Background:  There have been concerns of an increased incidence of fatty liver disease among HIV⁺ individuals receiving ART that could be related to drug toxicity, hepatitis B virus (HBV) or hepatitis C virus (HCV) co-infection or the metabolic derangements associated with ART.

Methods:  We conducted a case control study of all liver biopsies performed in HIV⁺ subjects co-infected with HBV or HCV at Washington University, Saint Louis, Missouri, between 1999 and 2003. The HIV^­ control group was matched for date of the biopsy; body mass index (±1 kg/m²) and alcohol use. Clinical data were collected retrospectively and biopsies were prospectively reevaluated by 2 observers. The degree of inflammation and fibrosis was determined using the Scheuer classification (grading/staging from 0 to 4), and the degree of fatty infiltration by using a modified Brunt Score (grading from 0 to 3)

Results:  We reviewed biopsies from 140 patients (70 HIV⁺ and 70 HIV^­). Racial/ethnic distribution (50% white), AST, ALT, cholesterol, and triglyceride (TG) levels and the frequency of diabetes were similar in both groups. Male gender (92% vs 68%, p <0.001), younger age (41 vs 48 years, p <0.001), and HBV infection (33% vs 14%, p = 0.02) were more common in the HIV⁺ group. There were no statistically significant differences in the frequency of steatosis among HIV⁺ and HIV^­ subjects. Among HIV⁺, steatosis was present in only 27 (39%) patients (grade 1 29%, grade 2 10%) vs 24 (34%) of HIV^­ (grade 1 21%, grade 2 11% and grade 3 1%). Among HIV⁺, 24 of 27 had had macrovesicular steatosis and only 3 microvesicular steatosis. Univariate analysis for steatosis risk factors revealed no statistically significant correlation of steatosis with demographic variables, duration of ART, type of hepatitis co-infection or CD4 levels. In a multivariate logistic regression model, body mass index was the only independent predictor of hepatic steatosis (p <0.0125) among HIV-infected individuals. The odds ratio was 1.27 x per each 1.0 kg/m² body mass index increase. For the highest body mass index >30, the odds ratio was 12.14 (1.3 to 113.54).

Conclusions:  Steatosis was as common in HIV⁺ as in HIV^­ individuals and was not associated with ART or hepatitis co-infection. As in the general population, body mass index remained a strong predictor for steatosis in HIV⁺ individuals and controls. Our data do not suggest an increase in incidence of hepatic steatosis among HIV-infected patients above and beyond the current epidemic of obesity in the United States.