A cross-disorder GWAS meta-analysis of nearly 800,000 individuals implicates the DAGLA gene and eight other endocannabinoid loci in five major mental disorders, but replication and functional validation are needed before these findings can inform clinical practice or drug development.
Why This Matters
Major psychiatric disorders, including depression, bipolar disorder, ADHD, autism spectrum disorder, and schizophrenia, share substantial genetic overlap, yet the specific biological pathways driving that overlap remain poorly mapped. The endocannabinoid system is a plausible candidate given its established roles in mood regulation, synaptic plasticity, and neurodevelopment. Identifying shared genetic architecture involving endocannabinoid genes could open a path toward transdiagnostic biomarkers or therapeutic targets. This 2023 meta-analysis represents the largest genetic investigation of endocannabinoid pathway involvement across multiple psychiatric conditions to date.
Clinical Summary
Psychiatric genetics has increasingly moved toward transdiagnostic designs that pool data across multiple conditions to identify shared biological pathways. This 2023 candidate gene GWAS meta-analysis, published in Molecular Psychiatry, drew on publicly available summary statistics from six Psychiatric Genomics Consortium datasets encompassing 284,023 cases and 508,515 controls of European ancestry. The investigators tested 2,241 SNPs across 33 pre-specified endocannabinoid system genes for association with risk of five psychiatric disorders: major depressive disorder, bipolar disorder, ADHD, autism spectrum disorder, and schizophrenia. The mechanistic rationale centers on the endocannabinoid system’s role in regulating synaptic transmission and neuroplasticity through endogenous lipid signaling molecules, particularly 2-arachidonoylglycerol (2-AG), which is synthesized by the enzyme diacylglycerol lipase alpha, or DAGLA.
Nine SNPs reached the study-specific significance threshold, with the lead variant, rs12805732 in the DAGLA gene, showing the strongest association both in the single-SNP analysis and in a complementary gene-based analysis using MAGMA. However, four of the nine significant SNPs displayed substantial heterogeneity across disorders (I-squared values exceeding 60%), meaning their effect sizes and directions varied meaningfully among the five conditions rather than reflecting a single unified genetic influence. The analysis was restricted to European-ancestry cohorts, limiting generalizability, and the candidate gene design means no novel endocannabinoid-related loci outside the 33 pre-specified genes could be discovered. The authors acknowledge that independent replication in external cohorts and functional validation studies are necessary before any clinical or pharmacological conclusions can be drawn from these findings.
Dr. Caplan’s Take
This study does something valuable: it applies a rigorous statistical framework to a question many of us working in cannabinoid medicine have been circling for years. The idea that the endocannabinoid system plays a role in psychiatric vulnerability is not new, but seeing DAGLA emerge as the strongest signal in a dataset of nearly 800,000 people is noteworthy. That said, patients who read headlines about this study will arrive in clinic asking whether their genetics explain their depression or whether cannabinoid therapies are supported by this evidence. They are not. This is a statistical association in a candidate gene analysis, not a causal pathway or a treatment rationale.
In practice, I use findings like these to contextualize why some patients may experience mood or anxiety changes with cannabis use, and to reinforce that the endocannabinoid system is biologically important in brain health. But I do not alter clinical recommendations based on candidate gene associations that have not been replicated. What I do is monitor the literature, discuss what we know and do not know honestly, and ensure patients understand that genetic predisposition studies are the very beginning of a research pipeline, not its conclusion.
Clinical Perspective
This study sits at an early stage in the research arc connecting endocannabinoid genetics to psychiatric phenotypes. It confirms that endocannabinoid system genes harbor variants associated with mental disorder risk in aggregate, consistent with smaller candidate gene studies and preclinical models implicating 2-AG signaling in stress responsivity and synaptic regulation. However, the substantial heterogeneity across disorders for nearly half the significant variants challenges the authors’ framing of these as truly “shared” risk loci. The evidence supports further investigation of DAGLA and related enzymes as biologically relevant candidates, but it does not support patient-facing claims about endocannabinoid genetic risk profiles or cannabinoid-based therapeutic targeting for psychiatric conditions.
From a pharmacological standpoint, DAGLA inhibition or modulation remains a preclinical research domain with no approved agents in clinical development for psychiatric indications. Clinicians should be aware that patients using exogenous cannabinoids, whether medical or recreational, are interacting with the same signaling system implicated here, but this study provides no basis for predicting individual response, risk, or benefit. Importantly, patients with psychiatric diagnoses who use cannabis should be monitored for symptom exacerbation irrespective of these genetic findings. The most actionable takeaway for clinicians is to track replication studies and functional work on DAGLA as they emerge, while resisting premature translation of these associations into practice.
