Uric Acid Measurement Improves Prediction of Cardiovascular Mortality in Later Life


Address correspondence to Professor David Melzer, Peninsula Medical School, Barrack Road, Exeter EX2 5DW, UK. E-mail: david.melzer@pms.ac.uk



To estimate the association between uric acid and cardiovascular mortality in older adults, independent of traditional risk factors, and to estimate the risk prediction gain by adding uric acid measurements to the Framingham Cardiovascular Risk Score (FCRS).


Longitudinal observational study of two population-based cohorts.


The Established Populations for Epidemiologic Studies of the Elderly, Iowa (Iowa-EPESE) and the Third National Health and Nutritional Examination Survey (NHANES III).


One thousand twenty-eight Iowa-EPESE participants and 1,316 NHANES III participants. Selected participants were aged 70 and older without overt cardiovascular disease, renal dysfunction, or diuretic use who lived for 3 years or longer after baseline.


Outcome was age at cardiovascular death during follow-up (12–20 years). Uric acid and cardiovascular risk factors such as smoking, systolic blood pressure, diabetes mellitus, obesity, serum cholesterol, and high-density lipoprotein cholesterol were measured at baseline.


High serum uric acid (>7.0 mg/dL) was associated with male sex, obesity, lipid levels, and estimated glomerular filtration rate at baseline. Fully adjusted hazard ratios (HRs) for cardiovascular death with high uric acid versus normal were 1.36 (95% confidence interval (CI) = 1.10–1.69) in Iowa-EPESE and 1.43 (95% CI = 1.04–1.99) in NHANES III; pooled HR was 1.38 (95% CI = 1.16–1.61). The net reclassification improvement achieved by adding uric acid measurement to the FCRS was 9% to 20%.


In individuals aged 70 and older without overt CVD, renal dysfunction, or diuretic use, serum uric acid greater than 7.0 mg/dL was associated with greater CVD mortality independent of classic CVD risk factors. Adding uric acid measurement to the FCRS would improve prediction in older adults.