ABC | Volume 110, Nº5, May 2018

Original Article Salmazo et al Frequency of subclinical atherosclerosis in Brazilian HIV-infected patients Arq Bras Cardiol. 2018; 110(5):402-410 Figure 4 – Frequency of plaques in HIV-positive patients according to risk stratification by Framingham risk score. 100 80 60 40 20 0 Percentage of carotid plaque (%) Intermediate/high Low Framingham risk category 62.8% 18.2% Plaque - yes Plaque - no p < 0.001 Other studies have reported similar results, suggesting that the increased risk of atherosclerosis in the presence of HIV is not directly associated with time since diagnosis, but with conditions involved in HIV infection. 10,38,39 Patients with PL were 11 years older than those without PL, and predominantly male. Fasting glucose, TC, LDL-c, and TGL were significantly higher in patients with the diagnosis of PL. In patients with PL compared to patients without PL, SBP and DBP were 10 and 6 mm Hg higher, respectively. These results agree with those of other studies and strengthen the concept that atherogenesis in HIV‑infected patients follows the classical risk factors described in other populations. 4,20,37,38 Our results indicate that older individuals had higher CIMT, regardless of the presence of HIV infection; however, there was an interaction between age and the presence of HIV to increase CIMT (p < 0.001). Although time since diagnosis did not affect PL frequency, HIV infection appears to enhance the effect of age on CIMT. In this case, younger individuals with HIV could have vascular changes compatible with those of older patients. Understanding this behavior is important in screening for atherosclerosis in HIV-infected patients, because the protective effect of lower age would have less relevance. In the HIV group, CIMT was associated with age (p < 0.001), BMI (p = 0.053), LDL-c (p = 0.005), and serum creatinine (p = 0.004); this was also seen in other studies. 17-19,37,40 There were no associations with gender, smoking, diabetes, hypertension, statin therapy, HDL-c, or TGL. Interestingly, vessels may have PL with normal CIMT, which means that increased intima-media thickness and PL are not necessarily directly associated processes. However, both reflect the presence of endothelial dysfunction and are considered to favor cardiovascular events. 17-23 Treatment with PI showed significant interaction with age and time since HIV diagnosis to increase TGL. Moreover, PI exposure was not associated with higher frequency of PL, in accordance with recent studies. 3,5,6,8,10,15,37-39 In our study HAART showed correlation with unfavorable lipid profile, but without interfering in PL frequency or arterial stiffness. The PWV was directly associated with age, CIMT, and SBP. These results are consistent with those from recent studies describing arterial stiffness indexes associated with age, hypertension, and vascular disease. 29-31 In addition, AIx showed a direct association with age, CIMT, and SBP, with no interaction between age and smoking to increase AIx. The elevation of PWV and AIx in patients with PL suggests that atherosclerosis is associated with functional alterations in the vessels; stiffer vessels have a higher risk of developing PL. Furthermore, patients with PL showed higher CIMT than patients without lesions, supporting the hypothesis that CIMT and atherosclerosis are associated, even when excluding the evolutionary nature of an alteration in the other. This study included 207 patients classified as low risk (82.1%), 31 as moderate (12.3%) and 14 as high risk (5.56%) according to FCRS. The literature shows that the greater the FCRS, the greater the CIMT. 23 In relation to age, younger patients were seen to have lower scores. This is in accordance with the concept that the FCRS, when applied to young individuals, can result in a low risk score, without implying that these individuals are not at risk of future cardiovascular events. It is important to note that almost 20% of the low-risk patients had carotid PL. Patients in the moderate/high-risk subgroup showed unfavorable lipid profile with low HDL-c and elevated TC and LDL-c, as described in the literature, but no difference in hs-CRP, when compared to those classified as low risk by the FCRS. 12,15,25,26 In addition, those classified as moderate/high‑risk had higher intima-media thickness and PWV (p < 0.001), consistent with the hypothesis of a higher chance of vascular disease in the group. Plaques were detected in 16% of the patients who were not treated with PI and had LDL-c lower than 130 mg/dL. 407