Objectives: The explosive and ongoing injecting drug use-related HIV-1 epidemic in Vancouver continues to receive international attention. This study was conducted to determine how patterns of cocaine use influence the risk of HIV infection.
Methods: The Vancouver Injection Drug Users Study is an open prospective cohort of injecting drug users that began in May 1996. At enrollment and at semi-annual follow-up visits an interviewer administers a detailed semi-structured questionnaire. Cox proportional hazards models were used to determine behavioral and drug use patterns reported in the 6 months prior to HIV seroconversion.
Results: One-hundred and nine incident HIV infections have been observed during a mean follow-up of 31 months, from 940 HIV-seronegative participants. During the 6 months prior to seroconversion, predictors of HIV infection were injecting cocaine use [adjusted hazards ratio (AHR), 3.72], incarceration (AHR, 2.74), unstable housing (AHR, 2.36), methadone maintenance treatment (AHR, 1.98), and Aboriginal ethnicity (AHR, 1.78). Injecting cocaine use was predictive of HIV infection in a dose-dependent fashion. Compared with infrequent cocaine users, participants who averaged more than three injections per day were seven times more likely to contract HIV. In addition, the time to HIV infection was accelerated among regular cocaine injectors independent of concurrent heroin use.
Conclusions: Injecting cocaine use was a strong, dose-dependent predictor of HIV seroconversion in this poly-drug using population. Injection cocaine users remain particularly vulnerable to HIV infection and treatment options for cocaine dependency remain woefully inadequate.