Malnutrition (subjective global assessment) scores and serum albumin levels, but not body mass index values, at initiation of dialysis are independent predictors of mortality: a 10-year clinical cohort study.

OBJECTIVE To examine the associations between demographic, clinical, lifestyle, and nutritional parameters at the start of dialysis and mortality, including the combined effects on nutritional parameters, which were seldom investigated in the literature. DESIGN Ten-year retrospective clinical cohort study. SETTING Dialysis unit of a metropolitan tertiary teaching hospital in Sydney, Australia. SUBJECTS Incident dialysis patients (n = 167; hemodialysis, 57.5%; male, 61.7%; age, 65.3 ± 13.6 years; diabetic, 24.5%) who commenced on a planned dialysis program. METHODS Associations were examined between all-cause mortality and baseline demographics, including age and gender; clinical and lifestyle characteristics, including glomerular filtration rate, smoking habits, presence of comorbidities (e.g., coronary artery disease, diabetes mellitus, and peripheral vascular disease); and nutritional parameters, including body mass index (BMI), serum albumin (s-albumin) levels, and subjective global assessment score (SGA). Associations with combination values for malnutrition, s-albumin (<3.3 vs. ≥3.3 g/dL), and BMI (<26 vs. ≥26 kg/m(2)) were also examined. RESULTS Median survival was 54.2 months (interquartile range, 23 to 83), and 52.1% of patients were malnourished (SGA score B and C) at the start of dialysis. Advanced age (classified as >65 years, P < .0001), presence of peripheral vascular disease (P < .0001), reduced s-albumin levels (P = .01), and malnutrition scores (P = .02) independently predicted mortality. Being overweight and obese (BMI: ≥26 kg/m(2)) did not show any advantage on survival (P = .73). Being malnourished and overweight (or obese) was associated with a 3-fold increase in mortality risk (adjusted hazard ratio [HR], 2.96; 95% confidence interval [CI], 1.12 to 7.33; P = .02) compared with being well nourished with a BMI <26 kg/m(2) (referent). Compared with being well nourished (SGA = A), being malnourished with normal or low s-albumin was associated with higher risk (HR, 2.06; 95% CI, 1.06 to 4.00; P = .03 and HR, 2.86; 95% CI, 1.65 to 4.94; P < .0001, respectively). There was no statistical difference between mortality risks through any combination of s-albumin and BMI values (P = .54). CONCLUSION Malnutrition and reduced s-albumin levels were found to be independent predictors of mortality, whereas being overweight and obese did not show protective effects.