BackgroundMalnutrition affects 1/3 of hemodialysis patients and associates with a higher risk of morbidity and mortality. Serum albumin is a marker of nutrition and inflammation, and predicts mortality, particularly when <3.8 g/dL. This study was performed to identify risk factors for hypoalbuminemia (<3.8 g/dL) and the particular temporal relationship and strength of association between protein intake (nPCR) and serum albumin when confounding variables are taken into account.MethodsDemographic, clinical, and dialysis-related data of 57 ESRD patients on hemodialysis over 24 months in 6 temporal segments were analyzed with serum albumin as a continuous, and categorical outcome (with 3.8 g/dl as cut-off) variable , against 13 potential independent variables [4 demographic factors, 3 nutrition-related, and 5 morbidity-related parameters, and % urea reduction ratio (URR)]. The temporal relationships between albumin and nPCR were analyzed for the concurrent & 3 subsequent months in each temporal segment.ResultsThe impact of nPCR on serum albumin (p < 0.05) was significant but with no discernible temporal relationship. Advancing age, longer vintage, female gender, diabetes mellitus, nPCR, serum phosphate and ferritin had significant correlation with albumin <3.8 g/dl (p < 0.05). Serum phosphate levels correlated positively, and fever, bacteremia, hospital stay and weight loss negatively, with mean serum albumin but did not negate the effect of nPCR. Regression analysis showed that mean albumin associated with nPCR, fever, hospital stay, bacteremia, dialysis vintage, age, sex, and diabetes mellitus; and that an albumin level of <3.8 g/dl associated with age, female sex, diabetes, lower nPCR, and higher ferritin.ConclusionSuggested target albumin levels were not met in elderly, female, and diabetic patients. The association of nPCR with albumin was not nullified by confounding demographic or morbidity-related factors. nPCR had no demonstrable temporal relationship with albumin.