Associations between bioelectrical impedance analysis‐derived phase angle, protein‐energy wasting and all‐cause mortality in older patients undergoing haemodialysis

Sho Kojima,Naoto Usui,Akimi Uehata,Akihito Inatsu,Atsuhiro Tsubaki
DOI: https://doi.org/10.1111/nep.14333
2024-06-12
Nephrology
Abstract:Summary at a glance Phase angle (PA) is associated with protein‐energy wasting (PEW) patients undergoing haemodialysis. However, PA needs to be considered with sex and age. This study revealed sex‐specific cut‐off values of PA predicting PEW in older haemodialysis patients. PA may be useful predicting PEW and mortality in older haemodialysis patients. Aim Protein‐energy wasting (PEW) is a common syndrome in patients undergoing haemodialysis (HD) and is associated with poor prognosis. Bioelectrical impedance analysis (BIA)‐derived phase angle (PA) is useful for predicting PEW, but sex and age need to be considered. We aimed to reveal sex‐specific cut‐off values of PA predicting PEW in HD patients aged ≥65. Methods This two‐centre retrospective cohort study included patients on HD who underwent BIA. PEW was detected using the International Society of Renal Nutrition and Metabolism (ISRNM) criteria as a reference. The PA was measured using a multifrequency bioimpedance device. Sex‐specific cut‐off values of PA predicting PEW were detected by receiver‐operator characteristic analysis. We investigated the association between PEW determined using sex‐specific cut‐off values for PA and all‐cause mortality. Results This study included 274 patients undergoing HD, with a median age of 75 (70–80) years, mean PA of 3.8 ± 1.1° and PEW of 43%. Over a median follow‐up duration of 1095 (400–1095) days, 111 patients died. Cut‐off values of PA predicting PEW were as follows: female, 3.00° (sensitivity, 87.3%; specificity, 77.5%), and male, 3.84° (sensitivity, 77.6%; specificity, 71.4%). The kappa coefficient between sex‐specific cut‐off values of the PA and ISRNM criteria had a moderate coincidence level of 0.55. PEW detected by PA was independently associated with all‐cause mortality (hazard ratio: 2.40; 95% confidence interval: 1.51–3.85; p
urology & nephrology
What problem does this paper attempt to address?