In our latest Beyond the Abstract, we dive into cannabis use and its impact on mental health, a controversial and understudied field, but one of vital importance for both patient safety and potential relief for millions. 

Three recent peer-reviewed studies diving into various mental health indications and the effects of cannabis use shed some light on the topic, but in doing so raise further questions. 

Each reveals a fundamental challenge in substance use research, particularly around observational methodologies. Observational evidence consistently shows relationships between cannabis and mental health problems, but determining causality and whether cannabis helps or harms fundamentally depends on study design and methodology.

The studies appear to produce contradictory findings. These divergent conclusions reflect different research questions and approaches. What emerges is a consistent pattern. 

Cannabis users report more mental health symptoms than non-users, but whether this reflects cannabis causing harm, people using cannabis to self-medicate existing symptoms, placebo effects, or shared underlying risk factors remains unresolved. The distinction between subjective relief and objective benefit, between association and causation, proves critical when evaluating cannabis mental health claims. 

Lifetime cannabis use not linked to cognitive decline in large observational study

A new peer-reviewed study of over 240,000 participants, published in BMJ Mental Health, found no association between lifetime cannabis use and longitudinal cognitive decline or increased dementia risk in older adults, and no evidence of causal effects when tested using genetic analysis.

University of Oxford researchers analysed UK Biobank data collected between 2006 and 2021, alongside US Million Veteran Program data from between 2011 and 2019, examining both actual cannabis users and people genetically predisposed to cannabis use. 

The observational analysis tracked up to 18,975 cannabis users and 60,598 non-users from UK Biobank, plus 12,222 individuals with cannabis use disorder and over 210,000 controls from the Million Veteran Program.

“Within the limits of these cohorts, we found no evidence that cannabis use contributes substantially to cognitive ageing or dementia risk in older adults,” the researchers noted. “However, these results do not establish the safety of cannabis, particularly at higher doses or with prolonged use.”

The study utilised Mendelian randomisation (MR), a methodology that uses genetic variants as a natural experiment. Since genes are randomly assigned at birth, researchers can test whether people genetically predisposed to use more cannabis actually experience worse outcomes. This approach reduces confounding from lifestyle factors that typically make cannabis users different from non-users.

Researchers identified 23 genetic variants linked to cannabis use disorder and eight variants linked to lifetime cannabis use. People carrying these variants showed no differences in cognitive performance or dementia risk compared to those without them.

However, the genetic analysis was underpowered as genetic effects on cannabis use are modest, meaning large sample sizes are needed to detect associations. The null findings could reflect either a genuine absence of causal effects or insufficient statistical power to detect them. Rather than proving cannabis doesn’t affect cognition, the analysis offers complementary evidence that reduces concerns about confounding in the observational data.

Cross-sectional analyses revealed modest positive associations. Cannabis users performed slightly better on numeric memory (β=0.07, 95% CI 0.06 to 0.09) and fluid intelligence tests (β=0.12, 95% CI 0.10 to 0.13), but these findings were not supported by longitudinal or genetic evidence.

Over seven years of follow-up, cannabis use showed no association with cognitive decline, even on tests where users initially scored higher. Among 12,222 veterans with cannabis use disorder, there was no increased dementia risk in either European or African ancestry groups.

The initial cognitive advantages among cannabis users likely reflect who chooses to use cannabis rather than the effects of the drug itself. Cannabis users were younger (average 58 vs 62 years), better educated (63% vs 44% with degrees), and less economically deprived. These differences persisted despite statistical adjustment, suggesting unmeasured factors like educational quality, early-life abilities, and lifelong health behaviours explain the associations.

Males showed stronger cognitive associations than females, though the reason remains unclear.

The findings suggest cannabis use was not associated with cognitive ageing or dementia risk in older adults. However, the study relied on self-reported frequency without measuring potency, product type, or cumulative lifetime exposure.

READ MORE…

Cannabis use and mental health problems show strengthening association in canada

Associations between cannabis use and anxiety, depression, and suicidality nearly doubled in strength between 2012 and 2022 among Canadians, according to an analysis of nationally representative survey data spanning cannabis legalisation and the COVID-19 pandemic

According to the peer-reviewed study published in The Canadian Journal of Psychiatry, examining temporal trends across a decade (2012-2022) spanning legalisation and the COVID-19 pandemic, cannabis use was associated with significantly higher levels across all three mental health outcomes. 

A Canadian-led research team from McMaster University analysed data from 25,113 Canadians in the 2012 Canadian Community Health Survey and 9,861 in the 2022 Mental Health and Access to Care Survey, using structured diagnostic interviews to assess past 12-month generalised anxiety disorder, major depressive episode, and suicidality alongside self-reported cannabis use frequency.

