Vitamin D receptor gene polymorphisms in association with diabetic nephropathy: a systematic review and meta-analysis

Background A large amount of researches have demonstrated that vitamin D receptor (VDR) gene polymorphisms are associated with diabetic nephropathy (DN) risk in diabetes mellitus (DM) patients. Nevertheless, the results are inconclusive and inconsistent. Methods We screened PubMed, Embase, Chinese National Knowledge Infrastructure and Chinese Wanfang databases for those relevant studies updated in May 2016. Results 7 studies involving 2564 subjects were recruited. We evaluated the genotypic and allelic differences between DN patients and DM controls. Overall analysis showed that no significant association was found among the ApaI, BsmI, FokI,TaqI gene polymorphisms and DN susceptibility in diabetic patients (all P values > 0.05). In the stratified analysis, TT genotype was related to DN susceptibility in Asians (TT vs Tt + tt: OR =2.21, 95% CI: 1.05–4.67, p = 0.04). The sensitivity analysis showed that the results in overall populations, Caucasians and Asians were dependable. Conclusions No significant association was found among the ApaI, BsmI, FokI, TaqI polymorphisms and DN risk in overall populations, the TaqI variants might related to DN susceptibility in Asians. Further researches are required to testify our meta-analysis.


Background
The vascular complications of diabetes mellitus (DM) can lead to high disability rates and mortality rates. Poorly controlled blood glucose leads to the occurrence and development of complications in patients with DM [1]. As one of the most serious diabetic microvascular diseases, diabetic nephropathy (DN) is the main reason of end-stage renal failure(ESRD) [2,3]. Generally, DN is a multifactorial disease attribute to the interaction of environmental and genetic factors [4,5]. Several factors always contributing to DN risk include abnormal renal hemodynamic responses, fatty acid metabolism caused by hyperglycemia, hypertension, and abnormal metabolism of vasoactive substances [6]. Recently, genetic predispositions have been found to play a key role in the development and progression of DN [7,8]. Thus, searching for genetic markers for DN can identify patients who may benefit from prevention.
Previous studies showed that vitamin D endocrine system played an important role in the development of DM [9][10][11]. High levels of vitamin D can enhance pancreatic b-cell secretion functions and improve insulin resistance [12,13]. There are accumulating evidences to suggest that vitamin D participated in a number of diseases such as DM and DN [14][15][16]. An increasing number of researches showed that there was a relationship between vitamin D receptor (VDR) gene polymorphisms and DN [17,18]. Vitamin D exerts its functions by binding to the nuclear or cytosolic VDR, which serves as a transcription activator and is a member of the steroid / thyroid hormone receptor family [17]. At present, the most common studied single nucleotide polymorphisms (SNPs) of VDR are FokI (rs10735810), BsmI (rs1544410), ApaI (rs7975232) and TaqI (rs731236) [19,20].
To provide an accurate estimation, we operated a systematic review to investigate the relation of DN and the four VDR gene polymorphisms(ApaI, BsmI, FokI and TaqI).

Search strategy
The related articles were searched from PubMed, Embase, Chinese National Knowledge Infrastructure and Chinese Wanfang databases. The The retrieval strategy were as follows: "VDR, vitamin D receptor, diabetic nephropathy (DN), ApaI, BsmI, FokI, TaqI, polymorphism, genotype". In addition, we retrieved references of selected articles. Here are the inclusion criteria: 1) case-control study; 2)DN as ending; 3) carried out in human population; 4) Clearly define genetic variants of case and control can offer sufficient calculational data. Here are the exclusion criteria: (1) case reports or review articles; (2) articles that did not offer the detailed genotype information; (3) studies on the correlation between DN and other genes; (4) studies investigate the role of VDR genotype in other diseases. The largest or most recent study was selected when continuous or multiple publications that occur under the same search condition [21].

Data extraction
For each study the following information was evaluated independently by two investigators according to the standard protocol:(1) name of first author; (2) year of publication; (3) original country; (4) ethnicity of the subjects; (5) genotype distribution or allele frequencies; and (6) number of cases and controls (Table 1). Genotypes for the VDR gene polymorphisms(ApaI, BsmI, FokI or TaqI) were designated with A, B, F and T for the absence of restriction sites and with a, b, f and t for their presence. In most studies, the ApaI, BsmI, FokI or TaqI genotype and allele frequency of the VDR gene were measured by PCR-restriction fragment length polymorphism (PCR-RFLP). In one study [22], pyrosequencing reactions were performed to determine the BsmI, FokI or TaqI genotype, Taqman 5′-nuclease assays were performed used to determine the ApaI genotype, the labels of FokI (C/T) are correspond to FokI (F/ f ),BsmI (A/G) are correspond to BsmI (B/b),ApaI (T/ G) are correspond to ApaI (A/a),TaqI(T/C) are correspond to TaqI(T/t).

Statistical analysis
The odds ratio (OR) with 95% confidence interval (CI) were used to evaluate the association among the four VDR gene polymorphisms and DN risk. Statistical analysis were divided into Caucasian and Asian populations in subgroup analysis. The Chi-square based Q-statistic test was used to assess the heterogeneity between the studies. P < 0.10 indicated there is a significant heterogeneity among the studies. If the P value was no more than 0.1, pooled OR was estimated using a random-effect model using the DerSimonian and Laird method (D + L), otherwise a fixed-effect model using the Mantel-Haenszel (M-H) method was carried out. The asymmetry funnel plots were used to estimate the publication bias. The exact test was carried out to assess whether the genotype distribution in control population were accord with Hardy-Weinberg equilibrium (HWE) expectations. A sensitivity analysis of the overall population was used by omitting one study in each turn. Cochrane Review Manager Version 5.1 (Cochrane Library, Oxford, UK) was used to performed the statistical analyses. All the P values were double tailed test and the significance was set at P < 0.05.

