A clinical newsletter digest highlights three 2025 studies on vaporized DMT for depression, cannabis-related hospitalization and later dementia risk, and valbenazine for tardive dyskinesia in older adults. Each finding carries intriguing clinical signals, but significant design limitations in every study prevent firm conclusions or immediate practice changes.

Clinicians working with aging populations face converging questions about the safety of rising cannabis use, the promise of psychedelic-assisted therapies, and the tolerability of tardive dyskinesia medications in frail patients. This newsletter digest touches all three issues with 2025 data, making it a timely reference point. But the clinical urgency of these topics also raises the stakes for overinterpretation. When patients and colleagues ask about these findings, practitioners need to understand not only what was reported but precisely how much weight each finding can bear.

This August 2025 issue of the Brown University Psychopharmacology Update summarizes three independent primary studies spanning distinct therapeutic domains. The first reports on an open-label trial of vaporized DMT in 14 adults with depression, noting rapid MADRS score improvements by day one that persisted through three months, along with reductions in suicidality measures. The second draws on a large Ontario retrospective cohort to examine whether cannabis-related emergency department or hospital visits predict subsequent dementia diagnosis. The third presents a post hoc subgroup analysis of adults aged 65 and older from the KINECT 3 and KINECT 4 valbenazine trials for tardive dyskinesia.

The cannabis-dementia study reported that at 10 years, 18.6% of individuals with cannabis-related acute care had been diagnosed with dementia, with a 1.7-fold greater 5-year risk compared to the general population and a 1.2-fold greater risk compared to those with acute care unrelated to cannabis. The valbenazine subgroup showed response rates reaching 89.3% at 48 weeks in elderly participants, with higher responses at 80 mg than 40 mg. Critically, the DMT study had no control arm and enrolled only 14 participants, while the valbenazine analysis pooled just 55 elderly patients from two trials with different designs. The cannabis cohort study, though the largest of the three, cannot establish causation due to residual confounding by psychiatric comorbidity and unmeasured cannabis use intensity. All three research groups note that confirmatory studies with stronger designs are needed.

DMT, Cannabis, and Tardive Dyskinesia: Parsing the Evidence Behind Three Headline-Grabbing Studies

Every month, clinical newsletters like this one compress dozens of research papers into digestible summaries, and in that compression, something important is almost always lost: the methodological texture that separates a compelling signal from an established fact. This particular issue of the Brown University Psychopharmacology Update covers three topics that will predictably generate patient questions and media attention. One claims rapid antidepressant effects from vaporized DMT. Another links cannabis-related hospitalization to dementia. A third suggests high efficacy of valbenazine in elderly patients with tardive dyskinesia. Each finding sounds important. But when you look beneath the surface, the evidential weight varies enormously. The DMT study enrolled 14 participants in an open-label design without a control group. Judging whether a new treatment works by asking 14 people who knew exactly what they were taking, with no comparison group, is like running a product taste test where every taster knows which brand is being promoted. Under those conditions, expectancy effects, placebo response, regression to the mean, and the natural fluctuation of depressive symptoms are all plausible explanations for whatever improvement was observed. The newsletter presents these findings in the same register as the much larger cannabis-dementia cohort, and that is where the format itself becomes misleading.

The Myran cannabis-dementia study is the genuinely important item here. A large Canadian cohort with multiple comparator groups and long follow-up is a serious piece of epidemiology. The finding that people with cannabis-related acute care visits have a 1.7-fold higher 5-year dementia risk than the general population deserves clinical attention, especially as cannabis use among older adults accelerates in jurisdictions where it is legal. But even this study carries a fundamental interpretive problem: the people who show up in emergency rooms for cannabis-related events are not typical cannabis users. They represent a clinically severe, psychiatrically complex subpopulation. Noticing that these individuals have higher dementia rates and concluding that cannabis causes dementia is like noticing that people who call 911 are more likely to be seriously ill and concluding that calling 911 makes you sick. The exposure marker captures severity and vulnerability, not just cannabis exposure. What I find most useful about this study is not the causal inference, which remains uncertain, but the identification of a recognizable, healthcare-engaged group of older adults who could benefit from structured dementia surveillance. That is actionable regardless of whether the association turns out to be causal.

The valbenazine elderly subgroup data are encouraging for an already-approved medication, but 55 patients drawn post hoc from two differently designed trials do not constitute robust evidence of age-specific efficacy. The 89% response rate at 48 weeks is notable, yet selected trial populations rarely predict real-world outcomes with accuracy. If I were speaking to a colleague, I would say: read the Myran primary paper, keep the DMT data filed under “watch and wait,” and treat the valbenazine numbers as supportive rather than definitive. If I were speaking to a patient who asked about these headlines, I would explain that we are seeing early signals, not finished answers. The broader lesson here is one worth internalizing: the format of a clinical newsletter digest produces a systematic optimism bias. Studies are selected for coverage in part because they are interesting or positive, and brevity prevents the inclusion of the methodological caveats that would calibrate reader confidence. Clinicians using these digests as a primary evidence interface should treat each item as a reading list rather than a conclusions list.

For clinicians navigating questions about cannabis safety in older adults, the Myran et al. study provides a useful framework: cannabis-related acute care events can be treated as a flag for heightened dementia surveillance, much as alcohol-related hospitalizations already trigger structured follow-up for cognitive and hepatic complications. This does not require assuming causation. It requires recognizing that patients who present for cannabis-related emergencies carry a burden of risk factors that independently elevate dementia probability, making them a population that benefits from proactive screening.

