The additive prognostic value of C-reactive protein for all-cause and cardiovascular mortality across the PREVENT-defined cardiovascular risk continuum: a nationally representative cohort study
V Jaiswal, A Bouras, A Jaiswal, A Ahmed, M J Ricciardi, M Banach, A QamarAbstract
Background
High-sensitivity C-reactive protein (hsCRP) is a recognized cardiovascular risk-enhancing factor, yet whether its prognostic value varies across the spectrum of baseline cardiovascular risk defined by the American Heart Association (AHA) PREVENT equations remains unknown.
Purpose
To evaluate the additive prognostic value of hsCRP for all-cause and cardiovascular mortality across PREVENT-stratified risk groups in a nationally representative US cohort.
Methods and Results
We analyzed 14,124 adults aged 30 to 79 years without baseline atherosclerotic cardiovascular disease (ASCVD) from the National Health and Nutrition Examination Survey (NHANES) 2003 to 2016, with mortality follow-up through December 31, 2019 (mean follow-up 10.2 years). Participants were stratified by PREVENT 10-year cardiovascular disease risk (low, borderline-intermediate, and high) and hsCRP category (less than 1, 1 to 3, greater than 3 mg/L). All-cause mortality was assessed using multivariable Cox proportional hazards models, and cardiovascular (CVD) mortality was assessed using cause-specific Cox models. During follow-up, 1,364 all-cause deaths (9.7%) and 335 CVD deaths (2.4%) occurred. For all-cause mortality, hsCRP greater than 3 mg/L was independently associated with increased risk in the borderline-intermediate risk group (hazard ratio [HR] 1.43, 95% confidence interval [CI] 1.18 to 1.74, P = 0.0003) and the low risk group (HR 1.39, 95% CI 1.06 to 1.83, P = 0.018), but not in the high-risk group (HR 1.25, 95% CI 0.97 to 1.62, P = 0.085). For cardiovascular mortality, hsCRP greater than 3 mg/L was significantly associated only in the borderline-intermediate risk group (HR 1.46, 95% CI 1.05 to 2.01, P = 0.023). Adding hsCRP to the base covariate model improved discrimination in the borderline-intermediate group (change in C-statistic 0.0064, 95% CI 0.0014 to 0.0136) and the high risk group (change in C-statistic 0.0108, 95% CI 0.0010 to 0.0248), but not in the low risk group.
Conclusion
Elevated hsCRP confers additive prognostic value for mortality primarily among adults at borderline-intermediate PREVENT-defined cardiovascular risk, the population segment where clinical decision-making regarding preventive therapy is most uncertain. These findings support the targeted use of hsCRP as a risk-enhancing factor in this population.