The Performance of a Body Composition–Based Equation in Estimating Overhydration of Hemodialysis Patients

Abstract

Achieving and maintaining optimal fluid status remains a major challenge in maintenance hemodialysis (MHD). The aim of this study was to establish a body composition–based multiple regression equation for use in the estimation of overhydration (OH) in the setting of MHD while using a body composition monitor (BCM) to guide patient dry weight management. We initially retrospectively analyzed factors associated with OH in 314 healthy Chinese individuals and obtained a multiple linear regression equation to determine OH level. Next, 49 stable MHD patients were enrolled to validate whether our multiple regression formula was applicable to such patients. Prior to hemodialysis, BCM measurements were performed; OHpre was defined as OH directly measured by BCM; while OHstd was defined as OH estimated by the multiple regression equation. In our multivariate linear regression analysis, PhA (β = − 1.266, 95% CI (− 1.532 ~ − 1.341), p < 0.001), LTM (β = 0.987, 95% CI (0.086 ~ 0.109), p < 0.001), and age (β = − 0.307, 95% CI (− 0.023 ~ − 0.015), p < 0.001) were independent predictors of OH in healthy Chinese individuals. The multiple linear regression equation that we developed for calculating OH was as follows: OHstd = 6.203 − 0.019 × age − 0.083 × gender − 0.006 × fat + 0.098 × LTM − 1.437 × PhA (F = 189.896, R2 = 0.755, p < 0.001). Linear correlation and Bland-Altman analyses were performed between OHpre and OHstd in MHD patients; correlation was found to be high (r = 0.786, p < 0.001). Bias between OHpre and OHstd was 0.45 L as assessed using 95% CI limits of agreement ranging from − 0.73 to 1.62 L. We found that our multiple regression equation formulated using data from healthy individuals provides applicable guidance for dry weight management in MHD patients.