In the present study, we analyzed the association of VDR gene SNPs rs154410, rs7975232, rs731236, rs2228570 and rs739837 with GDM in Wuhan, China. It was revealed that, VDR gene polymorphic markers were not found to be associated with GDM in central Chinese population. Furthermore, there were no gene-gene interactions on the GDM risk among the examined VDR gene SNPs.
The rs739837 SNP is located at the three-primer untranslated region (3′-UTR) of the VDR gene. This region does not affect amino acid sequence and is not likely to affect the function of the gene . However, the variant rs739837 might affect the expression of VDR gene by binding with microRNA . Several studies investigated the relationship between this locus and T2D and reported that rs739837 was associated with susceptibility to T2D [15, 17, 24]. To date, two studies had studied the relationship between rs739837 and GDM. Shi et al. reported no relationship between the genotypic model of rs739837 and GDM, whereas Wang et al. found a statistical correlation between the rs739837 polymorphism and GDM risk[26, 27]. However, neither of the two studies had analyzed the combined effect with other VDR gene SNPs. As susceptibility was attributable not to a single polymorphism or allele, but rather to multiple polymorphisms ,we evaluated rs739837 and other four widely studied VDR gene SNPs and their gene-gene interactions on the risk of GDM. The result showed that there was no statistical correlation between rs739837 and GDM. Besides, no evidence was found at the gene-gene interactions. To our knowledge, this study is the first to investigate the combined effect between rs739837 and other VDR gene SNPs on the risk of GDM. The results need to be verified in future studies.
The rs1544410, rs7975232, and rs731236 SNPs are known as BsmI, ApaI and TaqI according to their restriction enzymes. A meta-analysis on the three polymorphisms with the risk of T2D produced negative results. As for the association between the three polymorphisms and GDM, the results were inconsistent [18, 19, 22, 30]. Our study reported a negative result, which was consistent with the finding of Apaydin et al. BsmI, ApaI and TaqI are all located in the 3′-UTR and have been shown to be in strong linkage disequilibrium . Polymorphisms ofTaqI,BsmI and ApaI are probably non-functional because they are either located in intron (BsmI and ApaI in intron 8), which will be removed during mRNA post transcriptional modification or result in no amino acid sequence change (TaqI in exon 9).
The rs2228570 is known as FokI according to its restriction enzyme. FokI polymorphism was linked to risk of GDM in Turkish women and Iranian population [18, 31]. However, studies in other countries could not establish association between FokI and GDM [19, 20]. In the present study, we reported no evidence of allelic or genotypic association of the FokI polymorphism with GDM in central Chinese population. The FokI is located at the 5′ end region of the VDR gene. It is reported as an independent marker of the VDR gene because it has not been shown to be in linkage disequilibrium with any other VDR polymorphisms . It produces either a 424 or a 427 amino acid VDR protein. These two isoforms are thus structurally distinct, unlike those VDR gene that contain polymorphisms present in the 3′-portion of the gene that are either silent codon changes or are found in introns or in the 3′-untranslated regions . Even the relationship between the FokI polymorphism and T2D is still controversial [14–16, 29], probably because the socio-demographic characteristics, experimental methods, and sample sizeare different in the studied populations.
Our study had several strengths. First, we employed MDR method to explore the gene-gene interactions on the GDM risk among the selected SNPs. The identification and characterization of gene-gene interactions had been limited mainly by a lack of powerful statistical methods and a lack of large sample size . To overcome these limitations, the MDR method was developed. It was used for detecting and characterizing high-order gene to gene interactions  and was shown to have good power in relatively small case-control studies [23, 35]. Second, we adjusted potential confounding factors such as age, pre-pregnancy BMI etc. to explore the association in different genotype models. At last, we used a relatively large sample size, which was able to provide enough statistical power.
However, there were also some limitations in this study. First, the level ofplasma vitamin D was not measured in all subjects. Second, the information of environmental and lifestyle factors was lacked, which had been reported recently to be important determinants of GDM development [36, 37]. Finally, there were some potential biases that came from the cross-sectional nature of the case–control study. Thus, cohort studies concerning the above-mentioned factors will be required in future to validate the findings of the study.