rethinkTHC Search
Menu
Study breakdown

Cannabis use leaves a distinct DNA methylation signature in people with first-episode psychosis

Cross SectionalModerate evidence
The takeaway

A meta-analysis of two cohorts found that current cannabis use was associated with specific DNA methylation changes at genes involved in immune and mitochondrial function, with high-potency THC showing its own distinct epigenetic signature.

Researchers studying the biological mechanisms linking cannabis use and psychosis, and readers interested in epigenetics.

Cannabis-specific methylation changes at CAVIN1 and MCU genes, independent of tobacco effects

What the researchers found

Current cannabis use was associated with differential methylation at a site in the CAVIN1 gene, independent of tobacco effects. High-potency cannabis (THC >= 10%) was associated with a site in the MCU gene. Pathway analysis identified epigenetic variation near genes linked to immune and mitochondrial function. An interaction analysis suggested that first-episode psychosis status may moderate how cannabis affects DNA methylation.

Why it matters

Epigenetic changes represent a molecular mechanism through which cannabis exposure could have lasting biological effects. Identifying cannabis-specific methylation changes, separate from tobacco, and finding that psychosis status may modify these effects adds mechanistic insight to the cannabis-psychosis relationship.

The numbers in context

682 participants across two cohorts (188 current users, 494 never users). One CpG site (cg11669285) in CAVIN1 reached array-wide significance for current use. One CpG in MCU reached significance for high-potency (THC >= 10%) use. Both were independent of the tobacco methylation signature.

How the study worked

Two independent cohorts totaling 682 participants (188 current cannabis users, 494 never users) with first-episode psychosis and controls. DNA methylation profiles were generated from blood samples using Illumina arrays. Meta-analysis across cohorts identified significant sites, controlled for tobacco-related methylation signatures.

What this study cannot tell us

Cross-sectional design cannot determine whether methylation changes preceded or followed cannabis use or psychosis onset. Blood-based methylation may not reflect brain tissue changes. The modest sample of 188 users limits power to detect smaller effects. Only current use was assessed, not lifetime exposure patterns.

How to read the evidence

Meta-analysis of two independent cohorts with 682 participants and array-wide significance thresholds, though cross-sectional and blood-based.

When this study was published

Published in 2025.

The bigger picture

The finding that immune and mitochondrial pathways are epigenetically affected by cannabis aligns with known cannabinoid biology. The interaction between cannabis use and psychosis status on methylation patterns raises the possibility that people who develop psychosis may process cannabis exposure differently at the molecular level.

Questions still open

  • Do these methylation changes persist after stopping cannabis use? Are the immune and mitochondrial pathway changes functionally significant? Could methylation profiles serve as biomarkers for cannabis-related psychosis risk?

Common questions

What is DNA methylation?
It is a chemical modification to DNA that can turn genes on or off without changing the genetic code. Environmental exposures like cannabis can alter methylation patterns, potentially affecting how genes function.
Did this prove cannabis causes psychosis?
No. The study found that cannabis use is associated with specific molecular changes, and that psychosis status may modify these changes. But cross-sectional data cannot establish which came first.
Were the effects from THC specifically?
High-potency cannabis (THC >= 10%) showed a distinct methylation site compared to general cannabis use, suggesting THC concentration matters for epigenetic effects.

Read the original research

Methylomic signature of current cannabis use in two first-episode psychosis cohorts.

Molecular psychiatry, 30(4), 1277-1286

Citation

Dempster, Emma L; Wong, Chloe C Y; Burrage, Joe; Hannon, Eilis; Quattrone, Diego; Trotta, Giulia; Rodriguez, Victoria; Alameda, Luis; Spinazzola, Edoardo; Tripoli, Giada; Austin-Zimmerman, Isabelle; Li, Zhikun; Gayer-Anderson, Charlotte; Freeman, Tom P; Johnson, Emma C; Jongsma, Hannah E; Stilo, Simona; La Cascia, Caterina; Ferraro, Laura; La Barbera, Daniele; Lasalvia, Antonio; Tosato, Sarah; Tarricone, Ilaria; D'Andrea, Giuseppe; Galatolo, Michela; Tortelli, Andrea; Pompili, Maurizio; Selten, Jean-Paul; de Haan, Lieuwe; Menezes, Paulo Rossi; Del Ben, Cristina M; Santos, Jose Luis; Arrojo, Manuel; Bobes, Julio; Sanjuán, Julio; Bernardo, Miguel; Arango, Celso; Jones, Peter B; Breen, Gerome; Mondelli, Valeria; Dazzan, Paola; Iyegbe, Conrad; Vassos, Evangelos; Morgan, Craig; Mukherjee, Diptendu; van Os, Jim; Rutten, Bart; O'Donovan, Michael C; Sham, Pak; Mill, Jonathan; Murray, Robin; Di Forti, Marta. (2025). Methylomic signature of current cannabis use in two first-episode psychosis cohorts.. Molecular psychiatry, 30(4), 1277-1286. https://doi.org/10.1038/s41380-024-02689-0

Explore the wider topic