Abstract
IMPORTANCE: The vascular risk attributable to HIV infection is rising. The heterogeneity of the samples studied is an obstacle to understanding whether HIV is a vascular risk across geographic regions. OBJECTIVE: To test the hypothesis that HIV infection is a vascular risk factor, and that the risk conferred by HIV varies by geographical region. DATA SOURCES: A systematic search of publications was carried out in seven electronic databases: PubMed, The Cochrane Library, EMBASE, Web of Science, LILACS, ClinicalTrials.gov, and WHO International Clinical Trials Registry Platform from inception to July 2015. STUDY SELECTION: We included longitudinal studies of HIV+ individuals and their risk of vascular outcomes of ≥ 50 HIV+ cases and excluded studies on biomarkers of vascular disease as well as clinical trials. DATA EXTRACTION AND SYNTHESIS: Data was extracted by one of the authors and independently confirmed by the other two authors. We used incidence rate (IR), incidence risk ratio (IRR) and hazard ratio (HR) with their 95% confidence intervals as measures of risk. MAIN OUTCOME: All-death, myocardial infarction (MI), coronary heart disease (CHD), any stroke, ischemic stroke (IS) or intracranial hemorrhage (ICH). RESULTS: We screened 11,482 references for eligibility, and selected 117 for analysis. Forty-four cohorts represented 334,417 HIV+ individuals, 49% from the United States. Compared with their European counterparts, HIV+ individuals in the United States had higher IR of death (IRR 1.78, 1.69-1.88), MI (IRR 1.61, 1.29-2.01), CHD (IRR 2.27, 1.92-2.68), any stroke (IRR 1.94, 1.59-2.38), IS (IRR 1.56, 1.23-1.98), and ICH (IRR 4.03, 2.72-6.14). Compared with HIV- controls and independent of geographical region, HIV was a risk for death (HR 4.77, 4.55-5.00), MI (HR 1.60, 1.49-1.72), any CHD (HR 1.20, 1.15-1.25), any stroke (HR 1.82, 1.53-2.16), IS (HR 1.27, 1.15-1.39) and ICH (HR 2.20, 1.61-3.02). Use of antiretroviral therapy was a consistent risk for cardiac outcomes, while immunosuppression and unsuppressed viral load were consistent risks for cerebral outcomes. CONCLUSIONS: HIV should be considered a vascular risk, with varying magnitudes across geographical and anatomical regions. We think that strategies to reduce the HIV-related vascular burden are urgent, and should incorporate the disparities noted here.