Regarding the pharmacological findings, vaporized DMT remains an investigational compound without regulatory approval for clinical use, and the safety profile cannot be meaningfully assessed from 14 participants. Valbenazine is already approved for tardive dyskinesia, and this subgroup analysis modestly reinforces existing prescribing confidence in older adults, though tolerability monitoring remains essential given polypharmacy concerns in this age group. One concrete recommendation: when older patients or their families ask about cannabis and dementia risk based on news coverage of these findings, clinicians should distinguish between heavy, problematic use requiring acute care and moderate, controlled use, as the study population represents the former and results should not be generalized to the latter.

This is a clinical newsletter digest, not a primary study, systematic review, or meta-analysis. It sits near the bottom of the evidence hierarchy, functioning as an editorially curated awareness tool that points readers toward primary publications. The single most important inference constraint is that the newsletter performs no independent critical appraisal of the studies it summarizes, meaning readers must consult the original publications to assess risk of bias, effect size precision, and generalizability.

The cannabis-dementia association aligns with a growing body of observational research linking heavy substance use to accelerated cognitive decline, though prior studies have focused predominantly on alcohol. Myran et al. extends this literature by specifically examining cannabis-related acute care in a population-based cohort, providing a more granular exposure marker than self-reported cannabis use in survey-based studies. The DMT antidepressant findings are consistent with the broader psychedelic therapy research arc, including psilocybin and ketamine trials, but represent a much earlier phase of investigation with far less methodological rigor than the landmark Goodwin et al. (2022) psilocybin trial in the New England Journal of Medicine. The valbenazine elderly subgroup analysis builds on prior KINECT trial publications confirming efficacy in younger adults, modestly extending confidence to a demographic that was underrepresented in the original pivotal trials.

For the cannabis-dementia study, the most consequential analytic choice was the use of cannabis-related acute care as the exposure variable rather than self-reported cannabis use or cannabis purchase records. If the investigators had used a lower-severity exposure marker that captured moderate users alongside heavy users, the effect size would almost certainly have been attenuated, because the acute-care population carries a disproportionate burden of psychiatric comorbidity and health risk factors. Conversely, if the analysis had additionally stratified by frequency of cannabis-related visits or co-occurring psychiatric diagnoses, it might have revealed a dose-response gradient or identified subgroups where the association was stronger or weaker. For the DMT study, the inclusion of any control arm, even a wait-list control, would have fundamentally changed the interpretability of the results, potentially revealing that much of the observed improvement was attributable to expectancy and natural recovery.

The most likely overinterpretation is concluding that cannabis use causes dementia in older adults. The Myran et al. study identifies an association between cannabis-related acute care and subsequent dementia diagnosis, but the people captured by this exposure definition are not representative of all cannabis users. They represent a high-severity subpopulation with elevated psychiatric and medical comorbidity. Generalizing these findings to moderate or occasional cannabis users exceeds what the data can support. A second common misreading involves treating the DMT open-label results as evidence of antidepressant efficacy when the design cannot distinguish drug effects from placebo response, regression to the mean, or natural symptom fluctuation. The absence of a control arm is not a minor limitation; it renders the observed effect size uninterpretable as a measure of treatment efficacy.

This newsletter digest identifies three clinically interesting signals across psychedelic therapy, cannabis safety, and tardive dyskinesia treatment. The cannabis-dementia association from Myran et al. is the most methodologically credible finding and points toward a recognizable high-risk group that could benefit from dementia surveillance. The DMT and valbenazine findings are too preliminary or too narrowly powered to change practice. Clinicians should treat these summaries as a reading list, not a conclusions list, and consult the primary publications before drawing clinical inferences.

Does cannabis cause dementia?

This study does not establish that cannabis causes dementia. It found that people who went to the emergency room or hospital for cannabis-related problems had a higher rate of dementia diagnosis in the following years. However, these individuals represent a high-risk group with many other factors that could independently increase dementia risk. Moderate, controlled cannabis use was not specifically studied, and these findings should not be generalized to all cannabis users.

Is DMT an approved treatment for depression?

No. DMT is not approved for clinical use as an antidepressant by any regulatory agency. The study described here tested vaporized DMT in just 14 people without a comparison group, which means we cannot determine whether the improvements observed were caused by the drug or by other factors such as the expectation of benefit or the natural course of the illness. Controlled trials with larger groups of participants are needed before any clinical conclusions can be drawn.

Is valbenazine safe for older adults with tardive dyskinesia?

The data described here are encouraging but limited. Only 55 adults aged 65 and older were included in this analysis, drawn from two different clinical trials. The response rates were high, and the side effect profile appeared similar to that seen in younger adults. However, the small sample size and the post hoc nature of the analysis mean that more research specifically designed for elderly patients is needed before definitive safety and efficacy conclusions can be made for this age group.

Should I be worried about my cannabis use if I am over 45?

If your cannabis use has never resulted in an emergency room visit or hospitalization, this particular study does not directly apply to your situation. The elevated dementia risk was found specifically among individuals whose cannabis use was severe enough to require acute medical care. That said, any substance use in midlife and beyond is worth discussing with your physician, who can help you weigh risks based on your individual health history, patterns of use, and other risk factors.

Have thoughts on this? Share it: