Vitamin D receptor gene polymorphisms and the risk of the type 1 diabetes: a meta-regression and updated meta-analysis

Background The association between the polymorphisms in the vitamin D receptor (VDR) gene and the risk of type 1 diabetes mellitus (T1DM) has been evaluated in several studies. However, the findings were inconclusive. Thus, we conducted a meta-analysis to comprehensively evaluate the effect of VDR gene polymorphisms on the risk of T1DM. Methods All relevant studies reporting the association between VDR gene polymorphisms and susceptibility to T1DM published up to May 2020 were identified by comprehensive systematic database search in ISI Web of Science, Scopus, and PubMed/MEDLINE. Strength of association were assessed by calculating of pooled odds ratios (ORs) and 95% confidence intervals (CIs). The methodological quality of each study was assessed according to the Newcastle–Ottawa Scale. To find the potential sources of heterogeneity, meta-regression and subgroup analysis were also performed. Results A total of 39 case–control studies were included in this meta-analysis. The results of overall population rejected any significant association between VDR gene polymorphisms and T1DM risk. However, the pooled results of subgroup analysis revealed significant negative and positive associations between FokI and BsmI polymorphisms and T1DM in Africans and Americans, respectively. Conclusions This meta-analysis suggested a significant association between VDR gene polymorphism and T1DM susceptibility in ethnic-specific analysis.

Vitamin D (VitD) is a steroid molecule that has many roles in the body, such as regulation of the immune cells. In addition to immune responses, VitD is also involved in the etiopathogenesis of several disorders, such as cancer, autoimmune disorders, cardiovascular disorders, asthma, and diabetes [6][7][8][9]. In animal model of T1DM, VitD suppresses the occurrence of diabetes, by regulating the T helper (Th) 1/Th2 cytokine balance in the local pancreatic lesions [10,11]. Moreover, VitD inhibits T cell activation and secretion of proinflammatory cytokines, such as interleukin (IL)-1, IL-6, IL-12, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ, which are involved in the pathogenesis of T1DM [12][13][14]. Mostly, VitD exerts its function through vitamin D receptor (VDR), which is found in the nuclei of target cells, such as lymphocytes, macrophages, and pancreatic cells. VDR is a member of the nuclear hormone receptors superfamily and has been linked to insulin sensitivity and secretion [15].
Four common single nucleotide polymorphisms (SNPs) of VDR gene are FokI (rs2228570), TaqI (rs731236), BsmI (rs1544410), and ApaI (rs7975232). Among them, ApaI, BsmI, and TaqI polymorphisms are located in the 3′-end of VDR gene which lead to silent mutation associated with increased VDR mRNA stability. In contrast, FokI SNP is located in the start codon that produces a protein with shorter size (424 amino acids), which is more active than the long form (427 amino acids) [8,16,17]. Over the course of past few decades, the VDR gene polymorphisms have been associated with susceptibility to numerous autoimmune disorders [8,18,19].
In recent years, several studies have investigated the association between VDR gene SNPs and T1DM in all over the world, which have yielded conflicting results. The reasons for these discrepancies might be small sample sizes, clinical heterogeneity, and low statistical power. Therefore, a comprehensive meta-analysis might be the best way to solve these problems. Two previous meta-analyses performed by Tizaouia et al. in 2014 [20] and Guo et al. in 2006 [21] reported that VDR gene polymorphisms were not associated with the susceptibility to T1DM. However, Zhang et al. in 2012 [22] demonstrated that BsmI polymorphism was significantly associated with the risk of T1DM. Furthermore, Sahin et al. in 2017 indicated that BsmI and TaqI polymorphisms were associated withT1DM risk in children with less than average 15 years old [23]. Qin et al. in 2014 evaluated the association of only BsmI SNP with T1DM risk and demonstrated its association in the overall analysis, as well as in Asians, Latino, and Africans [24]. In 2014, Wang et al., by including 20 studies, reported that BsmI polymorphism might be a risk factor for susceptibility to T1DM in the East Asian population, and the FokI polymorphism was associated with an increased risk of T1DM in the West Asian population [25].
Since several articles published after the last metaanalysis, here we conducted an updated meta-analysis with the aim of providing a much more reliable conclusion on the significance of the association between VDR gene polymorphisms and T1DM risk.

