The Long-Term Relationship Between Cannabis and Heroin Use: An 18- to 20-year Follow-Up of the Australian Treatment Outcome Study (ATOS)

Cannabis use is common among individuals with opioid use disorder, but it remains unclear whether cannabis use is associated with an increase or a reduction in illicit opioid use. To overcome limitations identified in previous longitudinal studies with limited follow-ups, the authors examined a within-person reciprocal relationship between cannabis and heroin use at several follow-ups over 18 to 20 years.

The Australian Treatment Outcome Study (ATOS) recruited 615 people with heroin dependence in 2001 and 2002 and reinterviewed them at 3, 12, 24, and 36 months as well as 11 and 18–20 years after baseline. Heroin and cannabis use were assessed at each time point using the Opiate Treatment Index. A random-intercept cross-lagged panel model analysis was conducted to identify within-person relationships between cannabis use and heroin use at subsequent follow-ups.

After accounting for a range of demographic variables, other substance use, and mental and physical health measures, an increase in cannabis use 24 months after baseline was significantly associated with an increase in heroin use at 36 months (estimate=0.21, SE=0.10). Additionally, an increase in heroin use at 3 months and 24 months was significantly associated with a decrease in cannabis use at 12 months (estimate=−0.27, SE=0.09) and 36 months (estimate=−0.22, SE=0.08). All other cross-lagged associations were not significant.

Although there was some evidence of a significant relationship between cannabis and heroin use at earlier follow-ups, this was sparse and inconsistent across time points. Overall, there was insufficient evidence to suggest a unidirectional or bidirectional relationship between the use of these substances.