摘要

Objectives: The G1057D polymorphism in insulin receptor substrate (IRS)-2 gene is associated with type 2 diabetes mellitus (T2DM) risk, but results in published literatures are controversial. In addition, the effect of obesity as a modifier on this association is also inconsistent. Thus, this meta-analysis was performed to assess the above-mentioned association. @@@ Methods: A comprehensive search was performed to identify case-control or cohort studies (from 1990 to 2014) of the aforementioned association. The 12 statistic was used to examine between-study heterogeneity. Fixed or random effect model was selected based on heterogeneity test among studies. Publication bias was estimated using modified Egger's regression test. @@@ Results: Nine articles with ten studies were included. After excluding studies deviated from Hardy-Weinberg equilibrium (HWE) in controls, results showed a significant association of D allele with reduced 12DM risk in dominant (OR = 0.825,95% CI: 0.705-0.965) and codominant (OR = 0.857, 95% CI: 0.763-0.964) models, but no significant association in recessive (OR = 0.806,95% CI: 0.628-1.035) model. For studies stratified by obesity, after excluding studies deviated from HWE in controls, no significant association of D allele with 12DM risk was found in three inherited models in obese group; however, a significant protective effect of D allele was observed in dominant (OR = 0.714, 95% CI: 0.533-0.958), recessive (OR = 0.438, 95% CI: 0.253-0.760) and codominant (OR = 0.706, 95% CI: 0.565-0.883) models in non-obese group. @@@ Conclusions: This meta-analysis suggested that D allele of G1057D polymorphism have a significant effect on reduced risk of 12DM, and obesity is a modifier of this association. This result needs to be confirmed by further studies.