A well-known animal study found that a cannabinoid had an antidepressant-like effect at low doses. At higher doses that effect disappeared, and the firing of serotonin neurons was suppressed. That is not a finding about how much to take. It is a warning that the relationship is a curve, and that being on the wrong side of it can produce the opposite of what you wanted.
Read this before anything else. This article is not a recommendation to use cannabis for depression or bipolar disorder, and nothing in it should be used to change psychiatric medication. Stopping an antidepressant or a mood stabiliser abruptly is dangerous. In Thailand cannabis flower is a controlled herb dispensed only on prescription form PT 33 by an authorised practitioner. A personal or family history of psychosis, bipolar disorder or serious depression is a reason for particular caution (see the forum article on contraindications and cautions). If you are struggling with your mood right now, please speak to a doctor or someone you trust rather than acting on anything you read on a forum. In Thailand you can call the Department of Mental Health hotline 1323 (24 hours). Elsewhere in Southeast Asia, call your local emergency number.
1. The Study That Started This
Bambico, Gobbi and colleagues at McGill University, published in the Journal of Neuroscience in 2007. Rats were given the synthetic cannabinoid WIN55,212-2. The researchers assessed them using the forced swim test, a standard behavioural measure in antidepressant research, and recorded the firing activity of serotonin neurons in the dorsal raphe nucleus.
At lower doses: the rats showed an antidepressant-like effect in the forced swim test, and their serotonin-producing neurons fired more. Both effects depended on CB1 receptors. The serotonin effect relied on the ventromedial prefrontal cortex: cutting the medial prefrontal connections abolished it, and injecting the compound directly into that region reproduced it.
At higher doses: the benefit did not merely fade. The antidepressant-like effect in the swim test was no longer seen, and the firing activity of serotonin neurons fell, through a mechanism that no longer depended on CB1 receptors. The McGill press release (2007) went further. It reported Dr Gobbi’s description that serotonin in the rats’ brains dropped below the level of the control group. Keep in mind that the paper measured the firing of serotonin neurons, not brain serotonin levels.
That is the whole finding, and both halves belong to it.
2. What an Inverted U Actually Means
Most people carry an intuition that medicines work in a straight line: if a little helps, more helps more, and too much is simply wasteful. A large part of pharmacology does not behave that way.
A biphasic or inverted-U dose-response curve rises to a peak and then falls, and past a certain point it can cross below the starting line. Under that shape, three statements are all true at once:
- The compound has a genuine beneficial effect.
- Taking more of it produces less benefit.
- Taking substantially more produces an outcome worse than taking none.
Nothing about the subjective experience tells you where on that curve you are. This is the single most consequential fact in this article. It is why “it isn’t working, so I’ll take more” is a specifically dangerous response in this domain, not merely an ineffective one.
3. Four Reasons Not to Carry That Study Into Human Practice
The McGill result is real and important. It is also a rat study, and the gap matters.
It was not THC. WIN55,212-2 is a synthetic full agonist at the cannabinoid receptors. THC is a partial agonist. The pharmacology is different, so the curve cannot be assumed to be the same.
The forced swim test is not depression. It is a behavioural assay that responds to known antidepressants. It measures something antidepressants act on; it is not a model of human depressive illness.
The dose was controlled. The animals were injected with measured amounts. The McGill press release (2007) named the practical obstacle: the dose of natural cannabis is difficult to control, particularly when it is smoked. On a curve where overshooting cancels the effect, smoking is precisely the wrong delivery method for a compound with this profile.
The researchers did not conclude “use cannabis.” According to the McGill press release (2007), the team’s direction was towards a new class of drugs that enhance the effects of the brain’s own endocannabinoids. In the same release, Dr Gobbi warned that excessive cannabis use in people with depression poses a high risk of psychosis.
That warning comes from the senior author of the study most often cited to support cannabis for depression.
4. What the Human Data Shows
Churchill and colleagues, Psychological Medicine, 2025: an updated systematic review and meta-analysis of 22 longitudinal studies, mostly in participants under 18 at the start.
