Predictors of acarbose therapeutic efficacy in newly diagnosed type 2 diabetes mellitus patients in China

Table 3 Stepwise multiple regression analysis of ΔHbA1c predictors

Variables	Partial regression coefficient	Standard error	Standardized regression coefficient	t	P
Model 1
FPG (mmol/L)	0.111	0.039	0.146	2.805	0.005
Weight (kg)	−0.012	0.004	−0.112	−2.758	0.006
2 h PG (mmol/L)	0.042	0.021	0.109	2.036	0.043
DBP (mmHg)	0.010	0.005	0.080	1.995	0.047
Model 2
FPG (mmol/L)	0.117	0.034	0.146	3.416	0.001
Disease duration (months)	0.040	0.017	0.087	2.351	0.019
Model 3
Disease duration (months)	0.038	0.019	0.081	2.005	0.046

Model 1: With 12-week ΔHbA1c as the dependent variable, the R² of the regression equation was 0.586 and the constant was 3.609; Model 2: With 24-week ΔHbA1c as the dependent variable, the R² of the regression equation was 0.589 and the constant was 4.001; Model 3: With 48-week ΔHbA1c as the dependent variable, the R² of the regression equation was 0.546 and the constant was 4.653

ISSN: 2050-6511