Pre-dialysis hypoalbuminemia: Is it nutritional, medical, or dilutional?

Most dialysis units draw serum albumin levels before hemodialysis (HD) trea tment, when patients (pts) are fluid-overloaded and not at their dry weight . During treatment, fluid is removed form the plasma by ultrafiltration (UF ), which results in weight loss. Since hemodilution can lower serum albumin levels, we wondered whether albumin levels drawn post-dialysis might be hi gher due to the removal of excess fluid by UF. The purpose of our study was to determine whether albumin levels change significantly during a single H D treatment due to fluid removal. We studied pre- and post-dialysis (pre-HD and post-HD) albumin levels and weights for 122 patients in our dialysis u nit in May 1996. Albumin was analyzed by the bromcresol green (BCG) method. Fluid removal during dialysis was calculated as the difference between the pre- and post-HD weights. Our unit's goal was to achieve serum albumin levels greater than or equal t o 4.0 g/dl. We found 1) Only 32.8% of pts had albumin levels greater than o r equal to 4.0 g/dl when levels were drawn pre-HD. Yet, when levels were re peated after treatment (post-HD), 75.4% of pts had levels that met the goal . 2) Albumin levels generally increased from an average pre-HD level of 3.8 +/- 0.39 g/dl to an average post-HD level of 4.4 +/- 0.74 g/dl (p < 0.0000 1). 3) The change in albumin level correlated strongly with weight loss (fl uid removal) (r = 0.56, p < 0.00001). 4) The nPCR, which indicates protein intake in stable patients, was more strongly correlated with post-dialysis (rather the pre-dialysis) albumin levels (r = 0.39 vs. r = 0.25 for pts wit h current nPCRs, p < 0.05). Albumin levels generally increase significantly during HD. The strong corre lation between change in albumin and weight loss supports the idea that alb umin levels tend to increase due to fluid removal by UF. In conclusion, it may be inappropriate to use pre-dialysis serum albumin as an indicator of n utritional status, since levels drawn at that time are significantly altere d by the effect of excess fluid. When pre-dialysis hypoalbuminemia is obser ved, it may be important to ask whether the cause is nutritional, medical, or simply dilutional.