Abstract
OBJECTIVE: To determine the association between cardiovascular (CV) risk factors and atherosclerosis in patients with rheumatoid arthritis (RA). METHODS: The common carotid artery intima-media thickness (IMT) and plaque were evaluated by high resolution B-mode ultrasound in 74 consecutive patients with RA. Patients with an IMT > or = 0.60 mm and plaque were considered to have atherosclerosis and advanced atherosclerosis, respectively. Traditional risk factors as well as an extensive range of other clinical and laboratory variables were recorded. Methods used to analyze the data included logistic regression, classification and regression tree (CART), and factor analyses. RESULTS: Fifty-three (72%) patients had atherosclerosis, 23 (31%) had plaque, and 21 (28%) were free of atherosclerosis. In multivariable analysis, age and hypertension were independently associated with atherosclerosis and plaque (p < or = 0.04). Radiographic scores and polymorphonuclear cell counts were also strongly associated with plaque (p < or = 0.008). Uric acid concentrations were associated with atherosclerosis, and hypothyroidism was associated with plaque, both with borderline significance (p = 0.078 and 0.052, respectively). In CART analysis, age, polymorphonuclear cell counts, and joint space narrowing in the hands were considered to be the most important determinants of plaque, and 62% of patients could be classified correctly after cross-validation. Factor analysis (varimax rotation) revealed that age and uric acid levels were related to low glomerular filtration rates, polymorphonuclear cell counts to disease activity, and radiographic scores to disease duration, and hypertension was associated with high cholesterol levels. The 10-year risk for a coronary event estimated using the Framingham risk equation (calculated from traditional risk factors) was only 7% in patients with plaque. CONCLUSION: Atherosclerosis in RA is associated with the traditional CV risk factors age and hypertension, as well as nontraditional risk factors comprising current inflammation as reflected by polymorphonuclear cell counts, cumulative inflammation as disclosed by radiographic scores, and, to a lesser extent, with uric acid levels and hypothyroidism. Multiple risk factor assessment equations that are based on traditional risk factors only are likely to be insufficient to capture CV risk extent in RA.
In this issue
