Prescribed stimulants did not raise overdose risk overall during opioid agonist therapy
A large retrospective cohort from British Columbia suggests that prescribed stimulant medications do not increase overdose risk overall in people receiving opioid agonist therapy for opioid use disorder. Across 9,395 individuals followed for 18,273 person-years between January 2015 and February 2020, investigators captured 1,746 overdose events, including 37 fatal events. After adjustment for sociodemographic characteristics and substance use patterns, stimulant dispensation was not associated with a statistically significant increase in overdose risk overall, with an adjusted hazard ratio of 1.13 and a 95% confidence interval crossing unity.
The absence of an overall signal matters because co-prescribing stimulants in this population often raises concern about destabilizing substance use, especially when clinicians are balancing attention-deficit/hyperactivity disorder symptoms, daytime fatigue, and ongoing treatment for opioid use disorder. In a population already at high baseline risk of overdose, even a modest drug-related increase would carry important bedside consequences. Here, the aggregate data did not show that effect.
Divergent overdose patterns emerged by type of opioid agonist therapy
The more clinically striking finding came from the secondary analysis by OAT subtype. Among individuals receiving buprenorphine, stimulant dispensation was associated with a lower overdose risk, with an adjusted hazard ratio of 0.47 and a 95% confidence interval of 0.23 to 0.96. Among individuals receiving full agonist OAT — methadone or slow-release oral morphine — stimulant dispensation was associated with an increased overdose risk, with an adjusted hazard ratio of 1.51 and a 95% confidence interval of 1.09 to 2.07.
One sentence captures the clinical divide: the same stimulant exposure tracked in opposite directions depending on the opioid agonist backbone.
Several mechanisms could plausibly explain the split, even if the dataset cannot prove causality. Buprenorphine’s partial agonist pharmacology produces a different safety profile than full agonist therapy, particularly in terms of respiratory depression and overdose vulnerability. In patients stabilized on buprenorphine, stimulant treatment may reflect better engagement in care, improved functioning, or more structured prescribing, all of which could correlate with lower overdose risk. By contrast, among full agonist recipients, stimulant use may identify a subgroup with greater clinical complexity, heavier polysubstance exposure, or unstable treatment trajectories, factors that can amplify overdose risk even after statistical adjustment.
The data do not justify assuming that the stimulant itself is uniformly harmful or protective. They do, however, suggest that the interaction between stimulant prescribing and the specific opioid agonist regimen deserves closer attention at the point of care.
What the cohort can and cannot tell clinicians
The study used the British Columbia Provincial Overdose Cohort, a linked administrative database, which gives it breadth and real-world relevance. That design also imposes familiar limits. Administrative data capture dispensing and outcomes well enough to define population-level associations, but they cannot fully resolve why a stimulant was prescribed, whether the medication was taken as intended, how severe psychiatric symptoms were at baseline, or how often illicit stimulant and opioid use co-occurred outside the medication record. Those unmeasured factors could influence both treatment selection and overdose risk.
Even so, the study’s core clinical message is direct. Co-prescription of stimulants during opioid agonist therapy did not increase overdose risk in the cohort as a whole. The risk estimate changed substantially when the analysis separated buprenorphine from full agonist therapy, suggesting that the choice of opioid agonist remains central when clinicians think about the safety profile of adjunctive stimulant treatment.
For prescribers, the findings support more individualized decision-making rather than reflexive avoidance of stimulants in every patient receiving OAT. A patient stabilized on buprenorphine is not the same as a patient maintained on methadone or slow-release oral morphine, and the overdose data here reflect that difference. The appropriate next step is not a broad prohibition; it is careful risk assessment anchored to the specific OAT regimen, concurrent substance use, and the reasons a stimulant is being considered in the first place.
The unanswered question is whether the observed subgroup differences reflect pharmacology, treatment selection, or both. Until more granular prospective data arrive, the British Columbia cohort argues for caution, not panic, when stimulants are prescribed alongside opioid agonist therapy.