Exploring how ADHD, autism, OCD, Ehlers-Danlos/hypermobility, POTS, MCAS, and identity (transgender, gay/bisexual) intersect at the genetic, clinical, and epidemiological level. Which connections are real biology and which are ascertainment artifacts? This content was AI-generated and has been primarily AI fact-checked, not personally verified by the author.
Each cell represents the relationship between two conditions. Color intensity indicates strength; hue indicates direction. Values show genetic correlation (rg) where available, or clinical association strength. Click any cell to jump to the detailed evidence.
Click any card to expand. Prevalence is shown as population estimate. The bar represents twin-study heritability.
For each researched pair: genetic correlation, clinical comorbidity, direction, proposed mechanisms, and evidence strength.
From Genomic SEM and clinical evidence, these clusters emerge. Note that Tourette cross-loads on both Neurodevelopmental and Compulsive factors.
The p-factor (Caspi & Moffitt 2014) is a general dimension of psychopathology capturing shared liability across mental disorders -- analogous to the g-factor of intelligence. Phenotypic data consistently supports it, but at the genomic level, a bifactor model with a genetic p-factor fits worse than correlated factors models.
Four to five correlated genetic factors explain psychiatric comorbidity better than a single general factor: Compulsive (OCD, anorexia, Tourette), Psychotic (schizophrenia, bipolar), Neurodevelopmental (ADHD, autism), Internalizing (depression, anxiety, PTSD), and Substance Use. These factors are themselves correlated, explaining the appearance of a general factor.
~90% of GWAS-identified variants are pleiotropic. The 2019 PGC study found 109 pleiotropic loci across 8 psychiatric disorders, with 23 affecting 4+ disorders. The DCC gene (netrin-1 receptor) was associated with all 8 disorders. 75% of genome-wide significant psychiatric SNPs are pleiotropic.
The omnigenic model (Boyle et al. 2017) predicts that essentially all genes expressed in disease-relevant tissues contribute to disease risk, making cross-disorder overlap almost inevitable for conditions affecting the same cell types.
DCC (netrin-1 receptor) -- associated with all 8 major psychiatric disorders. Guides axonal growth during neurodevelopment.
RBFOX1 -- associated with 7 disorders. RNA-binding protein regulating neuronal splicing.
CACNA1C -- 5+ disorders. Voltage-gated calcium channel.
NRXN1 -- rare CNVs increase Tourette (OR=20.3), autism, schizophrenia risk.
These converge on neurodevelopmental pathways active from the second prenatal trimester.
The 2025 hEDS GWAS is a breakthrough: first molecular genetic evidence linking connective tissue and neurodevelopmental conditions. Genetic correlations: ME/CFS (rg=0.35), chronic pain (rg=0.33), IBS (rg=0.31), migraine (rg=0.30), fibromyalgia (rg=0.21), depression (rg=0.17), autism (rg=0.13).
The strongest cross-domain link (connective tissue to neurodevelopment) is indirect -- through shared associations with ME/CFS, chronic pain, and depression, rather than direct locus overlap. The hEDS-autism rg of 0.13 is significant but small.
No GWAS exists for: gender dysphoria/transgender identity, POTS, or MCAS. This means genetic correlations with other conditions cannot be computed for these. This is the largest evidence gap in this entire research project.
Scott Alexander has written ~40+ posts touching on this cluster. No single post treats the full phenomenon, but his body of work is one of the most sophisticated lay treatments of these questions. Key posts below.
All claims in this research are tagged with a confidence tier. Higher tiers indicate stronger, more replicated evidence.