Cannabis use doubled the odds of opioid use on the same day, contradicting the substitution hypothesis | Research

The takeaway

Among 211 adults with problem substance use tracked daily for 90 days, the odds of using non-medical opioids were approximately doubled on days when cannabis was also used, regardless of pain levels.

Addiction medicine researchers, cannabis-opioid policy analysts, harm reduction advocates

Opioid use odds doubled on cannabis-use days (aOR=1.86)

What the researchers found

On days when cannabis was used, the adjusted odds of non-medical opioid use were 1.86 times higher (95% CI: 1.44-2.41). This association did not differ between people with moderate-to-severe pain versus less pain, contradicting the hypothesis that people substitute cannabis for opioids to manage pain.

Why it matters

Ecological studies suggesting cannabis legalization reduces opioid deaths have led to policy interest in cannabis as an opioid substitute. This individual-level data tells a different story: among people who actually use both, cannabis and opioid use cluster together rather than substituting.

The numbers in context

211 participants; 64% male; mean age 43; 13,271 observation days; 70% IVR completion; same-day cannabis-opioid use aOR=1.86 (95% CI: 1.44-2.41); no significant interaction with pain levels

How the study worked

Prospective daily diary study of 211 adults with problem substance use who use non-medical opioids, recruited May 2016-June 2019 in New York. Participants responded to daily interactive voice response for 90 days (13,271 observation days total, 70% completion rate).

What this study cannot tell us

Sample limited to adults with established problem substance use and non-medical opioid use, not representative of all cannabis or opioid users. New York/suburban setting. Same-day association does not prove cannabis causes opioid use. Observational within-person design.

How to read the evidence

Strong within-person daily diary design with high completion rate, providing more rigorous individual-level evidence than cross-sectional studies.

When this study was published

Published in 2021 using 2016-2019 data.

The bigger picture

The disconnect between population-level ecological findings (cannabis laws associated with fewer opioid deaths) and individual-level data (cannabis and opioids used together) suggests that any protective effect of cannabis legalization may operate through mechanisms other than direct substitution.

Questions still open

Would the pattern differ in people who use cannabis medically rather than recreationally? Does the ecological-level protective effect of cannabis laws operate through reduced prescribing, cultural shifts, or other mechanisms rather than individual substitution?

Common questions

Do people substitute cannabis for opioids?

Not in this study. Among adults with problem substance use, opioid use was nearly twice as likely on days they used cannabis. The pattern was the same regardless of pain levels, suggesting cannabis was not being used as a pain substitute for opioids.

Why do ecological studies show the opposite?

Population-level studies suggest cannabis laws reduce opioid deaths, but this may not be driven by individual substitution. The protective effect could operate through reduced opioid prescribing, cultural shifts, or other population-level mechanisms rather than people replacing opioids with cannabis.

Read the original research

Is Cannabis being used as a substitute for non-medical opioids by adults with problem substance use in the United States? A within-person analysis.

Addiction (Abingdon, England), 116(5), 1113-1121

View on PubMed ↗Publisher’s paper ↗

Citation

Gorfinkel, Lauren R; Stohl, Malki; Greenstein, Eliana; Aharonovich, Efrat; Olfson, Mark; Hasin, Deborah. (2021). Is Cannabis being used as a substitute for non-medical opioids by adults with problem substance use in the United States? A within-person analysis.. Addiction (Abingdon, England), 116(5), 1113-1121. https://doi.org/10.1111/add.15228