You are here
Stroke Incidence Highest in Women and Black HIV-Infected Participants in ALLRT Cohort
Felicia Chow1; Michael R. Wilson1; Kunling Wu2; Ron Ellis3; Ronald Bosch2; Benjamin P. Linas4
1Univ of California San Francisco, San Francisco, CA, USA;2Harvard Sch of PH, Boston, MA, USA;3Univ of California San Diego, San Diego, CA, USA;4Boston Univ Sch of Med, Boston, MA, USA
While rates of stroke are higher in HIV-infected compared with age-matched HIV-uninfected individuals, many questions persist regarding the nature of cerebrovascular disease in HIV. We leveraged the large ACTG Longitudinal Linked Randomized Trials (ALLRT) cohort and its parent studies to investigate stroke incidence and associated risk factors in HIV-infected individuals.
We conducted a prospective observational cohort study of ART-naïve participants without a history of stroke who initiated ART from June 1998 to June 2011. The primary outcome was first-ever stroke or transient ischemic attack (TIA) documented at study follow up visits through centralized reporting. Age-adjusted Poisson regression models with time-updated covariates were used to identify traditional and HIV-specific risk factors for incident stroke/TIA.
Of 6,933 participants included in the analysis, 20% were women, 37% were non-Hispanic Blacks and 21% were Hispanic. Median pre-ART age was 37 years, pre-ART CD4 count was 243 cells/uL and pre-ART HIV RNA was 57,624 copies/mL. Fifty-four stroke/TIAs occurred over 32,023 person-years (PY). The incidence rate of stroke/TIA in women was 0.29 per 100 PY versus 0.14 per 100 PY in men [age-adjusted relative risk (RR) for female sex 1.72, 95% CI 0.96-3.08]. Incidence of stroke/TIA in non-Hispanic Blacks was 0.25 per 100 PY compared with 0.08 per 100 PY in Hispanics/Other (age-adjusted RR 2.94, 95% CI 1.22-7.14) and 0.16 per 100 PY in Whites (RR 1.67, 95% CI 0.95-2.94). In a multivariable model (Table), traditional risk factors that conferred greater risk of stroke/TIA were older age, LDL ≥ 160 mg/dL and hypertension. Of HIV-related factors, time-updated CD4 count ≤ 200 cells/uL and HIV RNA > 200 copies/mL were associated with increased risk of stroke/TIA. We found no statistically significant association of injection drug use, hepatitis C infection or recent use of any class of ART with stroke/TIA.
Age-adjusted incidence of stroke/TIA was highest in women and non-Hispanic Blacks in this cohort of HIV-infected participants. Investigation into the association between female sex and non-Hispanic Black race with stroke/TIA in HIV is merited. In addition to an association with several modifiable traditional risk factors, stroke/TIA was also associated with immunodeficiency and poor virologic control. This raises the possibility that immunologic sequelae of uncontrolled viremia may also contribute to stroke risk in HIV infection.