Methods
This meta-analysis was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines, including search strategy, inclusion and exclusion criteria, data extraction and quality assessment, and statistical analysis [26].

Search strategy
Three electronic databases (PubMed/MEDLINE, Scopus, and Web of Science) were systematically searched for studies regarding the association of VDR gene polymorphisms, including FokI (rs2228570) and/ or TaqI (rs731236) and/or BsmI (rs1544410) and/or ApaI (rs7975232), and T1DM susceptibility, which were published before May 2020. The following combinations of search terms were used: ("T1D" OR "type 1 diabetes" OR "diabetes") AND ("VDR" OR "vitamin D receptor") AND ("polymorphisms" OR "SNP" OR "variation" OR "mutation"). The reference lists of review articles were also manually searched for additional pertinent publications. Original data in English language and human population studies were collected.

Inclusion and exclusion criteria
Eligible studies must meet the following criteria: a) All studies assessing the association of VDR gene polymorphisms and T1DM risk; b) All studies reporting sufficient data to calculate the odds ratio (OR) and its 95% confidence intervals (CIs); c) All studies with distinct case and control groups (case-control and cohort design). The exclusion criteria were: a) studies that their genotype or allele frequency could not be extracted; b) letters, non-English publications, animal studies, case reports, reviews, comments, book chapters, and abstracts; c) duplicate and republished studies. The application of these criteria recognized 39 studies eligible for the quantitative analysis.

Data extraction and quality assessment
According to a standardized extraction form, the following data were independently extracted by two reviewers: the author's name, journal and year of publication, country of origin, ethnicity, number of case and control for each gender separately, genotype and allele frequencies in cases and healthy groups, mean or range of age, genotyping method, total sample size of cases and controls. The third reviewer finalized the extracted data, and potential discrepancies were resolved by consensus. For quality assessment of the included publications, the Newcastle-Ottawa Scale (NOS) was applied [27]. In this respect, studies with 0-3, 4-6 or 7-9 scores were, respectively, of low, moderate, and high-quality.

Statistical analysis
Deviation from Hardy-Weinberg equilibrium (HWE) for distribution of the allele frequencies was analyzed by χ2test in control groups.  [28]. Accordingly, heterogeneity was considered significant if I 2 values exceeded 50% or the Q statistic had a P value of less than 0.1 and random-effects model (DerSimonian-Laird approach) was carried out [29]. Otherwise,     the fixed-effects model (Mantel-Haenszel approach) was performed for combination of data [30]. In order to assess the predefined sources of heterogeneity among included studies, subgroup analysis and metaregression analysis based on year of population, and ethnicity were performed. Stability of our results was assessed by sensitivity analysis. Potential publication bias was estimated by Egger's linear regression test, and also Begg's test was employed to estimate the funnel plot asymmetry (P value< 0.05 considered statistically significant) [31,32]. The data analyses were carried out using STATA (version 14.0; Stata Corporation, College Station, TX) and SPSS (version 23.0; SPSS, Inc. Chicago, IL).

Study characteristics
Regarding to aforementioned keywords, a total of 1116 studies were initially retrieved. Of these studies, 456 publications were duplicate, 559 and 62 publications excluded by title & abstract and full text examination, respectively. Finally, 39 studies qualified for quantitative analysis. It should be noted that while the latest meta-analysis by Tizaouia et al. [20] in 2014 included 23 studies, we performed the updated meta-analysis by adding 16 more articles. Also, no studies were found by hand search (Fig. 1). The eligible studies were published from 1998 to 2019 and had an overall good methodological quality with NOS scores ranging from 6 to 8. Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) and Taq-man were used by majority of included studies as genotyping method. Tables 1 and 2 summarized the characteristics and genotype frequency of the included studies.

