An observational study of older adults in Oregon found that cannabis use was associated with more pain interference and depressive symptoms over time, not less, challenging assumptions that self-treatment with cannabis improves chronic pain outcomes. However, the cannabis-using group was extremely small, and the study design cannot prove that cannabis caused these worse outcomes.

Cannabis use among adults over 65 has risen sharply over the past decade, driven in large part by state-level legalization and growing public perception that cannabis is a safer alternative to opioids for managing chronic pain. Yet older adults remain profoundly underrepresented in cannabis clinical trials, leaving clinicians with almost no rigorous evidence to guide conversations about benefit or risk in this population. Longitudinal real-world data, even with significant limitations, help fill a critical gap in understanding how cannabis use tracks with pain and mood outcomes in aging patients living with persistent pain.

Chronic non-cancer pain is one of the most common conditions driving older adults to explore cannabis as a therapeutic option, often in the hope of reducing reliance on opioids or improving daily functioning. This study is a secondary analysis of the ORCATECH (Oregon Center for Aging and Technology) cohort, a longitudinal research platform originally designed to track aging and technology adoption among community-dwelling older adults. From 541 enrollees, 353 met criteria for chronic non-cancer pain, defined as pain persisting 90 or more consecutive days without a cancer, HIV/AIDS, opioid use disorder, or palliative care diagnosis. Cannabis use was ascertained by self-report, while opioid use was tracked through prescription fills, creating an asymmetry in exposure measurement that complicates direct comparison.

Contrary to the authors’ hypothesis, cannabis use (reported by roughly 5% of the sample, or about 18 individuals) was associated with significantly greater pain interference with daily activities and more depressive symptoms over time on both weekly and annual measures. Prescription opioid use (about 11%, or 39 individuals) was similarly associated with greater pain intensity and interference. Neither exposure was linked to improvements in physical function. The authors acknowledge that the cannabis subgroup is severely underpowered, that confounding by indication likely explains much of the opioid signal, and that no data on cannabis dose, formulation, or frequency were available. They characterize the cannabis findings as hypothesis-generating and call for adequately powered, controlled studies in this population.

I appreciate that this research team looked at cannabis longitudinally in older adults, a population where we badly need data. The direction of the association is striking and worth noting. But I want to be very cautious about what we do with a finding built on approximately 18 cannabis users in a single-site Oregon cohort. The most likely explanation, in my view, is confounding by indication: the older adults turning to cannabis were probably the ones struggling most with pain and mood to begin with, and we have no information about what they were using, how much, or how often. That is not a minor gap. It is the whole question.

In my practice, I see a wide range of outcomes among older patients using cannabis for chronic pain. Some do very well with carefully chosen, low-dose formulations, and some do not benefit at all. What consistently matters is clinical oversight: dose titration, product selection, monitoring for mood effects, and honest conversation about expectations. This study does not change that approach, but it does reinforce why we cannot simply assume cannabis is helping when we are not measuring what patients are actually using.

This study sits early in the research arc for understanding cannabis outcomes in geriatric chronic pain populations. It provides a temporal signal but falls well short of the evidence needed to change clinical behavior. Clinicians should recognize that the association between cannabis use and worse outcomes here almost certainly reflects selection bias: patients with more refractory pain and comorbid depression are the ones most likely to try cannabis. The same confounding by indication phenomenon is well documented for opioids in observational pain studies, and the opioid findings in this cohort follow the same expected pattern. Until randomized or at minimum well-controlled prospective studies with adequate sample sizes and detailed exposure characterization are available, these results should inform research priorities rather than clinical decision-making.

For clinicians managing older adults who are using or considering cannabis for chronic pain, the pharmacological considerations remain substantial. Older adults metabolize cannabinoids differently due to changes in body composition, hepatic function, and polypharmacy-related drug interactions, particularly with anticoagulants, statins, and CNS depressants. The absence of any dose, formulation, or route-of-administration data in this study means it cannot inform product-level guidance. The one actionable takeaway is this: do not assume cannabis is working simply because a patient continues to use it. Monitor pain interference and mood systematically, and revisit the therapeutic plan if outcomes are not improving.