated the quality of each study based on the nine-point Newcastle–Ottawa Scale (NOS). The characteristics and results of NOS for all the included studies are shown in Table 3. The NOS scores for all eligible studies in this Meta analysis exceeded 6 points, indicating that our analysis is updated and is of good quality.

Table 3 Characteristics of the included studies in the meta-analysis

Statistical analysis for meta-analysis

The association between the selected polymorphisms and congenital heart defects was evaluated for each study by the crude odds ratios (ORs) with 95% confidence intervals (CIs). For each study, HWE was assessed by the chi-square goodness of fit test. For all studies, we estimated the association under three different genetic models [Allele contrast, dominant model and recessive model]. Statistical heterogeneity between studies was assessed by Cochran’s Q test and I-square (I2) > 50% indicated the significance [31]. When I2 > 50%, a random-effect model should be taken otherwise fixed model is used. To calculate the OR and draw inference for each study, we used both random effects model and fixed effect model. Sensitivity analyses were conducted by omitting any single study, which predisposed the observed heterogeneity excessively and there should be no change in OR’s. Egger’s test and Begg’s funnel plot is used to solve the problem of Publication bias. All statistical analyses were performed in the MetaGenyo online Statistical Analysis System software [32].

Trial sequential analysis (TSA)

Meta-analysis may result in Type I error owing to an increased risk of random errors (play of chance) which can be due to dispersed data and repeated significance testing. Bias from low trial with low methodological qualities, publication bias and small trial bias may result in false p-value. Trial Sequential analysis is a methodology that can be used in meta-analysis to control random errors, and to assess whether the studies included in the meta-analysis have surpassed the requisite sample size. TSA was performed to calculate the required information size on the basis of overall 5% risk of Type-I error and a power of 80% for checking the reliability of meta analysis [33].

Results

Case–control study

Based on echocardiography reports, the different CHD phenotypes were categorised (Table 4). The observed prevalence of different CHD phenotypes in present study was highest for ventricular septal defect (VSD: 34%) and atrial septal defect (ASD: 26%) followed by tetralogy of fallot (TOF: 14%) and patent ductus arteriosus (PDA: 8%) and least for endocardial cushion defect (6%). The frequency of complex CHD forms (more than one CHD condition) were as follows: 4% for ASD with PDA, 2% for VSD with AV-canal defect, 4% for VSD with pulmonary arterial hypertension (VSD-PAH) and 2% for endocardial cushion defect along with dextrocardia.

Table 4 Prevalence of CHD phenotypes in present study

The genotypic and allelic frequencies along with Chi square values for Hardy–Weinberg calculations for the all the three polymorphisms in study participants are depicted in Table 5. There observed frequencies of genotypes were in concordance with HWE in both the groups for all the polymorphisms except for MTHFR C677T in patient group. The genotypic frequency of CC, CT and TT (MTHFR C677T) in CHD patients was 88%, 8% and 4% whereas in controls it was 90%, 9% and 1% respectively. The frequency of variant allele T (0.08) was higher in CHD patients than controls (0.05) whereas wild allele C was reported to be in slightly higher frequency in controls (0.95) as compared to patients (0.92). The genotypic frequencies for MTHFR G1793A in CHD patients were 58%, 38% and 4% for GG, GA and AA respectively. The frequencies in control group were 90% for GG and 10% for GA genotypes; however we did not observe any AA genotype in controls. In general there was higher frequency of risk allele ‘A’ in CHD patients (0.23) in comparison to controls (0.05). The distribution of observed MS genotypes in CHD patients were 60%, 36% & 4% for AA, AG and GG genotypes respectively. In control group the distribution was as follows: 73% for AA, 26% for AG and 1% for GG genotype. The CHD patients were showing higher frequency of risk allele ‘G’ (0.22) than controls (0.14).

Table 5 Showing genotypic and allelic distribution of selected gene polymorphisms among cases and controls

In order to investigate the possible association of these three polymorphisms with susceptibility of CHD, ORs with 95% confidentiality intervals was calculated for different genetic models which are presented in Table 6.

Table 6 Association between selected gene polymorphisms and CHD

For both MTHFR C677T and MS A2756G polymorphisms, we observed that even though the values calculated for ORs under different models were above 1, but none of the values reached statistical significance level (p > 0.05). The present study proclaimed lack of association of MTHFR C677T and MS A2756G gene polymorphism with the risk of CHD in our population. Furthermore, the GA vs GG genotype depicted a strong significant association of MTHFR G1793A gene polymorphism. The G vs A frequency showed that the allele ‘A’ is adding a significant risk of approximately 5.7 folds in the development of CHD in the studied population. Distribution of MTHFR haplotypes in cases & controls and their association towards CHD susceptibility is depicted in Table 7.

