Influence of nonclassical cardiovascular risk factors on the accuracy of predicting subclinical atherosclerosis in rheumatoid arthritis.

Abstract

To determine whether nontraditional risk factors increase the accuracy of predicting the presence of carotid artery plaque based on traditional cardiovascular risk factors only in patients with rheumatoid arthritis (RA). We identified risk factors that were independently associated with ultrasonographically located plaque. In predicting carotid artery plaque, the area under the curve (AUC) of the receiver-operating characteristic (ROC) curve for the combination of traditional and nontraditional risk factors was compared to the AUC of the ROC curve for traditional risk factors and nontraditional risk factors considered separately in 91 patients with RA. Thirty-one (34%) patients had carotid artery plaque. The 3 traditional risk factors of age > 55 years, hypertension, and ever-smoking, and the 3 nontraditional risk factors of a disease duration > 8 years, polymorphonuclear cell count > 4.5 x 10(6)/l, and hypothyroidism were each independently associated with the presence of plaque (odds ratios 2.08-8.78; p = 0.001-0.02). The percentage of patients with plaque was 0, 10%, 50%, and 83% in patients with 0-1, 2, 3, and 4-6 of these risk factors, respectively. In predicting plaque, the AUC of the ROC curve for the combination of traditional and nontraditional risk factors (0.90 +/- 0.03) was greater than that for either traditional (0.80 +/- 0.05; p = 0.006) or nontraditional (0.80 +/- 0.04; p = 0.005) risk factors considered separately. The combination of disease duration, polymorphonuclear cell counts, and thyroid status increased the accuracy of predicting subclinical atheroma in patients with RA. We believe that our findings merit external validation.