Bioimpedance Analysis to Predict Clinical Outcomes of Liver Cirrhosis in a Prospective Cohort Study

Background and Aims: Bioimpedance analysis (BIA) is a non-invasive method for estimating body composition, which plays a role in the development of cirrhosis. We investigated if BIA measures were associated with cirrhosis complications. Methods: We analyzed data from 566 participants in a prospective cohort study of cirrhosis patients from tertiary care clinics who underwent BIA at enrollment and regular follow up (20.3% female; 47.0% non-Hispanic White, 23.1% Hispanic, 27.7% Black; 47.7% metabolic and alcohol-associated liver disease [MetALD], 30.4% metabolic dysfunction-associated steatotic liver disease [MASLD], 1.9% alcohol-associated liver disease [ALD], 9.0% active hepatitis C virus [HCV]). We performed multivariable logistic regression to assess the association of BIA measures with etiology and prevalent decompensation and Cox proportional hazards regression to assess incident decompensation and HCC. Results: After adjustment for sex and race, body fat mass was associated with 14-fold increased odds of MASLD (adjusted odds ratio [aOR] 14.10, 95% confidence interval [CI] 5.25-37.85) and decreased odds of ALD (aOR 0.73, 95% CI 0.60-0.88). Lower odds of prevalent decompensation were seen in the highest tertile of trunk lean mass (aOR 0.39, 95% CI 0.21-0.73) and left arm lean mass (aOR 0.40, 95% CI 0.21-0.75) compared with their lowest tertiles. The middle tertile of visceral fat surface area showed an inverse association with incident decompensation (adjusted hazard ratio 0.19, 95% CI 0.04-0.88). There were no significant associations between BIA measures and incident HCC. Conclusions: BIA could be used as a prognostic tool for patients with cirrhosis, supporting its incorporation into clinical workflows.