Quantitative synthesis
Meta-analysis of the association between FokI (rs2228570) polymorphism and T1DM risk  (Fig. 2). The results of pooled ORs, heterogeneity tests and publication bias tests in different analysis models are shown in Table 3.    [52,65] and one study each was in Australia [50] and Americans [61]. Meta-analysis rejected any significant association between TaqI SNP and the risk of T1DM susceptibility. Moreover, the results of subgroup analysis by ethnicity were not significant under five genotype models. In subgroup analysis, since there was only one study for the Australians [50], Americans [61], and two studies for Africans [52,65], these studies were excluded from the analysis. The results of pooled ORs, heterogeneity tests and publication bias tests in different analysis models are shown in Table 3.

Meta-analysis of the association between BsmI (rs1544410) polymorphism and T1DM risk
To examining the association between BsmI polymorphism and T1DM risk, 34 case-control studies with 4826 cases and 7159 controls subjects were included. It was detected that 15 studies with 1938 cases and 4450  (Fig. 3). No significant association was detected for European, Asian and African population. The results of pooled ORs, heterogeneity tests and publication bias tests in different analysis models are shown in Table 3.

Meta-analysis of the association between ApaI (rs7975232) polymorphism and T1DM risk
Finally, 24 case-control studies with 2436 cases and 4074 controls were identified eligible for quantitative synthesis of the association between ApaI polymorphism and T1DM risk. Overall, 10 studies were conducted in Europe [15,35,36,38,41,44,59,60], 10 studies were in Asia [42,46,48,54,55,57,58,[62][63][64], 2 studies in Africa [52,65] and one study each was in Australia [50] and America [61]. Because of limited number of studies performed in Australia, America and Africa these studies were excluded from subgroup analysis. The results demonstrated no significant association between the ApaI polymorphism and risk of T1DM in the overall population and ethnic-specific analysis (Fig. 3). The results of pooled ORs, heterogeneity tests and publication bias tests in different analysis models are shown in Table 3.

Evaluation of heterogeneity and publication bias
During the meta-analysis of VDR gene polymorphism evidence of substantial to moderate heterogeneity was detected. However, partial heterogeneity was resolved while the data were stratified by ethnicity. Publication bias was evaluated by funnel plot, Begg's test and Egger's test. There was no obvious evidence of asymmetry from the shapes of the funnel plots (Fig. 4), and all P values of   Begg's test and Egger's test were > 0.05, which showed no evidences of publication biases.

Sensitivity analysis
The leave-one-out method was used in the sensitivity analysis to explore the effect of individual data on the pooled ORs. The significance of ORs was not altered through omitting any single study in the dominant model for FokI, TaqI, BsmI and ApaI SNPs, indicating that our results were statistically robust (Fig. 5).

Bayesian meta-regression analysis
Meta-regression and subgroup analyses were performed to explore potential sources of heterogeneity among included studies ( Table 4). The findings of metaregression indicated that ethnicity can be the potential source of heterogeneity, therefore, subgroup analysis was performed to attenuate the effect of these parameters. (Fig. 6).

