Risk reclassification ability of uric acid for cardiovascular outcomes in essential hypertension

Abstract

Background: Hyperuricemia is associated with incident cardiovascular events in different settings of patients. We tested whether the inclusion of uric acid (UA) in Cox models including standard risk factors allows to better stratify cardiovascular risk in a cohort of 1522 naïve hypertensives with preserved renal function.

Methods: We used multiple Cox regression models to assess the independent effect of UA on cardiovascular outcomes, and Harrell'C index, Net Reclassification Index (NRI), and Integrated Discrimination Improvement (IDI) as indicators of the additional prognostic value of UA beyond and above that provided by standard risk factors and estimated glomerular filtration rate (e-GFR). Study outcomes were fatal and nonfatal cardiovascular events and fatal and nonfatal coronary outcomes/death due to other cardiovascular events.

Results: UA resulted strongly related to both outcomes in unadjusted Cox regression analyses (P<0.001). Inclusion of UA into multiple Cox regression models including Framingham risk factors and e-GFR did not affect the association between UA and outcomes (fatal and nonfatal cardiovascular events, HR=1.44, 95% CI=1.36-1.55, P<0.001; fatal and nonfatal coronary outcomes/death due to other cardiovascular events, HR=1.48, 95% CI=1.36-1.61, P<0.001). Inclusion of UA into basic Cox models provided an increase in all indexes of prognostic accuracy for both outcomes: Harrell'C index: +5%; NRI: +24.9%; IDI: +7.6%, all P<0.001; and Harrell'C index: +5%; NRI: +25%; IDI: +6.3%, all P<0.001, respectively.

Conclusions: UA is an independent predictor of cardiovascular outcomes and increases prognostic accuracy of Cox models, including Framingham risk factors and e-GFR, in hypertensives with normal renal function, allowing a risk reclassification.

Keywords: Cardiovascular risk; Essential hypertension; Uric acid.