Background: HIV patients are at higher risk for atherosclerosis. This risk has been attributed to effects of ART, immunodeficiencyor HIV-associated inflammation. To define the pathogenesis of HIV-associated atherosclerosis, we studied carotid intima-media thickness (IMT) in a unique group of HIV-infected patients who fully control HIV replication in the absence of therapy (HIV controllers). These patients were compared to treated patients both with and without detectable viral loads, and to uninfected controls.
Methods:  We measured lipids and carotid IMT by high resolution ultrasound in 363 HIV patients and 92 controls. HIV patients were stratified based on treatment (yes/no) and viral load (detectable/undetectable). We adjusted for traditional risk factors and HIV characteristics including duration of exposure to ART. The primary outcome was mean maximal IMT of 12 pre-selected segments.
Results: The median age was 47 years (IQR 41 to 53); 84% were male. The median IMT was higher in each of the 4 HIV-infected sub-groups than for uninfected controls (see the table); these differences remained significant after controlling for traditional risk factors (p ≤0.001 for each adjusted pair-wise comparison). Among HIV patients with an undetectable viral load, treated patients had a 0.12-mm higher IMT than untreated patients after adjusting for traditional risk factors (p = 0.03). Increasing duration of protease inhibitor use was associated higher carotid IMT (Spearman's ρ: 0.21, p <0.001).
Conclusions:  After adjustment for traditional risk factors, both HIV infection and duration of exposure to protease inhibitors were independently associated with higher levels of subclinical atherosclerosis. The treatment-independent effect of HIV infection on IMT appeared to be due to factors other than HIV replication or advanced immunodeficiency, as evidenced by high IMT in our HIV controllers.

* p values are for each HIV sub-group compared to controls