see all topics...

Journal of the American Society of Nephrology,
September 1996, Volume 7, #9.

Bioelectrical impedance analysis predicts survival
in hemodialysis patients.

G. M. Chertow, E. G. Lowrie, N. L. Lew, and J. M. Lazarus.

Renal Division, Brigham and Women's Hospital, Harvard Medical School, Boston, MA and National Medical Care, Inc., Waltham, MA.

Abstract: We performed a cross-sectional study of bioelectrical impedance analysis in 3009 hemodialysis (HD) patients with follow-up to 12 (median 8) months to evaluate the relation between impedance and survival. 296 deaths were documented over the study period; patients were censored at transplantation or if lost to follow-up. Mean ( SD) age was 60.5 15.5 years, 47% were females, 47% African-American, and 36% diabetic. Dialysis vintage was 3.8 3.7 years. Mean reactance (Xc) was 40.9 13.8 ohms, mean resistance (R) was 497.7 99.0 ohms. The mean Xc/R ratio 0.084 0.030. Xc/R was directly correlated with albumin (r = 0.24), prealbumin (r = 0.27), and creatinine (r = 0.40), and inversely correlated with age (r = -0.33) and Quetelet's index (r = - 0.15). Xc/R was significantly higher in men, Africa Americans, and non-diabetics (P < 0.0001 compared with women, Caucasians, and diabetics, respectively). To obviate linearity assumptions, Xc/R was ranked into quintiles, and the relative risks of death (RR) and 95% confidence intervals (95% CI) within each quintile were estimated from proportional hazards regression, with and without covariate adjustment. Survival was similar in patients within the upper three quintiles of Xc/R. The RR in the two lowest quintiles was 2.7 (95% CI 1.1 to 2.2, P = 0.009), respectively. After adjustment for age, sex, race, diabetes, albumin, creatinine, and urea reduction ratio, the RR in the lowest quintile was 1.5 (95% CI 1.2 to 2.1, P = 0.003). Results were similar when Xc and R were adjusted for body stature.

Conclusion: Xc/R is correlated with biochemical surrogates of nutritional status in hemodialysis patients, and provides prognostic power even after adjustment for case mix and laboratory variables. Longer-term follow-up, and longitudinal assessment of Xc/R will be required to determine the optimal role for BIA in HD patient assessment.
copyright © 1998 - biodynamics corporation