Meta-analysis of microRNA-146a rs2910164 G>C polymorphism association with autoimmune diseases susceptibility, an update based on 24 studies

Abstract

Background: Published data showed that the susceptibility of autoimmune diseases (ADs) was associated with the polymorphism rs2910164 in microRNA-146a (miR-146a). However, the results remain controversial so far. Two meta-analyses published in 2013 and 2014 came to opposite conclusions. In order to derive a more precise estimation of the relationship, we performed this meta-analysis.

Methods: We searched the PubMed, OvidSP and CNKI databases (published prior to September 8th, 2014) and extracted data from eligible studies. The procedure of meta-analysis was performed by using the Stata 12.0 software. Random effect model or fixed effect model were chosen respectively, according to the between study heterogeneities.

Results: A total of 24 case-control studies, 11 more than previous meta-analysis on this topic, were involved. We took stratified analyses by different ethnicities and different types of diseases in different genetic models. In Caucasian subgroup, significant increased risks of GC genotype and GC+CC genotype with ADs susceptibility were found in heterozygote model (GC vs GG, OR = 1.38, 95% CI 1.04-1.83, p = 0.024) and dominant model (GC+CC vs GG, OR = 1.37, 95% CI 1.01-1.85, p = 0.041), respectively. Meanwhile, in other disease subgroup, significant increased risks of C allele, CC genotype and GC+CC genotype were found in allele model (C vs G, OR = 1.16, 95% CI 1.04-1.31, p = 0.010), homozygote model (CC vs GG, OR = 1.42, 95% CI 1.10-1.84, p = 0.006) and dominant model (GC+CC vs GG, OR = 1.25, 95% CI 1.04-1.51, p = 0.020), respectively.

Conclusions: MiR-146a rs2910164 G>C polymorphism was associated with the susceptibility of ADs.

Fig 1. Flowchart for identification of studies included in the meta-analysis.

In 411 articles, 51 were found not related to ADs and 43 were found not related to miR-146a by scanning the titles. After that, 233 articles were recognized as reviews, 37 were found not related to human patients and 8 articles were repeated papers by reviewing the abstracts. The full-texts of the left 39 articles were carefully reviewed, in which 18 articles were found not about rs2910164. At last, 21 articles were remained for this meta-analysis, which included 24 case-control studies for rs2910164.

Fig 2. Forest plots of ADs risk associated with rs2910164.

(A-B) Forest plots of ADs risk associated with rs2910164 stratified analyzed by ethnicities. (A) Heterozygote model, GC vs GG, Caucasian subgroup, random model. (B) Dominant model, GC+CC vs GG, Caucasian subgroup, random model. (C-E) Forest plots of ADs risk associated with rs2910164 stratified analyzed by diseases. (C) Allele model, C vs G, Other diseases subgroup, fixed model. (D) Homozygote model, CC vs GG, Other diseases subgroup, fixed model. (E) Dominant model, GC+CC vs GG, Other diseases subgroup, fixed model. OR: odds ratio; 95% CI: 95% confidence interval.

Fig 3. Cumulative meta-analysis of the association between rs2910164 and ADs risk.

Every rhombus represents the pooled OR when studies accumulated over time, and the horizontal line represents the 95% CI of the pooled ORs. (A) Heterozygote model, GC vs GG, Caucasian subgroup, random model. (B) Dominant model, GC+CC vs GG, Caucasian subgroup, random model. (C) Allele model, C vs G, Other diseases subgroup, fixed model. (D) Homozygote model, CC vs GG, Other diseases subgroup, fixed model. (E) Dominant model, GC+CC vs GG, Other diseases subgroup, fixed model.

Fig 4. Sensitivity analysis of association of rs2910164 and ADs risk.

(A) Pooled analysis of association of rs2910164 and ADs risk. Allele model, C vs G. (B) Sensitivity analysis by iteratively removing one study at a time. (C) Sensitivity analysis by removing three studies with low quality. (D) Sensitivity analysis by removing six studies without HWE.

Fig 5. Publication bias on the rs2910164 polymorphism and ADs risk.