The researchers used a nationally representative, web-based panel of 2,400 U.S. adults ages 18 to 64 drawn from the 2023 National Wellbeing Survey, then applied post-stratification weights so the sample’s age, sex, race and ethnicity, and regional makeup matched U.S. Census benchmarks.

They tested associations between depression, anxiety, and self-rated mental health on one side, and cannabis use disorder diagnosis, past-12-month cannabis use frequency, and six specific reasons for use on the other, using weighted logistic regression for binary outcomes and weighted linear regression for use frequency. Every model included interaction terms to check whether ADHD status changed these associations.

As self-rated mental health declined, respondents had greater odds of reporting cannabis use to escape problems, to manage daily stressors, and to help with falling asleep. The authors describe these three motives together, coping with distress and managing symptoms rather than using cannabis recreationally, as the self-medication pattern the survey was designed to characterize.

This distinction matters because it separates why someone uses cannabis from simply whether they use it. The survey’s design let the authors show that symptom-management reasons specifically, not recreational use, were the motives that tracked with worse self-rated mental health.

Reporting a diagnosis of depression was associated with substantially greater odds of also carrying a cannabis use disorder diagnosis, an adjusted odds ratio of 1.84 with a 95% confidence interval of 1.25 to 2.69. That is a statistically significant association, but it is a cross-sectional one, so the survey cannot say whether depression drives heavier or more problematic cannabis use, whether cannabis use disorder contributes to depressive symptoms, or whether some third factor drives both.

This finding is one of the clearest reasons this survey belongs in a clinical conversation rather than only a population-health one. A depression diagnosis is not a minor variable here, it is tied to nearly double the odds of a formal cannabis use disorder diagnosis in this sample.

The survey specifically tested whether the relationships between mental health and cannabis use differed for respondents who also reported an ADHD diagnosis, and found that many of these associations did in fact differ by ADHD status.

The published abstract does not give the individual stratified estimates for each motive by ADHD status, which is a real limitation for translating this finding into specific guidance. What it does establish is that ADHD cannot be treated as a side note when interpreting mental health and cannabis use associations, it appears to meaningfully change the relationship.

Therapeutic motives and sleep-related motives for cannabis use were more commonly endorsed by women and by racial and ethnic minority adults in this sample, a demographic pattern the authors highlight alongside the mental health findings.

This detail is useful for clinicians because it suggests self-medication motives are not evenly distributed across the population, and that conversations about why a patient uses cannabis may need to account for demographic context alongside mental health status.

The authors state directly that these are cross-sectional results identifying co-occurring patterns of mental health and cannabis use, and that the findings support trauma-informed, integrated behavioral health approaches rather than cannabis use alone to manage psychiatric symptoms.

They do not claim to have shown that cannabis use worsens or improves mental health outcomes, only that the two are statistically entangled in ways that differ by diagnosis and by ADHD status, and that this entanglement deserves an integrated clinical response rather than being addressed as two separate issues.

This survey sits inside a well-established public health finding that people with depression, anxiety, and ADHD use cannabis at higher rates than the general population, and that many describe using it to cope with symptoms rather than for recreation. What has been harder to pin down at a national level is how these self-medication motives specifically relate to formal markers of problematic use, like a cannabis use disorder diagnosis, and whether ADHD changes that picture.

This survey adds a nationally representative, quantified look at exactly that question, showing a self-medication pattern tied to worse self-rated mental health, a clear depression to cannabis use disorder association, and a demonstrated but not fully detailed ADHD interaction. It reinforces a message that keeps recurring across this literature: cannabis use in people with psychiatric conditions is frequently a coping strategy, and clinicians who treat it purely as a substance use problem, without addressing the underlying mental health need it may be serving, are likely missing part of the clinical picture.

I see this pattern in clinic regularly. Patients with depression, anxiety, or ADHD often describe cannabis as the thing that gets them through a hard day or lets them sleep, not as a recreational choice, and this survey puts real national numbers behind that clinical impression. The finding that a depression diagnosis nearly doubles the odds of a cannabis use disorder diagnosis is not something I take lightly, it tells me that self-medication, even when it feels like it is working in the moment, carries a measurable risk that deserves an honest conversation rather than either blanket approval or blanket dismissal.

What this survey cannot tell me, and what I try to be precise with patients about, is which direction that relationship runs, or whether a different approach to their depression, anxiety, or ADHD might reduce their need to self-medicate at all. The authors’ call for trauma-informed, integrated behavioral health care rather than cannabis use alone is the right clinical instinct. My job is to ask why a patient is using cannabis the way they are, treat the underlying condition alongside it, and not mistake symptom relief for something we have proven is safe long-term.

A Four-Step Reading Frame