Table 3.

Multivariable analyses of predictors for NAFLD and fibrosis in people with HIV.

	NAFLD			Fibrosis
	Multivariable logistic regressiona			Multivariable logistic regressionb
Variable	OR	CI	P value	OR	CI	P value
Waist circumference (cm)	1.1	1.067–1.135	<0.001
Type 2 diabetes				5.056	1.386–18.44	0.014

Multivariable logistic regression analysis of data is shown. With all factors showing a P value less than 0.05 in the univariable analysis and the variables, age and sex, a multivariable regression model was built. Age, sex, BMI and ALT were not predictive of NAFLD. Likewise age, sex, BMI, ALT, AST and alcohol intake were not predictive of fibrosis. Metabolic syndrome was excluded from the multivariable analysis to avoid multicollinearity. The variable triglycerides was excluded because of limited numbers available. Odds ratio (OR) and 95% confidence interval (CI) are shown. Boldface indicates statistical significance. A P value less than 0.05 was considered statistically significant. NAFLD, nonalcoholic fatty liver disease.

^aMultivariable logistic regression: age, sex, BMI, waist circumference, ALT (n = 230).

^bMultivariable logistic regression: age, sex, BMI, waist circumference, ALT, AST, type 2 diabetes, alcohol intake (n = 212).