L. D. Browne 1,2, O. O'cealligh 1, E. Quinn 1, C. Borghi 3, A. G. Stack 4
¹ Graduate Entry Medical School, University of Limerick, Limerick, Ireland; ² Health Research Institute, University of Limerick, Limerick, Ireland; ³ Dipartimento di Scienze Mediche E Chirurgiche, Università di Bologna , Bologna, Italy; 4 Department of Nephrology, University Hospital Limerick & Health Research Institute, University of Limerick, Limerick, Ireland
Introduction and Aims: Observational studies have yielded inconsistent results regarding the association of serum uric acid (UA) with mortality. The relationship appears to vary across different clinical populations and the threshold values at which mortality increases are poorly understood. The objective of this systematic review and meta-analysis was to investigate the risk of mortality associated with UA in the general population, and in patients with cardiovascular disease (CVD), chronic kidney disease (CKD) and end stage kidney disease (ESKD).
Methods: We searched electronic databases up to 1 July 2018 for observational studies reporting associations for three or more groups of serum UA with all-cause mortality in the general population and among patients with either CVD, CKD or ESKD. Study-specific associations between serum UA and adjusted relative risks (RR) were estimated using restricted cubic splines with three knots at 10th, 50th and 90th percentile of the UA distribution and a generalised least squares method before pooling study estimates with a multivariate random-effects meta-analysis.
Results: We included 3,034,994 participants from 49 cohorts who experienced 117,597 all-cause deaths. The overall pattern of association between serum UA and all-cause mortality was non-linear (p-value for non-linearity < 0.001). The mortality risk increased for UA concentration above 6.0 mg/dL [RR: 1.02 (1.01-1.03)] ,with an almost linear increase in risk for higher concentrations (7.0 mg/dL, [RR: 1.10 (1.06-1.14)]). There was evidence of heterogeneity across studies (I² = 85.5). The shape of the UA-mortality association was similar for participants in the general population, those with CKD or CVD but differed significantly from those with ESKD (p<0.001). For ESKD patients the pattern was U-shaped, with the lowest mortality risk observed at UA 7.6 mg/dL [RR 0.55 (0.35-0.86)].
Conclusions: With the exception of patients with ESKD, UA has a J-shaped association with all-cause mortality with increasing risk above UA of 6 mg/dL. Large randomised clinical trials of urate-lowering therapy should test whether targeting this threshold will confer a survival benefit.