Table 7 Association of MTHFR haplotypes with risk of CHD

The frequency of C-G haplotype was higher in both cases and controls (0.690 & 0.895 respectively). There was complete absence of T-A haplotype in both study groups. The haplotype combination C-A was significantly associated with CHD risk (OR = 5.67 [2.58–12.48], p = 2.71e‐006) and C-G was significantly involved in protection against CHD development (OR = 0.26 [0.14–0.48], p = 1.00e‐005) in the population under reference. By analysing LD scores in two study groups it was observed that the MTHFR variants were in complete LD in both patients (D' = 0.999, r2 = 0.026) and controls (D' = 1, r2 = 0.003).

Meta-analysis

We found 26 eligible studies having 3450 cases and 4447 controls with reference to MTHFR C677T polymorphism and 6 studies with 697 cases and 777 controls concerning MS A2756G polymorphism. The main study characteristics are summarized in Table 3. The study selection process has been depicted in PRISMA diagram (Fig. 4). By pooling all the studies, it was found that there is statistically significant association between MTHFR C677T polymorphism and congenital heart defects under all applied genetic models (Dominant model: OR = 1.38, 95% CI: 1.14- 1.69; recessive model: OR = 1.49, 95% CI: 1.83–1.87; allele model: OR = 1.33, 95% CI: 1.14–1.55) as shown in Table 8 and Fig. 5, 6, and 7. When we stratified the studies according to ethnicity, a significant association was observed between this locus and CHD only in Asian populations (Dominant model: OR = 1.50, 95% CI: 1.12- 2.01; recessive model: OR = 1.67, 95% CI: 1.21–2.31; allele model: OR = 1.42, 95% CI: 1.15- 1.76), but not in Caucasian populations (dominant model: OR = 1.24, 95% CI: 0.95- 1.62; Recessive model: OR = 1.27, 95% CI: 0.99–1.63; allele model: OR = 1.21, 95% CI: 0.97–1.50) as given in Table 8.

Fig. 4
figure 4

PRISMA flow diagram showing the selection of various studies for the meta –analysis

Table 8 Overall meta-analysis and subgroup analysis by ethnicity for MTHFR C677T polymorphism
Fig. 5
figure 5

Pooled OR (Dominant model) and 95% CI for individual studies and pooled data for the association between the polymorphism C677T and congenital heart disease (CHD) in the overall population

Fig. 6
figure 6

Pooled OR (Recessive model) and 95% CI for individual studies and pooled data for the association between the polymorphism C677T and congenital heart disease (CHD) in the overall population

Fig. 7
figure 7

Pooled OR (Allele model) and 95% CI for individual studies and pooled data for the association between the polymorphism C677T and congenital heart disease (CHD) in the overall population

However, it was observed that Caucasian population was also showing association but it did not reach statistical significance. For MS polymorphism, none of the applied genetic models found association with CHD in overall population or even after subgrouping (Table 9 and Fig. 8). Sensitivity analysis for both MTHFR and MS revealed that there is no change in the pooled ORs by omitting individual studies (Fig. 9 and 10). The publication bias was also estimated by using funnel plot for log-odds ratio for dominant model against the reciprocal of its standard error (Fig. 11 and 12). Further Egger regression asymmetry test was also used to evaluate publication bias (Table 9). No publication bias was observed in the present meta-analysis. Meta- analysis could not be performed for MTHFR G1793A gene polymorphism as we were able to find only one study other than the study under reference. Meta-analysis could not be performed for MTHFR G1793A gene polymorphism as we were able to find only one study other than the study under reference.

Table 9 Overall meta-analysis and subgroup analysis by ethnicity for MS A2756G polymorphism
Fig. 8
figure 8

Pooled OR (Dominant model) and 95% CI for individual studies and pooled data for the association between the polymorphism MS/MTR A2756G and congenital heart disease (CHD) in the overall population

Fig. 9
figure 9

Sensitivity analysis of association between MTHFR C677T polymorphism and CHD

Fig. 10
figure 10

Sensitivity analysis of association between MS/MTR A2756G polymorphism and CHD

Fig. 11
figure 11

Funnel plot of the MTHFR C677T polymorphism and susceptibility to CHD (Dominant model) in the overall population

Fig. 12
figure 12

Funnel plot of the MS/MTR A2756G polymorphism and susceptibility to CHD (Dominant model) in the overall population

Trial Sequential Analysis (TSA)

Trial sequential analysis was performed to calculate the requisite sample size for the meta-analysis of MTHFR C677T gene polymorphism. It revealed that sufficient number of studies have been included in the meta-analysis of this polymorphism. The results of TSA were in accordance with the findings of the conventional meta-analysis and revealed that C677T polymorphism was significantly associated with the risk of CHD (Fig. 13). For MS A2756G polymorphism, TSA could not be performed owing to very little information of sample size which revealed that there is need of more replicas of case control studies to reach the conclusive remarks on role of said polymorphism in conferring risk of CHD. Similarly for MTHFR G1793A gene polymorphism, TSA could not be performed as only two studies were available for meta-analysis.

