Association of Serum Uric Acid Level and All-Cause Mortality Among Dialysis Patients in Taiwan

BACKGROUND: Uric acid (UA) has complex characteristics in metabolic function of chronic kidney disease. High UA level is associated with gout formation and high-purine food intake, and it is elevated in acute kidney injury by two different mechanisms recognized as hyperuricemia and uricosuria. In contrast, UA serves an important beneficial role in the elimination of nitrogenous compounds. Previous studies showed debating results. A study showed that UA level had negative correlation of all-cause mortality among hemodialysis (HD) patients. Another large population study showed a U-shaped correlation. Our goal is to determine the relationship between serum UA level with all-cause mortality among uremia patients receiving HD and peritoneal dialysis (PD) in Taiwan. METHODS: A total of 49,557 dialysis patients received long-term dialysis (more than 3 months) in Taiwan Renal Registry Data System (TWRDS) were recruited, after exclusion of those lacking information of iron and UA. Other exclusion criteria included age (< 20 or, > 90 years old), renal transplant and malignancy. We used cox regression model for data analysis and adjustment for underlying diseases like diabetes mellitus, hypertension, previous myocardial infarction and coronary artery disease. Furthermore, we also compared intact parathyroid hormone (iPTH), iron profile, phosphate and normalized protein catabolic rate (nPCR) level in both HD and PD patients in high (> 6.5 mg/dL) and low (< 6.5 mg/dL) UA levels. RESULTS: A total of 43,863 HD patients and 4,203 PD patients were enrolled. In the PD group, the mean age was 53.75 ± 14.67, 46.28% of them were male and the mean UA level was 7.20 ± 1.53 mg/dL. In the HD group, the mean age was 61.92 ± 13.54, 52.18% of them were male, and the mean UA level was 7.22 ± 1.51 mg/dL. Hazard ratios (HR) of mortality among age > 65, male sex, and PD were 1.06, 1.22, 1.41, respectively. Furthermore, HRs of mortality among UA level between 3.0-3.9, 4.0-4.9mg/ dL were 1.53, 1.31, respectively, even after adjusted with underlying diseases, hemoglobin, albumin and nPCR. HRs of the low UA groups were higher than one compared with those whose UA level was above 6.0 mg/dL. In addition, there is no obvious difference of iPTH, iron profile and phosphate level between the high and low UA groups. CONCLUSION: Generally, the relationship of UA and all-cause mortality is J-shaped after exclusion of whose UA level above 10 mg/dL. Iron and phosphate did not interact with UA. Further investigation is warranted for the role of UA in metabolism especially in dialysis patients.