Normalization of protein intake by body weight and the associations of protein intake with nutritional status and survival

Objective: To examine the associations of total protein intake (TPI) and dietary protein intake (DPI) with baseline nutrition and subsequent mortality. Design: Retrospective analysis of incident dialysis patients. Setting: National cohort from The United States Renal Data System data. Patients: Incident dialysis patients (n = 5,059) with blood urea nitrogen (BUN) and urea clearances reported on form 2728. Methods: TPI was calculated from BUN and urea clearance. DPI was defined as TPI divided by weight. Urinary creatinine (UCr) calculated from creatinine clearance and serum creatinine was used as a marker of muscle mass. The associations of TPI and DPI with each of serum albumin <= 3.3 g/dL (50th percentile), UCr <= 0.56g/d (25th percentile), body mass index < 18.5, and death were examined. Results: Compared with patients in the highest quartile of TPI (> 60.2 g/d), those in the lowest quartile (<= 32.4 g/d) had 1.89-fold higher odds (P < .001) of low serum albumin, 10.22-fold higher odds (P < .001) of low UCr, and 3.83-fold higher odds (P < .001) of low body mass index in multivariable logistic regression models, and an 18% increase (P < .001) in hazard of death. Compared with patients with DPI > 1.2 g/kg/d, those with DPI < 0.8 g/kg/d had nonsignificantly higher odds of low serum albumin, 2.38-fold higher odds (P < .001) of low UCr, and 0.44-fold lower odds (P < .001) of low body mass index, and a 15% (P = .04) decrease in hazard of death. Conclusions: Higher TPI is associated with better nutrition at baseline and subsequent survival. Normalization of TPI by body weight provides contradictory information on nutritional status as well as survival. (c) 2005 by the National Kidney Foundation, Inc.