| Objective:To evaluate the efficacy of thiazolidinediones (TZDs) in the treatment of type2diabetes mellitus (T2DM) and provide the evidence for the clinical decisions.Method:Randomized controlled trials of T2DM patients that compared TZDs (rosiglitazone, pioglitazone) with other treatments for more than3months and reported at least one of HbAlc, FPG,2hPG, TG, TC, BMI, DBP and SBP those8indices were included. Software Review Manager4.2was used to assess the Weighted mean difference (WMD) and95%confidence interval (95%CI) of the indices. Sensitive analysis was used according to the different accepted standard, and funnel plot together with fail-safe number were made to confirm publication bias.Results:The analyses (rosiglitazone versus placebo) included7studies, and the WMD and95%CI of the HbAlc,FPG, TG,and TC were1.56(1.23,1.88),3.04(2.26,3.82),0.34(0.27,0.42) and0.03(-0.41,0.47); The analyses (TZDs versus metformin) included9studies, and there were5indices were analysed:the WMD and95%CI of HbAlc were-0.21(-0.32,-0.10), the WMD and95%CI of FPG were0.05(-0.47,0.60), the WMD and95%CI of BMI were-0.71(-1.03,-0.39), the WMD and95%CI of TG were0.28(0.21,0.35) and the WMD and95%Cl of TC were-0.34(-0.54,-0.14); The analyses of combined studies included23studies, and combined treatments had beneficial effects on HbAlc, FPG, 2hPG and TG. The sensitivity and subgroup analyses showed that the results of Meta analysis were stable. The fail-safe numbers showed that there was no publication bias.Conclusion:Rosiglitazone had beneficial effects on HbAlc, FPG and TG compared with placebo; Metformin had beneficial effects on HbAlc, BMI and TC compared with TZDs; The analyses of combined studies showed that combined treatments had beneficial effects on HbAlc, FPG,2hPG and TG. |