Cannabis Use in Cocaine Dependence Treatment: A Randomized Trial of Buprenorphine and Naltrexone
Clinical Takeaway
In this study, THC-positive urine drug screens were associated with higher odds of submitting cocaine-positive urine drug screens during treatment for cocaine use disorder. This highlights the importance of addressing polysubstance use in patients undergoing addiction treatment.
#7 Relationships between cannabis and cocaine use in a randomized trial of combined buprenorphine and naltrexone for DSM-IV cocaine dependence.
Citation: Smith Crystal L et al. Relationships between cannabis and cocaine use in a randomized trial of combined buprenorphine and naltrexone for DSM-IV cocaine dependence. Addictive behaviors. 2026. PMID: 42561565.
Design: 5 Journal: 0 N: 2 Recency: 3 Pop: 2 Human: 1 Risk: 0
Abstract: BACKGROUND: Cannabis is the most commonly used drug in the United States, and among people who use cannabis, polysubstance use is common and understudied. We aimed to examine the association of tetrahydrocannabinol (THC) positive urine drug screen (+UDS) with the odds of submitting a cocaine + UDS during cocaine use disorder treatment. METHODS: We conducted a secondary data analysis of a previously reported double-blind, placebo-controlled clinical trial, CTN0048. Participants meeting criteria for opioid abuse/dependence were assigned to receive extended-release naltrexone and one of three conditions of buprenorphine (placebo, 4 mg/day, 16 mg/day) for 8 weeks. Generalized estimating equations (GEE) were used to analyze urine samples (Liu et al., 2018) collected over time, examining the association between THC + UDS and cocaine + UDS during treatment. RESULTS: Participants (n = 301) averaged 46 (SD = 8.64) years of age, were majority male (78.41%), non-Hispanic (89.70%), and African American (66.45%). GEE results indicated that patients who submitted THC + UDS had significantly higher odds of submitting cocaine + UDS compared to participants who submitted THC-negative UDS across the 25 time points examined (OR = 1.47, 95% CI = 1.21-1.79, p = 0.00). Time (OR = 0.9998, 95% CI: 0.9997, 0.9999, p = 0.018) and the covariate of sex assigned at birth (OR = 1.77, 95% CI = 1.13-2.77, p = 0.013) were also significant in the model, indicating very small decreases in the odds of submitting a cocaine + UDS over time for all patients and 77% higher odds of submitting cocaine + UDS for females. CONCLUSION: THC + UDS was associated with increased odds of submitting a cocaine + UDS during treatment. Further investigation is needed to discern whether decreasing THC use will result in reduced cocaine use; however, these results suggest that it may be beneficial to counsel patients on cannabis use cessation both before and during treatment for cocaine use, as it is related to coc
What This Study Teaches Us
Among people in treatment for cocaine use disorder who also use opioids, those who test positive for THC are significantly more likely to also test positive for cocaine during the same treatment period. The association held across 25 urine drug screens over 8 weeks, suggesting this isn’t a one-time coincidence.
Why This Matters Clinically
If you’re treating someone for cocaine dependence and they’re also using cannabis, this signals increased risk of continued cocaine use despite treatment. It raises a practical question: should cannabis use be part of your treatment plan conversation, or is it a marker of something else driving polysubstance use?
Study Snapshot
| Study Design | Secondary analysis of a randomized, double-blind, placebo-controlled trial (CTN0048) |
| Population | N = 301, mean age 46 years, 78% male, 66% African American, all meeting criteria for opioid abuse/dependence and enrolled in cocaine use disorder treatment |
| Intervention | Extended-release naltrexone combined with one of three buprenorphine conditions (placebo, 4 mg/day, or 16 mg/day) for 8 weeks |
| Primary Outcome | Association between THC-positive urine drug screens and cocaine-positive urine drug screens during treatment, analyzed across 25 time points |
| Key Result | THC-positive UDS was associated with 47% higher odds of cocaine-positive UDS (OR = 1.47, 95% CI = 1.21-1.79, p = 0.00) |
Where This Paper Deserves Skepticism
This is a secondary analysis of a trial designed to study opioid and cocaine treatment, not cannabis specifically, which means the cannabis data wasn’t the primary focus and the study wasn’t powered to answer causation questions. The abstract doesn’t tell us the prevalence of cannabis use in this cohort, how it was assessed beyond urine screens, whether use was self-initiated or environmental, or what drove the 77% higher odds in females. Most importantly, this is association, not causation: we can’t tell from the data whether cannabis use is causing increased cocaine use, whether both reflect a common liability toward polysubstance use, or whether the two are independent markers of lower treatment engagement.
Dr. Caplan’s Take
I read this as a real clinical signal worth taking seriously in the assessment phase, not proof of a causal mechanism. When someone in cocaine treatment also tests positive for cannabis, I’m flagging that as a marker of risk. But I’m not concluding that cannabis caused the cocaine use or that stopping cannabis will stop the cocaine. There’s likely something else here: maybe it’s poor impulse control, maybe it’s using cannabis to manage withdrawal, maybe it’s access and social factors, maybe it’s that cannabis and cocaine are both available in the same circles. I’d use this finding to deepen the conversation about what’s driving the polysubstance use in that person, not to assume cannabis is the lever.
Clinical Bottom Line
THC-positive urine screens predict higher odds of concurrent cocaine-positive screens in people treated for cocaine use disorder with opioid comorbidity. This association warrants clinical attention during assessment and treatment planning, but doesn’t establish causation or guide specific intervention yet.
Clinical Angles to Consider:
- How does polypharmacy in older adults amplify the DDI risk here?
- Does fall risk, cognitive burden, or driving safety need special discussion?
- What starting dose and titration pace does this evidence suggest for geriatric patients?
|
Have thoughts on this? Share it:
