By Vijay Kumar MalesuReviewed by Lauren HardakerJun 2 2025

New research uncovers how cannabis addiction, not casual use, is genetically tied to severe mental health conditions, raising red flags for policy, treatment, and prevention.

Study: The genetic relationship between cannabis use disorder, cannabis use and psychiatric disorders. Image credit: PeopleImages.com – Yuri A/Shutterstock.com

In a recent study published in the journal Nature Mental Health, researchers in the United States investigated and compared the genetic relationships and causal links between cannabis use, cannabis use disorder (CanUD), and a wide range of psychiatric disorders.

Background

In the United States, over 52 million people aged 12 and older reported using cannabis in a year, and approximately 16 million met the criteria for CanUD. With the growing legalization and rising potency of cannabis products, understanding its impact on mental health is urgent.

CanUD often occurs with conditions such as major depressive disorder (MDD), schizophrenia (SCZ), bipolar disorder (BPD), and post-traumatic stress disorder (PTSD).

Despite increasing use, the direction of the relationship remains unclear- does cannabis use lead to mental illness, or vice versa? Further research is needed to distinguish the genetic mechanisms driving cannabis use and psychiatric risk.a

About the study

Researchers used genome-wide association study (GWAS) summary statistics to explore genetic relationships between lifetime cannabis use and CanUD in relation to various psychiatric and personality traits. Psychiatric disorders analyzed included SCZ, MDD, BPD, PTSD, anorexia nervosa (AN), anxiety, and attention deficit hyperactivity disorder (ADHD). Personality traits included neuroticism, extraversion, agreeableness, conscientiousness, and openness.

Genetic correlation analyses were conducted globally using linkage disequilibrium score regression and locally using the Local Analysis of [co]Variant Association (LAVA) framework to detect regional overlaps between traits.

Mendelian randomization was used with Lthe inkage Disequilibrium Adjusted Polygenic (MRlap) method to infer causality while accounting for biases like sample overlap and weak instruments. The Two-Sample Mendelian Randomization method was used for analyses involving non-overlapping samples.

Colocalization was assessed using the Hypothesis Prioritization Colocalization (HyPrColoc) algorithm to determine whether shared variants contributed to multiple traits. Genomic Structural Equation Modeling (genomic-SEM) uncovered latent genetic structures across traits.

Researchers also adjusted for socioeconomic status by conditioning on educational attainment and household income to control for potential confounding effects in the genetic associations.

Additionally, cross-ancestry genetic correlations were assessed using African ancestry data for CanUD, although the psychiatric and personality traits were based on European ancestry datasets

Study results

Global genetic correlation analyses showed strong positive genetic correlations between CanUD and nearly all psychiatric disorders except AN. The strongest correlation was between CanUD and ADHD (genetic correlation = 0.612), followed by MDD (0.448).

Cannabis use showed increased significant correlations with openness and conscientiousness, but weaker associations with most psychiatric disorders compared to CanUD.

Local genetic correlation analysis revealed 20 significant genomic regions shared between CanUD and psychiatric conditions, especially ADHD, MDD, PTSD, SCZ, and bipolar disorder.

Only two areas showed overlap for cannabis use, both involving SCZ. A notable shared genetic locus (GULOP*rs11783093), variants in strong linkage disequilibrium with one regulating the CHRNA2 gene, was identified between CanUD and SCZ.

While GULOP is a pseudogene, its functional relevance likely resides in CHRNA2, which encodes the cholinergic receptor nicotinic alpha-2 subunit and has been previously associated with substance use behaviors.

Mendelian randomization analyses demonstrated bidirectional causal relationships between CanUD and psychiatric disorders. The most substantial causal effect of CanUD was on SCZ (effect size = 0.688), followed by ADHD, BPD, and PTSD (measured using the PTSD Checklist – Total).

In reverse analysis, MDD demonstrated the most causal effect on CanUD. In contrast, cannabis use alone showed far fewer causal links. This indicates an increased risk of ADHD.

However, several psychiatric disorders, such as MDD, BPD, SCZ, and ADHD, increased the likelihood of cannabis use.

Genomic-SEM identified three latent genetic factors. One grouped PTSD, anxiety, and MDD; a second included AN, SCZ, and BPD; and a third grouped CanUD and ADHD.

Cannabis use showed genetic links to both the second and third factors, suggesting it shares some of the same genetic influences. These results were consistent, even after accounting for educational attainment and household income, indicating that the observed genetic relationships were not only due to socioeconomic background.

However, it’s important to note that when educational attainment was used as the adjustment, some genetic correlations for cannabis use (e.g., with ADHD and PTSD subtypes) that were previously non-significant became statistically significant.

Similar genetic patterns were also observed in individuals of African ancestry for CanUD. However, further analyses are needed in non-European populations due to limited GWAS data for most psychiatric traits in these groups.

No causal relationships were identified between either cannabis trait and anorexia nervosa (AN), despite a moderate genetic correlation observed between cannabis use and AN.

Conclusions

This study shows that CanUD is genetically and causally linked to several psychiatric disorders, more strongly than lifetime occasional cannabis use. CanUD shares a genetic connection to SCZ, MDD, BPD, ADHD, and PTSD, with the risk going both ways.

In contrast, cannabis use without disorder shows weaker and more limited associations. These findings highlight that individuals genetically predisposed to CanUD may also be at higher risk for psychiatric conditions.

Notably, the researchers found no evidence that genetic predisposition to cannabis use or CanUD offers any protective or therapeutic benefit for psychiatric disorders. This challenges common perceptions, especially in contexts where cannabis is promoted as a treatment for mental health conditions like PTSD.

As cannabis becomes more widely accepted and legalized, public health initiatives must consider these risks, emphasizing prevention and early intervention for cannabis dependence.

The study authors also note several limitations, including the lack of sex-specific data, underpowered GWAS data for AN, and limited availability of non-European datasets for most psychiatric traits.

They caution that a rush to legalize cannabis for medical or recreational use should be accompanied by awareness of its potential psychiatric consequences, especially for vulnerable populations.

Download your PDF copy now!