Serum uric acid, protein intake and mortality in hemodialysis patients

Background. The association between serum uric acid (SUA) and mortality has been conflicting among studies using hemodialysis (HD) patients. Given the close link between purine and protein in foods, we hypothesized that normalized protein catabolic rate (nPCR), a dietary protein intake surrogate, modifies the SUA-mortality association in the HD population. Methods. We identified 4298 patients who initiated HD and had one or more SUA measurement in a contemporary cohort of HD patients over 5 years (1 January 2007-31 December 2011), and examined survival probability according to the first uric acid measurement, adjusting for dialysis vintage, case-mix and malnutrition-inflammation complex-related variables. Results. Mean SUA concentration was 6.6 +/- 1.8 mg/dL. There was a consistent association of higher SUA with better nutritional status and lower all-cause mortality irrespective of adjusted models (P-trend < 0.001). In the case-mix adjusted model, the highest SUA category (>= 8.0 mg/dL) compared with the reference group (> 6.0-7.0 mg/dL) showed no significant mortality risk [hazard ratio (HR) 0.90, 95% confidence interval (CI) 0.72-1.13], while the lowest category (< 5.0 mg/dL) was associated with higher mortality (HR 1.42, 95% CI 1.16-1.72). The hypouricemia-mortality association was significantly modified by nPCR (P-interaction = 0.001). Mortality risk of low SUA (< 5.0 mg/dL) persisted among patients with low nPCR (< 0.9 g/kg/day; HR 1.73, 95% CI 1.42-2.10) but not with high nPCR (>= 0.9 g/kg/day; HR 0.99, 95% CI 0.74-1.33). Conclusions. SUA may be a nutritional marker in HD patients. Contrary to the general population, low but not high SUA is associated with higher all-cause mortality in HD patients, especially in those with low protein intake. Nutritional features of SUA warrant additional studies.