Logistic regression models for NAFLD as a predictor for AVS in patients with type 2 diabetes.

Sample size, n = 180. Data are expressed as odds ratios ±95% confidence intervals (CI) as assessed by univariable (unadjusted) or multivariable logistic regression analyses.
Other covariates included in multivariable logistic regression models, along with NAFLD, were as follows: model 1: age and sex; model 2: age, sex, BMI, smoking status, daily alcohol consumption, hypertension (i.e. blood pressure ≥140/90 mmHg and/or anti-hypertensive treatment) and dyslipidemia (i.e. LDL-cholesterol >130 mg/dl and/or triglycerides >200 mg/dl and/or HDL-cholesterol <40 mg/dl for women or <35 mg/dl for men and/or lipid-lowering treatment); model 3: adjustment for variables included in model 2 plus duration of diabetes, diabetes treatment, HbA1c, and estimated GFR.