Evaluation of Bioelectrical Impedance Analysis in Critically Ill Patients: Results of a Multicenter Prospective Study

Abstract

Background: In critically ill patients, muscle loss is associated with adverse outcomes. Raw bioelectrical impedance analysis (BIA) parameters (eg, phase angle [PA] and impedance ratio [IR]) have received attention as potential markers of muscularity, nutrition status, and clinical outcomes. Our objective was to test whether PA and IR could be used to assess low muscularity and predict clinical outcomes. Methods: Patients (≥18 years) having an abdominal computed tomography (CT) scan and admitted to intensive care underwent multifrequency BIA within 72 hours of scan. CT scans were landmarked at the third lumbar vertebra and analyzed for skeletal muscle cross‐sectional area (CSA). CSA ≤170 cm 2 for males and ≤110 cm 2 for females defined low muscularity. The relationship between PA (and IR) and CT muscle CSA was evaluated using multivariate regression and included adjustments for age, sex, body mass index, Charlson Comorbidity Index, and admission type. PA and IR were also evaluated for predicting discharge status using dual‐energy x‐ray absorptiometry–derived cut‐points for low fat‐free mass index. Results: Of 171 potentially eligible patients, 71 had BIA and CT scans within 72 hours. Area under the receiver operating characteristic (c‐index) curve to predict CT‐defined low muscularity was 0.67 (P ≤ .05) for both PA and IR. With covariates added to logistic regression models, PA and IR c‐indexes were 0.78 and 0.76 (P < .05), respectively. Low PA and high IR predicted time to live ICU discharge. Conclusion: Our study highlights the potential utility of PA and IR as markers to identify patients with low muscularity who may benefit from early and rigorous intervention.