. Author manuscript; available in PMC: 2017 Aug 1.

Published in final edited form as: Clin Gastroenterol Hepatol. 2016 Apr 1;14(8):1172–1180.e2. doi: 10.1016/j.cgh.2016.03.034

Table 3. Models predicting hepatic steatosis (LPR ≤ 0.33) in the Framingham Heart Study Cohort.

Models	Covariate	Odds Ratio (95% CI)	P- Value	C-statistic^**
Clinical model^*	Age (per 1 year increase)	1.00 (0.98, 1.03)	0.68	0.830
	Sex (Women vs. Men)	0.65 (0.47, 0.91)	0.01
	BMI (per 1 kg/m² increase)	1.20 (1.16, 1.24)	<0.0001
	Triglycerides (per 1 mg/dl increase)	1.01 (1.00, 1.01)	<0.0001
	Hypertension (yes vs. no)	1.94 (1.38, 2.72)	0.0001
	Diabetes (yes vs. no)	2.56 (1.31, 5.00)	0.006
Framingham Steatosis Index^*	Age (per 1 year increase)	1.01 (0.99, 1.04)	0.36	0.845
	Sex (Women vs. Men)	0.86 (0.61, 1.23)	0.42
	BMI (per 1 kg/m² increase)	1.19 (1.15, 1.23)	<0.0001
	Triglycerides (per 1 mg/dl increase)	1.01 (1.00, 1.01)	<0.0001
	Hypertension (yes vs. no)	1.81 (1.28, 2.57)	0.0009
	Diabetes (yes vs. no)	2.20 (1.12, 4.33)	0.02
	ALT/AST Ratio (> 1.33 vs ≤ 1.33)	3.00 (2.15, 4.20)	<0.0001

BMI, body mass index; ALT, alanine aminotransferase; AST, aspartate aminotransferase

Multivariable stepwise regression models. Age and sex were forced into the model. Candidate variables for the selection model included smoking status, BMI, waist circumference, diabetes, hypertension, total cholesterol, triglycerides, HDL cholesterol, or low-density lipoprotein (LDL) cholesterol. The ALT/AST ratio was added to the clinical model to derive the Framingham Steatosis Index.

^**

Difference in c-statistic between the clinical model and the FSI=0.015, 95% CI of difference (0.005, 0.025), p<0.001