- Cannabis use was associated with a higher risk of depression: OR 1.29 (95% CI 1.13–1.46).
- In exploratory meta-regression, heavy users had higher odds than non-heavy users. This reached significance only at the p < 0.10 level, so it is a hint of a dose-response relationship in the direction of harm, not proof of one.
- The 95% prediction interval was wide (0.75–2.20), risk of bias was medium across studies, and every study relied on self-reported cannabis use.
- The authors state that the findings cannot definitively establish causation and that confounding factors were not fully captured.
An earlier meta-analysis focused on adolescents points the same way. Gobbi and colleagues, JAMA Psychiatry, 2019 (11 studies, 23,317 individuals) found that cannabis use in adolescence was associated with depression in young adulthood (OR 1.37, 95% CI 1.16–1.62). It was also associated with suicidal ideation (OR 1.50, 95% CI 1.11–2.03) and suicide attempts (OR 3.46, 95% CI 1.53–7.84). The association with anxiety was not statistically significant. These data are also observational.
Read carefully, that is neither exoneration nor indictment. The association is modest, consistent in direction, possibly dose-related, and unresolved as to cause. This is the classic pattern where self-medication (depressed people use more cannabis) and causation (cannabis use worsens mood over time) both fit the data, and probably both operate.
What it is emphatically not is support for cannabis as a treatment for depression. There is no equivalent body of randomised controlled trials showing benefit.
5. Bipolar Disorder: The Claim, and What Surrounds It
The claim that circulates comes from Zucker Hillside Hospital in New York. It says that patients with bipolar I disorder who use cannabis perform better on tests of attention, processing speed and working memory than other bipolar I patients. The study behind it (Braga and colleagues, 2012) was a retrospective comparison of 200 patients, 50 of whom had a history of cannabis use disorder. That same group also had more lifetime psychosis, which the authors read as a possible sign of poorer prognosis.
Later reviews treat the paradox as unresolved. Only a few studies exist, and they do not agree on which cognitive domains, if any, differ between cannabis-using and non-using bipolar patients.
Before that is read as evidence of benefit, consider what else could produce it. The study’s own authors raised one explanation. To obtain and use cannabis regularly, a person must function well enough to do so: maintain contacts, manage money, leave the house. Illness severity varies enormously in this population. “The ones who use cannabis test better” may therefore be measuring who is well enough to use cannabis, not what cannabis does. This is confounding by indication in reverse, and no cross-sectional comparison can rule it out.
Set against that, the rest of the bipolar literature is predominantly unfavourable, based mostly on observational studies:
| Finding | Direction |
|---|---|
| Age of onset | Earlier onset of affective symptoms associated with cannabis use |
| Manic symptoms | Cannabis use associated with worsening manic symptoms in bipolar disorder, and with new-onset manic symptoms |
| Psychotic symptoms | Lifetime psychotic symptoms more common among cannabis users |
| Suicide attempts | More common among cannabis users |
| Course | Continued use associated with lower recovery and remission rates, higher recurrence and more work impairment over two years |
Sources: Pinto et al., 2019 (systematic review and meta-analysis, 53 studies); Gibbs et al., 2015 (systematic review and meta-analysis of prospective studies); Zorrilla et al., 2015 (EMBLEM). These are associations; they show what goes together, not what causes what.
One finding is more useful than all of those because it is actionable. In the EMBLEM study, bipolar patients who stopped cannabis during a manic or mixed episode had clinical and functional outcomes over two years similar to those who had never used it. That is about as close to a hopeful result as this literature gets, and it points the opposite way from the cognition claim.
6. How to Hold All of This at Once
The mechanism is real: cannabinoid signalling influences serotonergic neurons, and at low doses in animals that produces an antidepressant-like effect.
The curve is real too. Past a threshold in the same animals, the antidepressant-like effect was lost and serotonin-neuron firing was suppressed.
