Randomized trial of high-flux vs low-flux haemodialysis: effects on homocysteine and lipids.

BACKGROUND: Uncontrolled studies have found that high-flux haemodialysis favourably modifies homocysteine and lipid profiles. We sought to confirm these findings by carrying out a randomized prospective comparison of high-flux and low-flux polysulphone in chronic, stable dialysis patients. METHODS: Forty-eight patients were randomly assigned to either high or low-flux dialysis for 3 months. Serum levels of homocysteine, lipoprotein (a), and lipids were compared between the treatment groups at monthly intervals. RESULTS: All patient characteristics and laboratory variables were equally distributed between the groups at baseline. Over the study duration, we observed no differences between high- and low-flux treatment groups for the following outcomes: pre-dialysis homocysteine, lipoprotein (a), total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides (all P>0.05). Geometric mean (interquartile range) homocysteine at baseline was 20.0 (16.8-24.5) and 19.5 (15.3-22.0) micromol/l for the high-and low-flux groups respectively (P=0.80), and levels did not change significantly during the study. We did demonstrate a more pronounced intradialytic effect of high-flux dialysis on homocysteine levels, which fell during dialysis by 42%, compared to 32% with low-flux dialysis (P<0. 001). CONCLUSIONS: In this randomized controlled trial, the effects of high-flux and low-flux haemodialysis on homocysteine and lipid profiles were comparable. The greater intradialytic effect of high-flux dialysis on homocysteine did not translate into a significant difference in pre-dialysis levels after 3 months of study.