“From 2012 to 2022, both cannabis use and depression and anxiety disorders rose, with stronger links between them over time,” the study noted. 

Between survey years, the prevalence of generalised anxiety disorder doubled from 2.6% to 5.2%, major depressive episode rose from 4.7% to 7.6%, and cannabis use increased from 11.4% to 20.7%. 

In 2022, Canadians using cannabis 2+ times weekly were 4.5 times more likely to have generalised anxiety disorder (95% CI 3.8–5.2), 5.2 times more likely to experience major depressive episode (4.6–5.7), and 5.4 times more likely to report suicidality (4.7–6.1) compared to non-users. These prevalence ratios represented substantial increases from 2012 levels of 2.3 (1.6–3.4), 3.0 (2.3–3.9), and 3.0 (2.1–4.0), respectively.

This occurred across all cannabis use frequencies, not just heavy use. Monthly or more frequent use showed significantly stronger associations with anxiety in 2022 versus 2012, while associations with depression strengthened specifically for 2+/week use. For suicidality, both weekly and 2+/week use showed strengthened associations over time.

Associations between cannabis and depression or suicidality strengthened more among youth aged 15-24 than adults, while anxiety associations strengthened more among females. All associations persisted after adjusting for other substance use and sociodemographic factors.

Potential explanations include rising cannabis potency (average THC increased from ~14% in 2012 to ~20% by 2018, with extracts now reaching 95%+ THC), increasing social acceptability, and post-legalisation accessibility facilitating self-medication attempts.

Study limitations include different sampling methods between surveys (69% response rate with in-person interviews in 2012 versus 25% telephone response in 2022) and the inability to measure cannabis potency, product type, or cumulative exposure. The cross-sectional design documents trends rather than establishing causality.

“Cannabis use was consistently related to a higher prevalence of internalising problems, and these associations strengthened between 2012 and 2022.”

The researchers emphasised that ‘given continued escalations in co-occurring cannabis use and internalising problems, greater investment in early identification, intervention, and access to integrated substance use and mental health treatment is needed.’

Florida medical cannabis users report anxiety relief over 45-days

A separate study conducted by researchers at Florida Gulf Coast University, published in Scientific Reports in February 2026, appeared to contradict the above study’s findings in relation to anxiety. 

Floridian medical cannabis patients diagnosed with anxiety disorders reported consistent symptom relief across 45 days of tracking, with cannabis use emerging as the strongest predictor of anxiety reduction compared to prescription medications and lifestyle activities, according to the prospective observational study. 

However, it’s worth noting that the study’s design, which lacked randomisation, blinding, or placebo controls, suggests the findings show that people who choose cannabis for anxiety believe it works, rather than establishing whether cannabis pharmacologically reduces anxiety symptoms.

Researchers tracked 416 patients through smartphone-based daily reporting, generating a total of 11,164 observation days. Participants rated their anxiety before and after interventions using a simple 0-10 scale, not a validated clinical anxiety measure.

Cannabis users reported average anxiety reductions of 3.56 points, with 97% reporting positive relief. However, same-day before-and-after comparisons are highly susceptible to demand characteristics (reporting what’s expected), regression to the mean (extreme scores naturally moving toward average), and pharmacological effects unrelated to therapeutic benefit, such as intoxication-induced mood alteration, which differs from genuine anxiety treatment

The 97% positive relief rate likely reflects survival bias. This would mean patients experiencing worsening or no benefit would discontinue use and drop from tracking. The study recruited participants from CannaMD, a statewide cannabis physician network, creating selection bias toward those seeking cannabis treatment. Ninety-one percent had used cannabis before enrollment. 

Anxiety relief remained stable throughout the 45 days with no evidence of tolerance effects. Neither age, sex, route of administration, nor duration of previous cannabis use significantly influenced outcomes. Products were predominantly THC-dominant (86%), with smoking (60.5%) and vaping (23.8%) most common.

Critical limitations undermine the findings, however. The study recruited 91% experienced cannabis users (only 39 treatment-naive participants), creating severe prevalent user bias. People who previously found cannabis unhelpful would have stopped using it before the study began, leaving only those who already responded positively. The 97% positive relief rate likely reflects this selection bias rather than genuine treatment efficacy.

The comparison between novice and experienced users is drastically underpowered, with only 39 treatment-naive participants versus 401 experienced users, limiting any meaningful conclusions about differential effects. Participants were recruited from CannaMD, a statewide cannabis physician network, further biasing the sample toward those actively seeking cannabis treatment.

Additionally, the study lacked randomisation, blinding, or placebo controls; used a non-validated 0-10 anxiety rating rather than established clinical measures; and relied on same-day pre-post comparisons that cannot distinguish therapeutic benefit from intoxication effects, demand characteristics (reporting expected results), or regression to the mean.