The human association runs the other way. It is modest, possibly dose-related and in the direction of more depression, and its cause is unresolved.
The bipolar evidence is worse. In observational data it shows earlier onset, worse manic symptoms and a poorer course. Against that stands one contested cognition finding with an obvious alternative explanation.
And the treatment evidence is absent: there is no body of controlled trials showing cannabis treats either condition.
An honest reader ends up somewhere uncomfortable, which is the correct place. “Cannabis helps depression” and “cannabis causes depression” are both too simple for what is actually known. A stronger opinion does not close the gap between them. Only trials nobody has run would do that.
7. If This Is About You
Two things are worth saying plainly.
A personal or family history of psychosis, bipolar disorder or severe depression is the strongest caution in medical cannabis. It exists because THC can precipitate psychotic symptoms and destabilise mood in vulnerable people. It applies whether or not the cannabis was obtained through a prescription.
Do not stop your medication. Antidepressants and mood stabilisers have discontinuation effects, and in bipolar disorder stopping a mood stabiliser can precipitate an episode. Discussing cannabis with your psychiatrist is a reasonable conversation to have. Unilaterally changing what you take is not.
In Thailand, this belongs with a prescriber under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis), B.E. 2568. That means prescription form PT 33, up to 30 days’ supply per prescription, single use (no refills).
If you are having thoughts of harming yourself, this is not a topic to work out alone or on a forum. In Thailand, call the Department of Mental Health hotline 1323 (24 hours). Elsewhere in Southeast Asia, call your local emergency number. You can also contact a doctor or someone close to you today.
If you live with depression or bipolar disorder and have discussed cannabis with a psychiatrist, what did they say, and what reasoning did they give? Mapping how this conversation actually goes in Thailand and the region is more useful than another round of general argument.
References
- Bambico FR, Katz N, Debonnel G, Gobbi G. Cannabinoids elicit antidepressant-like behavior and activate serotonergic neurons through the medial prefrontal cortex. J Neurosci. 2007;27(43):11700–11711. doi:10.1523/JNEUROSCI.1636-07.2007
- McGill University. Study: Cannabis a double-edged sword. Press release, 23 October 2007. Study: Cannabis a double-edged sword | Newsroom - McGill University
- Churchill V, Sutton Chubb C, Popova L, Spears CA, Pigott T. The association between cannabis and depression: an updated systematic review and meta-analysis. Psychol Med. 2025;55:e44. doi:10.1017/S0033291724003143
- Gobbi G, Atkin T, Zytynski T, et al. Association of cannabis use in adolescence and risk of depression, anxiety, and suicidality in young adulthood: a systematic review and meta-analysis. JAMA Psychiatry. 2019;76(4):426–434. doi:10.1001/jamapsychiatry.2018.4500
- Braga RJ, Burdick KE, DeRosse P, Malhotra AK. Cognitive and clinical outcomes associated with cannabis use in patients with bipolar I disorder. Psychiatry Res. 2012;200(2–3):242–245. doi:10.1016/j.psychres.2012.05.025
- Zorrilla I, Aguado J, Haro JM, et al. Cannabis and bipolar disorder: does quitting cannabis use during manic/mixed episode improve clinical/functional outcomes? Acta Psychiatr Scand. 2015;131(2):100–110. doi:10.1111/acps.12366
- Pinto JV, Medeiros LS, Santana da Rosa G, et al. The prevalence and clinical correlates of cannabis use and cannabis use disorder among patients with bipolar disorder: a systematic review with meta-analysis and meta-regression. Neurosci Biobehav Rev. 2019;101:78–84. doi:10.1016/j.neubiorev.2019.04.004
- Gibbs M, Winsper C, Marwaha S, Gilbert E, Broome M, Singh SP. Cannabis use and mania symptoms: a systematic review and meta-analysis. J Affect Disord. 2015;171:39–47. doi:10.1016/j.jad.2014.09.016
General educational information, not medical advice. Nothing here is a recommendation to use cannabis for any psychiatric condition.