Fig. 13
figure 13

Trial Sequential Analysis (TSA) of the studies included in the meta analysis of MTHFR C677T gene polymorphism with CHD

Discussion

The folate-homocysteine metabolic pathway performs a paramount role in neural tube formation and cardiac development during embryogenesis. Low folate and high homocysteine levels are a closely related with the manifestation of congenital heart defects, which indicates that single nucleotide polymorphisms (SNPs) in the genes controlling this pathway may be the genetic risk factors for these disorders [34]. Therefore, we performed a case–control association study and an updated meta-analysis along with TSA to investigate the association of MTHFR and MS gene polymorphisms with risk of CHD. We did not find a significant association of MTHFR C677T and MS A2756G polymorphism with risk of CHD in our studied population. The results were consistent with studies done by various workers [5, 7, 35,36,37,38]. Regarding MTHFR G1793A polymorphism in link with CHD risk we found significant association under co-dominant, dominant and allelic model in present study. The genotypic frequencies reported in the present study were almost compatible with frequencies as reported by Toganel and co-workers and the investigators also observed a strong significant association this SNP with susceptibility of CHD [AA + GA vs GG: OR = 4.18; 95% CI (1.25- 13.98), p = 0.02] in a Romanian population whereas antithetical findings were reported in Chinese population [39, 40]. Xu and co-workers found that the variant genotypes of MTHFR G1793A polymorphism were significantly associated with a decreased risk of CHD, especially in patients with isolated peri-membranous VSD [40]. The correlation between the MTHFR G1793A gene polymorphism and the CHD risk has not been extensively studied so far. To the best of our knowledge there is no previous report from India and we are the first to analyse G1793A variation of MTHFR gene from North India. The present study is first of its kind concentrating on the effect of MTHFR (C677T and G1793A) haplotypes with vulnerability of CHD. The haplotype C-A was conferring nearly 5.7-fold disease risk and C-G haplotype was giving a shielding outcome of approximately 3.8-fold (1/0.26). Based on measure of LD, the two MTHFR SNPs were in complete LD in both CHD cases and controls. The possible limitations of the present study may be the enrolment of study samples from single region of UT J&K and lack of homocysteine measurements in the study subjects. Besides these limitations and to the best of our knowledge, the study under reference here is the first attempt that evaluates the association of MTHFR and MS gene polymorphisms in CHD.

Genetic association studies have been a powerful approach for identifying susceptibility genes for common diseases but it has been experienced that most of the initial positive associations were not reproduced in the subsequent replication studies because of small sample size or false-positive reports [41, 42]. Meta-analysis solves this problem as it increases the statistical power to detect gene–disease associations by combining results from the original and subsequent replication studies [42]. Similarly, when we conducted case–control association, we did not observe significant association of MTHFR C677T with risk of CHD, as it was a pilot study and carried on limited number of samples. But after performing meta-analysis, the results suggested a positive association of MTHFR C677T with the risk of CHD. The results of the overall analysis depicted an increased risk of CHD with the presence of MTHFR 677 T- allele in fetus. The putative risk allele-677 T had a 1.33 folds increased risk of CHD against the C-allele. From the subgroup analysis, the increased risk of the T-allele was widely detected in Asians but not in Caucasians. Our results are compatible with the previous Meta analyses that investigated the association of the MTHFR C677T polymorphism in CHD [34, 43]. Further, this association revealed through conventional meta-analysis has also been confirmed by performing Trial Sequential Analysis. Lack of association was reported for MS A2756G both in pooled and in sub-grouped meta-analysis and the findings are consistent with study done by Cai and co-workers [44]. The findings of MS polymorphism needs to be further investigated as there are not enough studies on association of this polymorphism with risk of CHD and during our search we also found only six eligible studies and TSA has not been performed in lieu of lack of sufficient number of studies. Further, we were not able to perform meta- analysis for MTHFR G1793A polymorphism as to best of our efforts; we found only a few case–control studies which were not sufficient for performing meta-analysis.

Conclusion

In conclusion, the results of meta-analysis and TSA support the role of MTHFR C677T gene polymorphism as susceptibility factor for Congenital Heart Defects. For MTHFR G1793A and MS A2756G gene polymorphisms, there is need to perform large number of homogenous studies to evaluate these crude results further.