A 2025 Canadian survey of 503 community-dwelling cannabis users aged 50 and older found that 61% reported at least one adverse effect in the past year, most commonly dry mouth, feeling high, and balance problems. Women and adults in their 50s faced higher odds, but the convenience-sample design and self-reported outcomes mean these figures cannot be generalized as definitive population risk estimates.

Cannabis use among older adults has accelerated since Canadian legalization, yet this growing population of users largely obtains cannabis from dispensaries and informal sources without clinical oversight. Unlike patients in medical cannabis programs, community-dwelling older adults face age-related vulnerabilities including altered drug metabolism, increased fall risk, and cognitive fragility that make adverse effect monitoring especially important. Understanding what these users actually experience in real-world settings fills a gap that clinical trial data and medical program registries cannot address.

Cannabis use among adults over 50 has risen substantially since Canadian legalization in 2018, yet most published safety data come from clinical cannabis programs or tightly controlled trials, leaving a knowledge gap about what community-dwelling older users actually experience. This cross-sectional survey, conducted online between February and September 2022 through aging-focused advocacy organizations and social media, recruited a convenience sample of 1,615 Canadian adults aged 50 and older, then restricted its adverse-effect analysis to 503 participants who reported using cannabis at least once monthly. The study used pre-defined adverse effect categories drawn from existing literature, coded by two independent investigators, and analyzed predictors through multivariate logistic regression adjusting for age, sex, multimorbidity, polypharmacy, reason for use, and frequency.

Among the 503 current users, 61.2% reported at least one adverse effect in the past year. Dry mouth was most common at 36.2%, followed by feeling high (25.9%), balance-related effects including dizziness and lightheadedness (22.1%), and mental alertness impairment such as grogginess and confusion (20.3%). Psychological effects like anxiety, paranoia, and hallucinations were rare at 7.0%. Female sex was independently associated with approximately doubled odds of any adverse effect (OR 1.99, 95% CI 1.33 to 2.97) and balance effects (OR 1.93). Adults aged 70 and older had roughly half the odds of any adverse effect compared to those aged 50 to 60. Multimorbidity, polypharmacy, and reason for use did not reach statistical significance. The authors acknowledge that the convenience sample, self-report design, absence of dose verification, and exclusion of former users who may have discontinued due to harms all substantially limit interpretability, and they call for prospective studies with objective endpoints.

Cannabis and Older Adults: What a Canadian Survey Reveals About Real-World Adverse Effects

More than half of older Canadians in a new survey said cannabis gave them at least one unwanted effect, but before that number reshapes clinical practice, it is worth understanding exactly what the data can and cannot tell us. This study does something genuinely useful: it looks beyond the controlled environment of medical cannabis clinics and asks what community-dwelling older adults are actually experiencing with products they purchase on their own, without a prescriber titrating doses or monitoring for drug interactions. The finding that 22% of participants reported balance disturbances and 20% reported impaired mental alertness is not a number I can dismiss. In my practice, when an older patient reports dizziness or confusion, those symptoms carry real weight because the downstream consequences, particularly falls leading to hip fractures, head injuries, and hospitalization, are among the most devastating events in geriatric medicine. The study’s greatest contribution is simply giving us a number to attach to a concern that many clinicians already feel intuitively but have struggled to quantify in a real-world, community-based context.

The central methodological problem, however, is survivorship bias, and it fundamentally shapes every prevalence number in this paper. Only current cannabis users, those still using at least once a month, were included. Anyone who tried cannabis, had a terrible experience with dizziness or cognitive impairment, and stopped using it altogether would not appear in this dataset. It is like surveying only people who are still skiing about how often they fall on the slopes; those who broke a leg last season and sold their skis are nowhere to be found in the results. This means the 61% adverse effect rate almost certainly underestimates the true burden of serious harms while potentially overestimating the prevalence of mild, tolerable effects that users have already decided to live with. The counterintuitive finding that adults 70 and older report fewer adverse effects than those in their 50s is best understood through this lens. It is not that aging protects against cannabis side effects; it is more likely that the most vulnerable older users already stopped and dropped out of the denominator. Without knowing who left and why, the apparent age gradient could be entirely an artifact of differential attrition.

Additionally, the study collected no information about cannabis dose, potency, or THC-to-CBD ratio at the time adverse effects occurred, and it did not assess baseline symptoms. Many older adults experience dizziness, dry mouth, and cognitive fog for reasons entirely unrelated to cannabis. Without knowing how dizzy someone was before they started using cannabis, we cannot be certain the cannabis caused the dizziness; blaming a new medication for a headache that the patient had every afternoon for years is a diagnostic trap we must avoid. When a patient asks me whether cannabis is safe, I tell them that one in five older adults in this study reported feeling unsteady or foggy, and that those numbers are likely conservative. I would recommend we discuss exactly what they are using, how much, and whether they have noticed any balance or thinking changes, because community cannabis purchases do not come with the safety checks of a clinical prescription. To a colleague, I would say these data justify incorporating cannabis adverse effect screening into every geriatric visit. To a policymaker, I would argue these findings make the case for better point-of-sale guidance and standardized product labeling. Self-reported prevalence surveys of adverse drug effects in self-selected current users will structurally underestimate serious harms due to survivorship and cannot establish causation or dose-response relationships, but they can identify patient subgroups warranting targeted clinical attention and generate testable hypotheses for prospective studies with objective endpoints.

