Background:  Measurements of carotid intima-media thickness (cIMT) and screening for coronary artery calcium (CAC) have previously been applied to HIV-infected populations and are increasingly used for CV risk stratification and as surrogate end points of CVD. Most studies in HIV infection have been limited to cross-sectional analysis or short follow-up times. We assessed the predictors of cIMT and CAC progression in a large and well-described longitudinal cohort of HIV-infected adults.    

Methods:  Common cIMT, CAC scores, vascular and HIV risk factors were evaluated at baseline and at 3-year follow-up in 239 HIV-infected adults, whose mean age was 45 ± 7 years, 27% were female and 47% non-white. Multivariate regression was used to determine predictors of cIMT and CAC progression. 

Results:  The mean change in cIMT per year of follow-up was 0.016 ± 0.031 mm. Yearly progression of cIMT was significantly higher in men than in women (P =0.03), but did not vary by lipid and glucose abnormalities, inflammatory markers, immunologic, and virologic status, exposure to ART or smoking. Significant predictors of yearly cIMT progression were age, diastolic BP, triglycerides, ApoB, and insulin; 28% had CAC progression. Of those with zero CAC at baseline, 32% had detectable scores at follow-up. Of those with detectable CAC at baseline, 26% had progression at follow-up. More men than women (31% vs 22%) and more with an intermediate/high than low Framingham risk scores (42% vs 25%) had CAC progression, but other risk factors were not different between those with and without CAC progression. For CAC score, age, weight and ApoE predicted progression. CIMT and CAC progression were correlated (P =0.03).

Conclusions:  While cIMT and CAC progression rates in HIV-infected patients appear higher than expected for this age group, traditional CV risk factors remain the strongest determinants of carotid and coronary atherosclerotic disease progression in HIV-infected patients. Technical optimization and stringent standardization of these measurements are required and may explain much of the heterogeneity of associations found in other studies. However, both modalities are validated and complementary measures of CV burden and can be applied to refine CVD risk scores to help avert future coronary events in HIV-positive adults.