Table 2.

HR (95% CI) for diabetes associated with statin use at visit prior to diabetes diagnosis

Adjusted models	Pooled	Placebo	Metformin	Lifestyle	Group p value
Model 1: demographic	1.36 (1.17 to1.59)	1.21 (0.93 to 1.57)	1.33 (1.02 to 1.73)	1.59 (1.21 to 2.10)	0.36
Model 2: 1+baseline diabetes risk factors	1.35 (1.15 to1.57)	1.18 (0.90 to 1.54)	1.37 (1.05 to 1.78)	1.53 (1.16 to 2.03)	0.41
Model 3: 2+updated statin confounders	1.27 (1.08 to1.50)	1.15 (0.87 to 1.53)	1.31 (0.99 to 1.73)	1.36 (1.00 to 1.86)	0.71
Model 4: 2+updated diabetes risk factors	1.27 (1.08 to1.49)	1.19 (0.91 to 1.55)	1.36 (1.04 to 1.76)	1.37 (1.04 to 1.81)	0.71
Model 5: dully adjusted	1.27 (1.08 to1.50)	1.20 (0.90 to 1.59)	1.33 (1.01 to 1.76)	1.43 (1.06 to 1.94)	0.70

Adjusted HRs for updated statin use are noted in bold at significance level of α=0.05 based on the following models: 1: age, sex, and race/ethnicity; 2: model 1+baseline diabetes risk factors (family history of diabetes, fasting plasma glucose, log Insulinogenic Index, log 1/fasting insulin, waist); 3: model 2+time-dependent statin treatment confounders (use of antihypertensive medications, diastolic blood pressure, systolic blood pressure, HDL-cholesterol, and LDL-cholesterol), which were updated until visit prior to statin initiation or diabetes assessment, plus baseline SES (years of education and household income) and reported history of cardiovascular disease at baseline; 4: model 2+time-dependent diabetes risk factors (log 1/fasting insulin, log Insulinogenic Index, waist) updated until the visit prior to the assessment of diabetes; 5: model 3+updated diabetes risk factors. Group p value represents the test for heterogeneity of HR among treatment group.

HDL, high density lipoprotein; LDL, low density lipoprotein; SES, socioeconomic status.