Abstract
BACKGROUND: Plasma volume (PV) calculated from hematocrit and body weight has applications in cardiovascular disease. The current study investigated the validity of the calculated PV for predicting volume overload and its prognostic utility in patients undergoing hemodialysis (HD). PATIENTS AND METHODS: Fifty-four HD patients were prospectively enrolled, and their actual PV (aPV) and relative PV status (PVS) were calculated. Bioelectrical impedance analysis (BIA) with assessment of and total body water (TBW), intracellular water (ICW), extracellular water (ECW), and overhydration (OH) and routine blood examinations were performed before dialysis. A second cohort of 164 HD patients was retrospectively enrolled to evaluate the relationship between the calculated PVS and the outcome, with an endpoint of all-cause mortality. RESULTS: aPV was significantly associated with TBW, ICW, ECW, OH, and ECW/TBW (all p < 0.001), and most strongly with ECW (r = 0.83). aPV predicted the extent of volume overload with an AUC of 0.770 (p < 0.001), but PVS did not (AUC = 0.617, p = 0.091). Median follow-up time was 53 months, during the course of which 60 (36.58%) patients died. Values for PVS (12.94 ± 10.87% vs. 7.45 ± 5.90%, p = 0.024) and time-averaged PVS (12.83 ± 11.20 vs. 6.78 ± 6.22%, p < 0.001) were significantly increased in patients who died relative to those who survived. A value of time-averaged PVS >8.72% was significantly associated with an increased incidence of all-cause mortality (HR = 2.48, p = 0.0023). CONCLUSIONS: aPV was most strongly associated with ECW measured using BIA. HD patients with higher time-averaged PVS had a higher rate of all-cause mortality.