PP217 - BIOIMPEDANCE ANALYSIS FOR EVALUATING OBESITY AND METABOLIC RISK IN LIVING LIVER DONORS

PP217

BIOIMPEDANCE ANALYSIS FOR EVALUATING OBESITY AND METABOLIC RISK IN LIVING LIVER DONORS

S. Kim^1,*, D. Flach², K. Hong³, Z. Elliott⁴, H. H. Han⁵, B. Kim ⁵, J. Kahn⁵, A. L. Padilla¹, Y. Kwon⁶, Y. Genyk⁷, N. Kaur¹

¹Hepatobiliary & Abdominal Organ Transplant, ²Gerontology, ³Internal Medicine, ⁴Biology, ⁵GI/Liver, University of Southern California, Los Angeles, ⁶Hepatobiliary & Abdominal Organ Transplant, UT Southwestern, Dallas, ⁷Hepatobiliary & Abdominal Organ Transplant, MedStar Georgetown, Washington D.C., United States

Rationale: BMI is widely used to screen living liver donor candidates for obesity-related risk, but cannot differentiate fat from lean mass, predisposing to misclassification. We evaluated whether bioimpedance analysis (BIA)-derived metrics, including percent body fat (%BF) and fat mass, provide superior characterization of hepatic steatosis and metabolic risk vs. BMI alone.

Methods: Single-center retrospective cohort of 314 adult living liver donor candidates (2019–2024). BMI and sex-specific %BF-based obesity definitions were compared; associations with hepatic fat fraction (MRI-PDFF) and metabolic biomarkers were assessed by Spearman/Pearson correlation and univariate linear regression.

Results: Of 314 evaluated candidates (approved n=188; deferred/declined n=55), BMI identified 38% as obese vs. 59% by sex-specific %BF thresholds, demonstrating misclassification. BMI and %BF were strongly but incompletely correlated (rₛ = 0.663, p <0.001). Hepatic fat fraction was moderately associated with BMI (rₛ = 0.338) and fat mass (rₛ = 0.322), and weakly with %BF (rₛ = 0.245; all p <0.001); it correlated strongly with ALT (rₛ = 0.452) and HbA1c (rₛ = 0.249) than other anthropometric measures. On linear regression, BMI explained 11.5% of hepatic fat fraction variance (R² = 0.115), %BF explained 6.0% (R² = 0.060), and fat mass was comparable to BMI (R² = 0.103); all three were limited independent predictors.

Conclusion: BMI and %BF are modest, partially redundant predictors of hepatic steatosis, and neither is clinically decisive in isolation. A 21-point gap in obesity prevalence estimates (38% vs. 59%) underscores the misclassification risk of relying on BMI alone. Uniquely, BIA enables identification of high-risk phenotypes invisible to BMI, including normal-BMI/high-fat and sarcopenic obesity profiles. Incorporating BIA into living donor evaluation may meaningfully improve risk stratification and guide prehabilitation.

Disclosure of Interest: None declared