Abstract P148: C-Reactive Protein (CRP) and the Prediction of Cardiovascular Death Among Mexican-Americans
Background: Addition of C-reactive protein (CRP) improves risk stratification provided by traditional Framingham risk factors (FRFs) in some populations, but findings vary by gender and by race/ethnicity. Whether CRP levels can improve risk prediction of cardiovascular mortality (CVD) among elderly Hispanics is unknown.
Hypothesis: CRP will add incremental CVD risk prediction in addition to FRFs.
Methods: We evaluated whether addition of baseline CRP to the Framingham risk score can improve CVD risk stratification for CVD death among 1,422 participants (536 males, 886 females) from the Sacramento Area Latino Study on Aging (SALSA). SALSA is a well-characterized, NIH funded cohort study of Mexican Americans aged > 60 followed since 1998–1999. Death was ascertained by participant contacts, obituary reviews and cause of death from death certificates. We evaluated the association of CRP levels with CVD death using Cox proportional hazards models adjusting for FRFs stratified by gender. We calculated gender stratified net reclassification improvement (NRI) in models with and without CRP to predict 10-year CVD risks using the established FRFs categories (<5%, 5–10%, 10–20%, > 20%).
Results: Mean baseline age was 70.7 and mean CRP level was 5.86 ng/ml. There were 167 CVD male deaths (mortality rate 39.2 per 1,000 person years) and174 among women (mortality rate 27.8 per 1,000 person years). Higher CRP levels were significantly associated with higher CVD mortality risk in men, HR of 1.29 (95% CI 1.11 – 1.50) but not women, HR 0.94 (95% CI 0.81 – 1.08) after adjustment for covariates. Addition of CRP was useful in reclassifying CVD death risk among men defined by FRF alone, and this was due to correct reclassification of those who did not experience CVD death from the highest risk group to a lower risk category. The majority of male participants (619/745=83%) were classified as having >20%risk in the model without CRP. Mortality rate for men initially classified as having > 20% risk was 37.9 per 1,000 person years. After addition of CRP, 6% were reclassified into an intermediate risk category (NRI 7.4%, p< 0.001) and their mortality rate was 10.3 per 1,000 person years. Overall, NRI for men was 4.2% (p=0.05). In contrast, addition of CRP among women was associated with increased misclassification to a higher risk category (NRI= − 13.4%, p=0.0001).
Conclusion: A risk prediction model that includes CRP improves cardiovascular risk classification among elderly Mexican-American men but not women as defined by traditional FRFs.
- © 2012 by American Heart Association, Inc.