Serum albumin and mortality after renal transplantation

The incidence, causes, and consequences of hypoalbuminemia after renal transplantation are not well defined. We examined clinical correlates of serum albumin measured at 3 months, 6 months, 1 year, and annually thereafter in 706 renal transplant recipients who survived at least 6 months with a functioning allograft. Follow-up was 7.0 +/- 4.2 years. Hypoalbuminemia (< or = 3.5 g/dL) was most common at 3 months (31%, n = 692), least common at 1 year (12%, n = 656), and then became increasingly common among survivors, for example, 14% (n = 466) at 4 years, 20% (n = 204) at 8 years, and 29% (n = 77) at 12 years after transplantation. By multiple linear regression, variables that correlated (P < 0.05) with lower serum albumin at 3, 6, 12, and 24 months included age, diabetes, proteinuria, and cytomegalovirus infection. Other independent correlates on at least one of these occasions included renal function and chronic disease (malignancy, liver disease, and cardiovascular disease). Serum albumin, as a time-averaged and time-dependent covariate, was a strong independent risk factor for death using Cox proportional hazards analysis (relative risk for each g/dL increment, 0.26; 95% confidence interval, 0.16 to 0.44 [1.00 = no risk]). The effects of albumin on mortality were independent of age, diabetes, serum lipids, renal function, chronic liver disease, malignancies, and cardiovascular disease. The effects of albumin on mortality were evident even when the analysis was restricted to patients dying several years after albumin was measured. Thus, hypoalbuminemia is common and serum albumin is a strong independent risk factor for all-cause mortality after renal transplantation.