Variability of laboratory parameters is associated with frailty markers and predicts non-cardiac mortality in hemodialysis patients.

BACKGROUND: The levels of many laboratory parameters are associated with the outcomes of dialysis patients, but the significance of their variability has not been well studied.
METHODS: A total of 384 patients receiving stable hemodialysis treatment during 2002 were followed up for mortality until the end of 2013. The within-patient coefficients of variation (CV) were calculated for 13 laboratory parameters from 1 year of data. We defined variability as CV and analyzed the survival of the patients according to the baseline CV values of each parameter by proportional hazard modeling.
RESULTS: During the 11-year observation period, 125 patients died. Higher CV levels for eight parameters, namely, blood urea nitrogen (BUN), sodium, hemoglobin, creatinine, total protein, albumin, potassium and phosphate, were significantly associated with all-cause mortality. The adjusted hazard ratios for a high BUN-CV (>15 %) and a high Na-CV (>1.3 %) against a lower CV were 1.92 (95 % CI 1.31-2.81) and 1.95 (1.36-2.80), respectively. The increased mortality risk associated with each variability was attributed to excess non-cardiac deaths. The CV values of most parameters were correlated with each other and often exhibited negative associations with age, diabetes, and mobility as well as the levels of hemoglobin, albumin, creatinine, Na, the protein catabolic rate, and the creatinine generation rate. Therefore, a high variability was generally associated with frailty-related adverse prognostic factors.
CONCLUSIONS: The variability of several blood parameters had a significant impact on all-cause and non-cardiac mortality. The levels of the variabilities were most likely related to poor physical conditions of the patients.