This study sits near the beginning of the research arc needed to inform evidence-based cannabis safety guidance for older adults. It is descriptive rather than explanatory, hypothesis-generating rather than hypothesis-testing. Prior literature has drawn predominantly from medical cannabis registries, geriatric clinics, and dispensary populations where some degree of clinical selection or oversight exists. By recruiting through aging advocacy organizations and social media, this survey captures a segment of the older cannabis-using population that is otherwise invisible to clinical research, though the self-selected nature of that recruitment brings its own distortions.

From a pharmacological perspective, the finding that balance and alertness impairment affect roughly one in five older users is consistent with known THC effects on vestibular function and cognitive processing speed, compounded by age-related reductions in hepatic clearance and increased receptor sensitivity. The absence of dose and product composition data is the study’s most frustrating clinical gap, because without it, clinicians cannot distinguish whether adverse effects cluster with higher-THC products, specific routes of administration, or particular dose thresholds. The non-significant findings for polypharmacy and multimorbidity should not be taken as reassurance, as the study was likely underpowered for these subgroup analyses. One actionable recommendation: clinicians caring for adults over 50 should routinely ask about cannabis use and specifically screen for balance disturbances, dizziness, and cognitive fog, treating these as medication-related adverse effects warranting product and dose review.

This is a cross-sectional observational study based on a convenience-sample online survey, placing it in the lower tiers of the evidence hierarchy. It can describe patterns of association and generate hypotheses but cannot establish causation, quantify attributable risk, or demonstrate dose-response relationships. The single most important inference constraint is that the convenience sample and self-reported outcomes make prevalence estimates unreliable as population-level statistics; they describe what this particular group of respondents experienced, not what all older Canadian cannabis users would report.

This study broadly confirms prior clinical-setting findings that dry mouth, dizziness, and somnolence are among the most common cannabis-related complaints in older adults, while extending those observations into a community-dwelling population with less clinical oversight. Earlier work from medical cannabis clinics and dispensaries has reported adverse effect rates ranging from 30% to over 50%, though direct comparison is complicated by differences in how adverse effects are defined and ascertained. One study of medically complex older adults in clinical settings reported higher rates of psychological adverse effects than the 7% found here, consistent with the hypothesis that this healthier, self-selected convenience sample may underrepresent patients with greater vulnerability. The sex-based findings align with a growing body of pharmacological evidence suggesting women experience different cannabinoid metabolism and receptor sensitivity, though the mechanisms remain incompletely understood.

The most consequential analytic choice was restricting the sample to current users only, which creates the survivorship bias discussed throughout this analysis. Had the authors included former cannabis users who stopped due to adverse effects, the prevalence of serious harms would almost certainly have been higher, and the apparent protective effect of older age may have diminished or reversed. Additionally, the decision to classify “feeling high” as an adverse effect inflates the overall 61% figure, as this is an intended outcome for many recreational users. If feeling high were excluded from the adverse effect count, the headline prevalence would drop meaningfully. Finally, the 81+ age subgroup (n=26) was combined with the 71 to 80 group in the regression analysis; analyzing this oldest group separately, if adequately powered, might have revealed a different risk profile in the population at greatest fall risk.

The most likely overinterpretation is treating the 61% figure as a reliable population prevalence rate and concluding that “cannabis causes adverse effects in most older users.” This number comes from a convenience sample of volunteers recruited through aging advocacy organizations, restricted to current users, and based entirely on self-report with no dose verification or clinical confirmation. Equally problematic would be reading the lower adverse effect odds in adults 70 and older as evidence that cannabis becomes safer with advancing age. This pattern almost certainly reflects survivorship bias, where those most harmed have already stopped using cannabis, rather than any biological protection conferred by aging. Finally, the non-significant findings for multimorbidity and polypharmacy should not be cited as evidence that these factors are irrelevant to cannabis safety; the study was likely underpowered to detect these associations.

This study provides a useful descriptive snapshot showing that self-reported adverse effects, particularly balance disturbances and alertness impairment, are common among community-dwelling older Canadian cannabis users. It does not establish causal relationships, population-level prevalence, or dose-response thresholds. For clinical practice today, its most actionable contribution is reinforcing that older cannabis users should be routinely screened for balance and cognitive symptoms, with particular attention to women and adults in their 50s and 60s.

Does this study prove that cannabis is dangerous for older adults?

No. The study describes what a group of volunteer respondents reported experiencing, but it cannot prove that cannabis caused those effects. Many symptoms reported, such as dry mouth and dizziness, are also common in older adults for reasons unrelated to cannabis. The study had no control group of non-users and no way to verify doses or products, so it generates hypotheses rather than definitive safety conclusions.

Should I stop using cannabis based on these findings?

These findings do not provide a basis for broad recommendations to stop or continue cannabis use. What they do support is having a conversation with your healthcare provider about what you are using, how much, and whether you have noticed any changes in balance, alertness, or cognitive function. If you are experiencing dizziness or unsteadiness, those symptoms deserve clinical evaluation regardless of whether cannabis is the cause.

Why did older participants (70+) report fewer side effects than those in their 50s?

The most likely explanation is survivorship bias. People who experienced serious side effects in their earlier years of cannabis use may have already stopped using it, so they would not appear in a survey of current users. It is also possible that long-term users have developed tolerance or that this age group self-selects for lower-risk products and doses. The finding should not be interpreted as evidence that older age is protective against cannabis adverse effects.

Are women really at higher risk for cannabis side effects?

In this study, women had roughly twice the odds of reporting any adverse effect and balance-related effects compared to men. This aligns with a growing body of pharmacological research suggesting that women metabolize cannabinoids differently and may have greater receptor sensitivity. However, the self-report design means we cannot rule out that women may also be more attentive to or more willing to report side effects. More research with objective measurements is needed to clarify the biological basis of this difference.

Have thoughts on this? Share it: