Sensitivity analysis & publication bias
Sensitivity analyses were conducted through omitting one study at a time. We also conducted sensitivity analysis by removing the study deviating from HWE. No material changes of the pooled ORs were found after omitting any single study or the HWE-violating study[15] in all genetic models for Arg16Gly polymorphism and in dominant, heterozygous and additive models for Gln27Glu polymorphism. However, as for Gln27Glu polymorphism, the corresponding pooled ORs were considerably changed after deleting a single study [17] at a time in the recessive model (OR=0.348, 95% CI: 0.093-1.308, P =0.118), allele model (OR=0.700, 95% CI: 0.428-1.145, P =0.155) and homozygous model (OR=0.334, 95% CI: 0.087-1.288, P =0.111). Egger’s test and Egger’s test were performed to detect the possibility of publication bias. The results demonstrated that publication bias under any genetic models was inapparent (data not shown). The funnel plots showed no obvious asymmetries (Figure 5).