#2819 BIOELECTRICAL IMPEDANCE PARAMETERS AS EARLIER PREDICTORS OF FRAILTY AND FUNCTIONAL IMPAIRMENT IN ECUADORIAN HEMODIALYSIS PATIENTS

Abstract Background and Aims Functional impairment (FI) and frailty are highly prevalent in hemodialysis (HD) population and are associated with morbidity/mortality. Bioelectrical impedance analysis (BIA) is a non-invasive test to assess body composition and water distribution. Few is known about BIA parameters as predictors of frailty and FI. The aim of the study was to study BIA parameters associated with frailty and FI and to determine their prevalence among Ecuadorian adult patients on HD. Method Observational-prospective-cohort study performed between January/2021 and October/2022 in one HD center. Clinical Frailty Scale (CFS) score ≥4 determined frailty diagnosis, and FI was labeled with Barthel Index (BI) <70. ROC curves were performed to determine the best cut-off for Phase Angle (PA) and Extracellular/Intracellular water ratio (E/I ratio) to predict frailty and FI (Fig. A). Pearson's correlation coefficient between PA and E/I ratio with CFS score and FI were calculated (Fig. B), and univariate logistic regression analysis was performed. Results A total of 115 patients were included. Baseline characteristics, findings and variables with statistical significance are resumed in Table 1. The AUC of PA to predict non-frailty patients and non-functional impaired patients was 0.87 (p<0.001) and 0. 79 (p<0.001), which corresponds to PA values of ≥4.35° and ≥3.62°, respectively. The AUC for E/I ratio to predict frailty and FI was 0.83 (p<0.0001) and 0.74 (p<0.0001), which corresponds to E/I ratio values of ≥0.97 and ≥0.99 respectively. Moderate-to-strong-negative correlation was found between PA and CFS score (r = −0.61, p<0.0001) as well as moderate-positive correlation between PA and BI (r = −0.47, p<0.0001) was found. Moderate-positive correlation between E/I ratio and CFS score (r = −0.56, p<0.0001) and moderate-negative correlation between E/I ratio with BI (r = −0.45, p<0.0001) were found. Univariate logistic regression analysis revealed that factors independently associated with frailty were age (OR = 1.04, CI-95%:1.02-1.06) and PA (≥4.35°) (OR = 0.17, CI-95%:0.11-0.24). Conclusion PA and E/I ratio obtained in a single non-invasive BIA can be used as screening tool, with sensitivity >80%, for earlier prediction of FI and frailty in HD patients allowing premature identification and intervention in this group of patients.