Characteristics of the studies
The search yielded 67 articles, 7 studies [18,[22][23][24][25][26][27] contain 1230 DN patients and 1334 diabetic controls were finally recruited into our meta-analysis according to the inclusion and exclusion criteria, all studies reporting the association among ApaI, BsmI, FokI, TaqI of VDR gene polymorphisms and DN susceptibility (Fig. 1). The main characteristics of these selected studies were summarized in Table 1, including the first author's name, year of publication, original country and genotype distribution.

Quantitative data synthesis
It has been shown in Table 2    between genetic variant and diabetic nephropathy. Besides, there was no significant correlation between other three polymorphisms and DN risk in Asians and Caucasians (Table 2). Altogether, the results of the metaanalysis can be described as negative. Particularly worth mentioning is that only one published study in Asian populations was included in the BsmI and FokI groups, the subgroup for Asians might not be reliable in these two groups. Two studies [24,27] were excluded because the genotype distributions in controls were obviously deviated from the HWE. Sensitivity analysis showed that the results were dependable for overall populations, Caucasians and Asians.

Publication bias
Begg's funnel plots and Egger's test were carried out to analyse the publication bias. Results showed that there was no evidence of publication bias in the 7 studies (data not shown). However, due to the small number of studies, publication bias cannot be ruled-out in this paper.

Discussion
DN is the most common microvascular complication in diabetes patients, and the most common reason of ESRD in the United States [28]. 40% DM patients suffer from DN, and the occurrence have no association with hyperglycemia, suggesting genetic factors may be involved in the development of DN [29,30].Currently, VDR has also been illustrated to be associated with the occurrence and development of DN [31].
The vitamin D has important biological functions, such as modulating immunity system, influencing insulin secretion and improving insulin resistance [32,33], which are involved in the etiology of DN and more likely to be influenced by VDR gene polymorphisms. Given the controversial results from published individual studies with small sample sizes, we carried out a metaanalysis to clarify the association between the VDR gene polymorphisms and susceptibility to DN.
The VDR gene is located on chromosome 12 at q12-q14, it has14eons, 8 protein-coding exons and 6 untranslated region. ApaI and BsmI VDR gene variants locate in intron 8 and TaqI variants locate in exon 9. The FokI variants locate in the codon initiating translation, which has been reported to cause severe protein dificiency [25]. A large amount of common chronic disorder diseases, such as type 1 and type 2 DM, have been found to be relevant to specified VDR gene polymorphisms [34]. The effects of VDR gene polymorphisms on gene expression are still not clearly understood. VDR gene variants may have effect on RNA translational efficiency [35]. Distinctions of VDR mRNA express level and stability can affect the activity of renin-angiotensin system and participate in the progress of DN [26,27].
In this study, we analyzed the VDR gene functional polymorphisms (ApaI, BsmI, FokI and TaqI) as potential genetic markers for DN. Nevertheless, the results of our research showed that none of these polymorphisms was significantly associated with DN risk in diabetic patients in overall population(all P values >0.05). Since genetic factors can affect the results of meta analysis, subgroup analysis were also preformed by race for the four polymorphisms. Our results indicate that there was a relationship between TT genotype and DN risk in Asians (TT vs Tt + tt: OR =2.21, 95% CI: 1.05-4.67, p = 0.04), however, p value equals to 0.04 seems not enough to support relationship between genetic variants and DN. Since only one published study in Asian populations was included in the BsmI and FokI groups, the stratified analysis for Asians might not be reliable in these two groups. Besides, there was no significant associations among the other three gene variants and DN in Caucasians and Asians (Table 2). Altogether, the results of the metaanalysis can be described as negative. Liu [34] reported that the FokI gene variants was related to DN susceptibility in Caucasians. We have a different result, and we performed a more comprehensive research on the association among four gene polymorphisms (ApaI, BsmI, FokI and TaqI) of VDR and DN risk. However, further large-scale studies should be carried out to testify our research.
For better understanding of the results, several limitations in our study should be mentioned. First, the number of subjects were relatively small, so we lacked sufficient statistical power to estimate the associations between VDR gene variants and susceptibility to DN. Second, metaanalysis is a retrospective study that may cause selection bias, which may have an impact on the reliability of results. Third, our study failed to acquire raw data from the selected articles, which may lead to further limitation in the assessment of potential effects of VDR gene polymorphisms in the occurrence and progression of DN.
Our research provide evidence that VDR TaqI polymorphism may be relevant to the risk of DN in DM patients among Asians. However, the number of articles included is really small, which is an important limit, and with such a small number of studies, subgroup (Asian/ Caucasian) results we reported have to be considered as exploratory. Martin et al.'s [22] is by far the largest sample size study in our research. Consequently, the overall meta-analysis report results very close to those of Martin et al. Therefore, large-scale studies are still needed to provide a more representative meta-analysis.

Conclusions
In conclusion, our meta-analysis showed that no significant association was observed among the ApaI, BsmI, FokI and TaqI variants and DN susceptibility in overall populations, the TaqI variants may be associated with DN risk in Asian populations. Future larger scale epidemiological investigation of this topic should be conducted to confirm or refute our findings.