The models were analyzed after substituting the continu ous variables related to MetS components with their di chotomous counterparts in the models. Variables were entered into the backward stepwise regression models if a P value 0. 10 was obtained. Bivariate association ana FTY720 Multiple Sclerosis lysis based on generalized linear model may treat correla tions of outcomes so as to be more power to detect the shared contributors to outcomes as compared to univari ate association analysis. Generalized linear model was performed to include dependences of both SHF and DHF simultaneous to identify shared predictors. The predictive performance of MetS severity score was evalu ated with respect to the area under the curve in a receiver operating characteristics curve.

Odds ra tios with 95% confidence intervals were calcu lated for the relative risk of MetS with SHD or DHF. Results were analyzed using the Statistical Package for Social Sciences for Windows version 16. 0. Tests were two sided and a P value of 0. 05 was considered significant. Results Baseline clinical characteristics of the 347 subjects were grouped according to SHF, DHF and control. The total sample included 208 males and 139 females in total sample. Gender, height, PBG, HbAlc, LDL and TC levels were similar among the three groups, while the other demo graphic parameters and biochemistry variables were sig nificantly different. LVEF were significantly different among the three groups. The prevalence of HT, DM, MetS and CAD were 71. 18%, 49. 26%, 26. 51% and 24. 78% in the patients, respectively.

The four chronic dis eases were more prevalent in SHF group than the other two groups. In the high risk patients, the CAD prevalence was no significant difference among three groups. The use of oral medications were sig nificant different among the three groups. The number of patients with DHF and SHF were 97 and 126 accounted for 27. 95% and 36. 71% in the patients, re spectively. SHF was more prevalent than DHF in total subjects. The subjects in New York Heart Association class IV and III was 11. 11% and 12. 98% in SHF groups, respectively. The subjects in NYHA class III was 12. 37% in DHF groups. Components of MetS v. s SHF and DHF Univariate association analysis using multinomial LR to include single independent variable showed that MetS and its components significantly associated with SHF or DHF.

Backward step wise multinomial LR model to include components of MetS controlling for covariates of age, gender, smoking, HR, Ccr, UA, LVMI and CAD, indicated that three components of MetS FPG, HT and TG significantly associated with SHF or DHF independently. HT and TG were detected to independently add to your list associate with DHF, while HT and FPG independently associate with SHF. LR models demonstrated that HT associated with DHF or SHF. Its regression coefficient was significant greater in LR model for SHF as compared with that for DHF. In patients with HT, the OR of DHF was 1.