Serum uric acid, mortality and glucose control in patients with Type 2 diabetes mellitus: a PreCIS database study

Aims To determine the association of serum uric acid with all-cause mortality and hyperglycaemia in patients with Type 2 diabetes. Methods Retrospective cohort analysis of 535 consecutive patients who had uric acid determinations between 1998 and 2004 and whose subsequent vital status was determined at a median of 4.5 years. The association with mortality was analysed with Cox proportional hazards models. The incremental predictive value of uric acid was examined with concordance indexes. The proportional risk of mortality was represented with the Kaplan-Meier survival curves by uric acid quartiles. Results We studied 370 men and 165 women aged 59.3 +/- 11.5 years. Mean uric acid was 371.7 +/- 106.2 mu mol/l. Patients with glycated haemoglobin (HbA(1c)) >= 9% had lower uric acid vs. the rest (342.2 +/- 112.1 vs. 383.5 +/- 106.2, P = 0.002). Overall mortality was 10.8%. For each 59 mu mol/l increase in uric acid there was a 41% increase in risk of death (unadjusted analysis). The association of uric acid with mortality remained after adjustment for covariates (hazard ratio = 1.21, 95% confidence interval 1.07-1.45) and after gender subanalyses. Uric acid increased the accuracy of prediction when added to a model including Framingham risk factors, components of metabolic syndrome and fibrinogen (P = 0.03). Mortality was higher in patients taking diuretics vs. the rest (15.9 vs. 7.3%), but uric acid predicted mortality in both subgroups. Conclusions Serum uric acid predicts mortality in Type 2 diabetic patients regardless of gender, HbA(1c), renal function and diuretic use. Intervention studies should determine whether uric acid is a potential therapeutic target or only a marker of mortality risk.