OBJECTIVETo clarify the contradictory findings in patients with obesity and type 2 diabetes by meta-analysis.

METHODSPubMed and Embase were searched for articles published up to March 2009. All studies on the association of FTO polymorphisms with obesity and type 2 diabetes were included. Pooled odds ratio was calculated using the model of fixed or random effects. Sensitivity analysis was performed to evaluate the stability of meta-analytic results.

RESULTSMeta-analysis suggested that rs9939609 A allele was more significantly associated with obesity risk than T allele (3 studies / 004 cases and 4544 control subjects): random effect odds ratio (OR)=1.28, 95%CI=1.05 and 1.55, P heterogeneity =0.05, I2=66.6%. Similar results were observed in rs8050136 polymorphism (3 studies/2404 cases and 5713 control subjects): fixed effect OR =1.25, 95%CI=1.13, 1.37, P heterogeneity=0.12, I2=51.9%. However, no significant association was found between genetics and risk of type 2 diabetes after control of potential confounders (at least for BMI) either for rs9939609 (fixed effect OR=1.05, 95% CI=0.97,1.13) or for rs8050136 polymorphism (fixed effect OR =1.07, 95%CI: 0.99, 1.16). Furthermore, the sensitivity analysis strengthened our confidence in validity of the association. Conclusion FTO polymorphisms are associated with obesity but not with type 2 diabetes in East Asian populations. Further large-scale studies are required to conclusively establish the association.