Background.  Diabetes and insulin resistance (IR) is common in human immunodeficiency virus–hepatitis C virus (HIV–HCV)-coinfected patients, a population also concerned with elevated cannabis use. Cannabis has been associated with reduced IR risk in some population-based surveys. We determined whether cannabis use was consistently associated with reduced IR risk in HEPAVIH, a French nationwide cohort of HIV–HCV-coinfected patients.

Methods.  HEPAVIH medical and sociobehavioral data were collected (using annual self-administered questionnaires). We used 60 months of follow-up data for patients with at least 1 medical visit where IR (using homeostatic model assessment of insulin resistance [HOMA-IR]) and cannabis use were assessed. A mixed logistic regression model was used to evaluate the association between IR risk (HOMA-IR > 2.77) and cannabis use (occasional, regular, daily).

Results.  Among the 703 patients included in the study (1287 visits), 323 (46%) had HOMA-IR > 2.77 for at least 1 follow-up visit and 319 (45%) reported cannabis use in the 6 months before the first available visit. Cannabis users (irrespective of frequency) were less likely to have HOMA-IR > 2.77 (odds ratio [95% confidence interval], 0.4 [.2–.5]) after adjustment for known correlates/confounders. Two sensitivity analyses with HOMA-IR values as a continuous variable and a cutoff value of 3.8 confirmed the association between reduced IR risk and cannabis use.

Conclusions.  Cannabis use is associated with a lower IR risk in HIV–HCV-coinfected patients. The benefits of cannabis-based pharmacotherapies for patients concerned with increased risk of IR and diabetes need to be evaluated in clinical research and practice.