Study at a Glance
- Study Type
- Candidate gene cross-disorder GWAS meta-analysis using summary statistics
- Population
- European-ancestry adults and children with MDD, bipolar disorder, ADHD, ASD, or schizophrenia
- Intervention
- Not applicable (observational genetic study)
- Comparator
- 508,515 controls without psychiatric diagnoses
- Primary Outcomes
- Association of 2,241 endocannabinoid gene SNPs with cross-disorder psychiatric risk
- Sample Size
- 284,023 cases and 508,515 controls (total approximately 792,538)
- Journal
- Molecular Psychiatry
- Year
- 2023
- DOI or PMID
- Summary statistics sourced from six PGC GWAS datasets; see original publication for specific DOI
- Funding Source
- Not specified in available data; constituent datasets funded through PGC consortium sources
What Kind of Evidence Is This
This is a candidate gene GWAS meta-analysis, a secondary analysis that applies genome-wide association meta-analytic methods to summary statistics from six existing Psychiatric Genomics Consortium datasets, but restricts its scope to 33 pre-specified endocannabinoid system genes. It sits in the middle tier of the evidence hierarchy for genetic epidemiology: large in sample size but limited by its candidate gene design, which means it can only confirm or refute associations within genes the investigators already hypothesized to be relevant. The single most important inference constraint is that no novel endocannabinoid-related loci outside these 33 genes can be discovered, and results remain correlational.
How This Fits With the Broader Literature
The finding that DAGLA harbors psychiatric risk variants aligns with preclinical research demonstrating that 2-AG signaling modulates stress-related behaviors, fear extinction, and synaptic plasticity in animal models. Earlier candidate gene studies of CNR1, the gene encoding cannabinoid receptor 1, have produced mixed results in psychiatric populations, and this meta-analysis notably identifies DAGLA rather than receptor genes as the strongest signal. The Cross-Disorder Group of the PGC has previously identified shared genetic architecture across psychiatric conditions using genome-wide approaches, and this study extends that framework to a specific biological system. However, unlike those hypothesis-free genome-wide analyses, the candidate gene restriction here means the findings should be viewed as hypothesis-confirming rather than hypothesis-generating, and replication in non-European populations and independent datasets remains essential.
Common Misreadings
The most likely overinterpretation is concluding that the endocannabinoid system has been established as a shared genetic mechanism across all five psychiatric disorders studied. The heterogeneity data directly undermine this reading: with four of nine significant SNPs showing I-squared values above 60%, the effects of these variants differ meaningfully across conditions. A related misreading involves treating these associations as evidence that cannabinoid-based therapies could address transdiagnostic psychiatric risk. Statistical association with a gene involved in endocannabinoid synthesis does not imply that modulating that system pharmacologically would be therapeutic, and the study provides no functional or clinical data to bridge that gap.
Bottom Line
This large meta-analysis provides preliminary genetic evidence that variants in the DAGLA gene and other endocannabinoid system loci are statistically associated with cross-disorder psychiatric risk. The findings are hypothesis-generating, not definitive, constrained by candidate gene design, European-only ancestry, absence of replication, and substantial heterogeneity across disorders. For clinical practice, these results warrant attention and tracking but do not yet support changes to patient care, cannabinoid prescribing, or genetic risk counseling.
References
- Cross-Disorder Group of the Psychiatric Genomics Consortium. Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide analysis. Lancet. 2013;381(9875):1371-1379. doi:10.1016/S0140-6736(12)62129-1
- Demontis D, Walters RK, Martin J, et al. Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder. Nat Genet. 2019;51(1):63-75. doi:10.1038/s41588-018-0269-7
- Wray NR, Ripke S, Mattheisen M, et al. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018;50(5):668-681. doi:10.1038/s41588-018-0090-3
- Stahl EA, Breen G, Forstner AJ, et al. Genome-wide association study identifies 30 loci associated with bipolar disorder. Nat Genet. 2019;51(5):793-803. doi:10.1038/s41588-019-0397-8
- Grove J, Ripke S, Als TD, et al. Identification of common genetic risk variants for autism spectrum disorder. Nat Genet. 2019;51(3):431-444. doi:10.1038/s41588-019-0344-8
- Schizophrenia Working Group of the Psychiatric Genomics Consortium. Biological insights from 108 schizophrenia-associated genetic loci. Nature. 2014;511(7510):421-427. doi:10.1038/nature13595
- Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26(17):2190-2191. doi:10.1093/bioinformatics/btq340
- de Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: generalized gene-set analysis of GWAS data. PLoS Comput Biol. 2015;11(4):e1004219. doi:10.1371/journal.pcbi.1004219
Have thoughts on this? Share it:
