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 Table 1 – Demographic and clinical variables of HIV-infected patients and the control group. Variables HIV group (n = 264) Control group (n = 279) p Age (years) 43.2 ± 10.5 37.9 ± 11.5 < 0.001 Sex (F/M) 125/139 144/135 0.321 O_ob (yes/no) 135 (51,1%)/129 176 (63.1%)/103 0.005 SAH (yes/no) 28/236 23/256 0.360 Smoking (yes/no) 115 (43.6%)/149 45(16.1%)/234 < 0.001 Diabetes (yes/no) 10/254 6/273 0.263 BMI (kg/m 2 ) 25.5 ± 4.5 27.4 ± 5.4 < 0.001 SBP (mm Hg) 121 (111;133) 120 (110;130) 0.535 DBP (mm Hg) 77 (71;85) 80 (70;80) 0.616 F: female; M: male; O_ob: overweight/obesity; SAH: systemic arterial hypertension; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure. CD4 + counts ranged from 442 to 16,338 cells/µL (median: 1,739 cells; interquartile range: 1,350-2,212 cells). Of the HIV-infected patients, 35 were without HAART. Table 1 shows the demographic and clinical variables of the HIV-infected patients and controls. Compared to controls, HIV-infected patients were six years older (43.2 ± 10.5 vs. 37.9 ± 11.5 years; p < 0.001), had lower BMI (25.5 ± 4.5 vs. 27.4 ± 5.4 kg/m²; p < 0.001), lower frequency of overweight/obesity (51.1 vs. 63.1%; p = 0.005), and higher active smoking incidence (43.6 vs. 16.1%; p < 0.001). Table 2 shows clinical and laboratory variables of the patients, separated in treatment with protease inhibitors (PI). Those exposed to PI showed longer time since diagnosis [140 (74-175) vs. 72.5 (20-120) months; p<0.001] and disease treatment duration [124 (56-155) vs. 44 (4‑101) months; p < 0.001] and elevated TGL levels [190 (119-280) vs. 140 (100-188.5) mg/dL; p < 0.001]; but PI exposure had no effect on LDL-c, HDL-c, fasting glucose, creatinine, or hs-CRP levels. Atherosclerotic plaques in carotid arteries and carotid intima-media thickness Plaques were detected in 37% of the HIV group and 4% of the control group (p < 0.001), as shown in Figure 1. Multivariate logistic regression analysis indicated that the presence of HIV, adjusted for age, overweight/obesity, and smoking, had an almost five-fold increase in the risk of carotid PL (OR 4.9, 95% CI 2.5 to 9.8; p < 0.001). Patients with PL were 11 years older than those without PL (51.4 ± 9.21 vs. 40.2 ± 9.40 years, p < 0.001), and had higher levels of fasting blood glucose [90 (78-100) vs. 83 (76.5‑90) mg/dL; p = 0.012], TC [200 (178-244) vs. 181 (156-208.5) mg/dL; p < 0.001], LDL-c [120.1 (96.2-148.4) vs. 96.8 (80-125) mg/dL; p < 0.001], TGL [188.5 (125.5‑288.5) vs. 150.5 (108-226) mg/dL; p = 0.010] and creatinine [0.80 (0.70-1.10) vs. 0.80 (0.70-0.90) mg/dL; p = 0.027]. Patients with PL also had higher systolic (SBP: 132 ± 21 vs. 121 ± 16 mm Hg; p < 0.001) and diastolic blood pressure (DBP: 83 ± 12 vs. 77 ± 11 mm Hg, p < 0.001). In addition, PL was detected in approximately 34% of men versus 17.4% of women. Regarding treatment with PI, exposure to this drug class was not associated with higher PL frequency. Nevertheless, results show significant interaction between PI and elevated TGL, even though no association with the presence of PL. Figure 2 illustrates the significant association between age and CIMT in both groups, indicating that the oldest individuals have a higher CIMT, regardless of the presence of HIV infection. However, there was a significant interaction between age and the presence of HIV towards increasing CIMT (p < 0.001). Arterial stiffness Comparing patients exposed and not exposed to PI, this drug class had no effect on arterial mechanical characteristics, expressed by PWV [7.10 (6.20-8.20) vs. 7.20 (6.30-8.40) m/s; p = 0.727] and AIx [28 (17-37) vs. 26 (13-38)%; p=0.315]. In addition, no effect was observed on CIMT [0.645 (0.570- 0.765) vs. 0.625 (0.565-0.740) mm; p=0.331]. Pulse wave velocity was associated with age (R = 0.573, p < 0.001), CIMT (R = 0.449, p < 0.001) and SBP (R = 0.557, p < 0.001), as shown in Figure 3. The association between PWV and age persisted in the model corrected for smoking. However, smoking interacted with age to increase PWV (p = 0.05). The AIx was also associated with age (R = 0.411, p < 0.001), CIMT (R = 0.274, p < 0.001), and SBP (R = 0.348, p < 0.001). Arterial stiffness index was elevated in patients with PL compared to no-plaque individuals with PWV [7.90 (7.0-9.5) vs. 6.80 (6.10-8.0) m/s; p < 0.001] and AIx [37 (25-42) vs. 24 (12-35) %; p < 0.001]. Furthermore, PL patients showed median CIMT about 0.170 mm greater than patients without injury [0.770 (0.680-0.910) vs. 0.597 (0.550-0.690) mm; p = 0.003]. Framingham coronary heart disease risk score The FCRS was estimated in 252 HIV-infected patients. Of those, 207 were classified as low-risk (82.1%), 31 as intermediate-risk (12.3%), and 14 (5.56%) as high-risk for developing CVD within 10 years. 404