Risk misclassification by cardiovascular risk scores in rheumatoid disease: high prevalence of carotid plaque in patients classified as low risk
A Leal-Valdez, V Cervantes-Ramirez, F M Garcia-Garcia, E C Garza-Gonzalez, N Castro-Saenz, O A Garza-Flores, H H Garza-Sepulveda, J E Escober-Torres, I J Colunga-Pedraza, J A Cardenas-de la Garza, D A Galarza-Delgado, R Flores-Ramirez, A Ordaz-Farias, R Vera-Pineda, J R Azpiri-LopezAbstract
Background
Cardiovascular (CV) risk calculators are widely used to estimate 10-year CV risk in the general population. However, even in non-inflammatory cohorts, these models show limited sensitivity for detecting subclinical atherosclerosis, as a proportion of individuals categorized as low risk already exhibit carotid plaque (CP). Immune-mediated rheumatic diseases are characterized by chronic inflammation and accelerated atherosclerosis, potentially amplifying risk underestimation by traditional algorithms.
Purpose
To evaluate CV risk misclassification in rheumatoid disease and to compare CP prevalence among patients classified as low risk across commonly used CV risk calculators.
Methods
We conducted a cross-sectional study including 668 participants aged 18–70 years: rheumatoid arthritis (RA, n=331), psoriatic arthritis (PsA, n=85), systemic lupus erythematosus (SLE, n=50), and controls (n=202). All patients fulfilled international classification criteria. Individuals with prior ASCVD, overlap syndromes, or pregnancy were excluded.
Carotid ultrasound was performed by a certified radiologist. CP was defined as focal wall thickening ≥50% greater than the adjacent wall or intima-media thickness ≥1.5 mm. CV risk was calculated using Framingham (Lipids and BMI), QRISK3, SCORE2, and PREVENT. Participants were categorized as low versus moderate/high risk according to recommended thresholds. Misclassification was defined as CP in individuals classified as low risk. ORs were estimated for overall comparisons and RRs within the low-risk stratum.
Results
Overall CP prevalence was 35.9% (RA 40.5%, PsA 38.8%, controls 30.7%, SLE 22.0%). RA had higher odds of CP versus controls (OR 1.54, 95% CI 1.06–2.23; p=0.026); PsA showed a non-significant trend (OR 1.43, 95% CI 0.84–2.43; p=0.216).
Among individuals classified as low risk, CP remained frequent. In RA, CP prevalence was 37.4% with Framingham Lipids (RR 1.75, 95% CI 1.20–2.53), 31.5% with Framingham BMI (RR 1.55, 95% CI 1.00–2.38), and 32.2% with QRISK3 (RR 1.11, 95% CI 0.77–1.60), compared with 21.4%, 20.4%, and 28.9% in low-risk controls, respectively. In PsA, CP prevalence was 32.8% with Framingham Lipids (RR 1.53, 95% CI 0.93–2.51), 22.2% with Framingham BMI (RR 1.09, 95% CI 0.56–2.12), and 31.6% with QRISK3 (RR 1.09, 95% CI 0.67–1.79). SCORE2 low-risk estimates were 37.8% in RA and 35.1% in PsA. In SLE, PREVENT-based subanalysis showed CP in 27–29% of low-risk individuals.
Discussion
Traditional CV risk scores substantially underestimate subclinical atherosclerosis in rheumatoid disease. Approximately one-third of individuals labeled as low risk already exhibit CP, particularly in RA, where misclassification is most pronounced, with approximately 11-75% higher relative plaque prevalence compared with low-risk controls. These findings support the incorporation of vascular imaging into CV risk assessment strategies in patients with inflammatory rheumatic disease.Figure 1.CP prevalence in low risk patientsFigure 2.CP prevalence and PREVENT