The Genetics of Trying Cannabis Are Different from the Genetics of Cannabis Addiction | Research

The takeaway

A massive genetic study found 11 gene variants linked to ever trying cannabis — prominently CADM2 — and showed the genetics of trying cannabis overlap with risk-taking and openness but differ from cannabis addiction genetics.

Geneticists, addiction researchers, public health professionals, and anyone interested in the biology of drug experimentation

What the researchers found

11 independent genome-wide significant variants for cannabis lifetime use, most prominently CADM2*rs7609594 (p=7.4×10⁻²⁰); genetic correlations with risk-taking, openness, and substance use; bidirectional causal relationships via Mendelian randomization; distinct from CUD genetic architecture.

Why it matters

The genetic distinction between trying cannabis and developing addiction is crucial — it means different people need different prevention approaches, and trying cannabis alone isn't a genetic pathway to addiction.

The numbers in context

258,823 effective sample; 11 significant variants; CADM2*rs7609594 p=7.4×10⁻²⁰; bidirectional MR causality with risk-taking and sexual behaviors; 6 traits locally correlated with CanLU at CADM2.

How the study worked

Multi-ancestral GWAS using All of Us data meta-analyzed with prior studies (effective N=258,823), followed by genetic correlation (LDSC), local genetic correlation (LAVA), Mendelian randomization, and phenome-wide association analysis.

What this study cannot tell us

Lifetime use is a binary measure missing frequency and quantity; multi-ancestral meta-analysis still dominated by European ancestry; GWAS associations are not actionable individual predictions; MR assumptions may be violated.

How to read the evidence

Large multi-ancestral GWAS with robust post-hoc analyses including Mendelian randomization, published in a top neuropsychopharmacology journal.

When this study was published

Published in 2026 using All of Us and other cohorts, providing the most comprehensive genetic analysis of cannabis lifetime use to date.

The bigger picture

CADM2's central role links cannabis experimentation to a broader biology of exploration and risk-taking — suggesting that cannabis 'gateway' effects may be driven by shared genetics of novelty-seeking rather than pharmacological progression.

Questions still open

Could CADM2-based risk stratification identify youth needing targeted prevention? How do CUD-specific genetic variants differ from CanLU variants mechanistically?

Common questions

Is trying cannabis genetic?

Partly — this study found 11 gene variants associated with ever trying cannabis, most prominently in the CADM2 gene, which is also linked to risk-taking and openness to experience.

Are the genetics of trying cannabis the same as cannabis addiction?

No — this study found the genetic architecture of cannabis experimentation is distinct from cannabis use disorder, suggesting different biological pathways for trying versus becoming addicted.

Read the original research

The genetics of cannabis lifetime use.

Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 51(3), 554-564

View on PubMed ↗Publisher’s paper ↗

Citation

Bright, Uri; Beck, Sarah; Galimberti, Marco; Gupta, Priya; Chen, Yu; Dao, Cecilia; Nunez, Yaira Z; Kranzler, Henry R; Zhou, Yu; Zhang, Yingzhe; Choi, Karmel W; Levey, Daniel F; Gelernter, Joel. (2026). The genetics of cannabis lifetime use.. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 51(3), 554-564. https://doi.org/10.1038/s41386-025-02255-4