Discussion
In this study, we performed a systematic review and meta-analysis to achieve a vivid and exact approximation of the associations between the VDR gene polymorphisms, including FokI (rs2228570), TaqI (rs731236), BsmI (rs1544410), and ApaI (rs7975232) and susceptibility to T1DM. Over the course of past years, a bulk of studies has addressed the association of VDR gene polymorphisms and risk of T1DM throughout various populations, resulting in conflicting findings [61,67]. Such discrepancies might stem from diversity in detection methods, differences in diagnostic criterions, clinical heterogeneity, small sample sizes, low statistical power, and interactions between genetic and environmental contributing factors according to variations in the geo-epidemiological factors. As a consequence, three previous meta-analyses by Guo et al. [21] [20] in 2014 (13 studies for TaqI, 23 studies for BsmI, 15 studies for ApaI, and 18 studies for FokI) were carried out to resolved the conundrum and attain an exact approximation. They indicated that VDR gene SNPs were not associated with T1DM risk, except than BsmI polymorphism association with T1DM predisposition that was observed in Zhang et al. [22] study. Upon the latest meta-analysis published in 2014, several original association studies evaluated the role of VDR gene polymorphisms with T1DM risk. As a result, the necessity for performing an updated meta-analysis is sensed to come up with resolution of the limitations of individual association studies and to gain a much more valid and comprehensive pooled estimation on the association of VDR gene polymorphisms with T1D risk.
Previous meta-analysis performed by Tizaouia et al. [20] in 2014 reported no significant association of VDR gene FokI polymorphism with risk of T1D. According to our meta-analysis, the pooled results in overall population across all genotype models demonstrated no significant association of VDR gene FokI polymorphism; nonetheless, subgroup analysis according to ethnicity showed a marginally-significant decreased susceptibility to T1DM in European population according to dominant genetic model and heterozygote comparison, while an increased risk of T1DM in African population according to all genotype models. In addition, our meta- analysis did not support any significant association between TaqI SNP and susceptibility to T1DM. Furthermore, the results of subgroup analysis according to ethnicity did not show any significant association in all genetic models. However, in the subgroup analysis, given that there was only one study in the Australian [50] and American [61] populations, and two studies in the African [52,65] population, the subgroup analysis was not performed in these populations. In line with our findings, previous meta-analysis by Tizaouia et al. [20] also did not show significant association of VDR gene TaqI polymorphism with risk of T1D. According to the previous meta-analysis, BsmI SNP was not the risk factor for T1D susceptibility. However, after excluding one study, a marginal significant (P = 0.051) association was found in the homozygous model. On the other side, our metaanalysis also revealed that BsmI polymorphism was not a risk for T1DM in all genetic models when all of the population were analyzed. Nonetheless, subgroup analysis demonstrated a strong negative significant association between BsmI SNP and the risk of T1DM in American population in all of the genetic model comparisons. Finally neither our meta-analysis nor the previous one by Tizaouia et al. [20] found any significant association of ApaI polymorphism and T1DM risk in overall as well as subgroup analyses. Taken together, although our meta-analysis included further studies compared to the previous study, the overall analysis was almost the same. Nonetheless, our subgroup analysis indicated association of VDR genetic polymorphisms with T1DM risk in different ethnical groups.
In their meta-analysis, Tizaoui et al. [20] indicated in the stratification analysis that publication year, age, gender, estimated VitD levels, and latitude modulated the association between VDR gene polymorphisms and T1D risk. Furthermore, another meta-analysis revealed a relationship between winter ultraviolet radiation (UVR) and VDR gene polymorphisms in T1DM, implying to the influence of the UVR on the association between VDR polymorphisms and T1DM susceptibility [72]. During the four cooler months, it was observed that latitude  strongly determines the available levels of VitD producing UV. As latitude increases, the amount of VitD producing UV decreases, which may prevent VitD synthesis in humans [73]. As a result, the latitude of the locations in which the individuals live may impress the susceptibility to develop T1DM. Despite we tried to conduct best meta-analysis of the VDR gene polymorphisms and susceptibility to RA, there was also a number of limitations that should be taken into account. First, there was significant heterogeneity across studies, which may lessen the certainty of the results. However, we tried to find and attenuate its effect by meta-regression and subgroup analysis. Consequently, heterogeneity was still an unavoidable problem that may influence the accuracy of the overall results. Second, only articles published in the English language were include in this meta-analysis. Third, our metaanalysis was based on crude approximation of the genetic variations regardless of adjusting the analysis by gender, age, VitD intake, and other environmental factors like exposure to sun light, as several studies noted the involvement of these parameters as well as geneenvironment and gene-gene interactions in the susceptibility and of RA and we could not analyze it owing to a lack of published well-structured data.

Conclusion
In conclusion, this study was a systematic review and meta-analysis of 40 case-control association studies to come up with the clear estimation of the associations between the VDR gene SNPs [FokI (rs2228570), TaqI (rs731236), BsmI (rs1544410), and ApaI (rs7975232)]  and susceptibility to T1DM. The findings of metaanalysis revealed no significant association of VDR gene SNPs with T1DM risk in the overall population. However, the subgroup analysis indicated significant associations between FokI and BsmI polymorphism and T1DM risk in African and American population. As a limitation, we did not evaluate a number of VDR gene SNPs that might act in interaction with environmental factors to determine the fate of T1DM pathogenicity. Further investigations on the VDR, above and beyond the genetic as well as traditional risk factors, may confer a possibility for identification of critical susceptibility factors in the disease development, which might be applicable in the personalized medicine for better and optimized therapy of T1DM patients.