Mental Health

Shared Genetics Across Psychiatric Disorders

An evidence-based guide to how major psychiatric disorders overlap genetically, what this suggests about underlying biology, and implications for diagnosis and treatment.

6 min read

Intro

Psychiatric disorders are typically diagnosed using symptom-based criteria. Large-scale genetic research increasingly shows that many conditions regarded as distinct in clinical practice share substantial genetic risk architecture and may converge on shared biological pathways.

This guide explains what cross-disorder genetic studies find, why this creates tension with traditional diagnostic systems, and what it may imply for diagnosis and treatment.

Key Points

  • Many psychiatric disorders share polygenic risk (many variants with small effects).
  • A major cross-disorder analysis across 14 conditions identified five genomic factors explaining ~66% of shared genetic liability.
  • The DSM is descriptive (syndromes), while genetics begins to illuminate mechanisms (pathways, cell types).
  • Comorbidity is expected when disorders share upstream vulnerabilities.
  • These findings support pathway-informed research and treatment development, not immediate replacement of diagnostic categories.

Background: What “Shared Genetics” Means

Most psychiatric conditions are influenced by many common genetic variants, each contributing a small amount of risk. A key complication is pleiotropy, where the same variant can influence multiple traits or diagnoses.

Shared genetics does not mean disorders are identical. It means:

  • there can be common upstream vulnerabilities
  • symptoms can diverge downstream due to development, environment, and other biological factors

The Nature Cross-Disorder Study (14 Disorders)

A large international collaboration analyzed GWAS data spanning 14 disorders:

  • ADHD
  • Alcohol-use disorder
  • Anorexia nervosa
  • Anxiety disorders
  • Autism spectrum disorder
  • Bipolar disorder
  • Cannabis-use disorder
  • Major depressive disorder
  • Nicotine dependence
  • Obsessive-compulsive disorder
  • Opioid-use disorder
  • PTSD
  • Schizophrenia
  • Tourette syndrome

The authors report five underlying genomic factors capturing shared liability across disorders, along with hundreds of pleiotropic loci and pathway / cell-type enrichments.

Five Genomic Factors

Instead of 14 fully distinct genetic signatures, disorders grouped into five broad dimensions:

  1. Schizophrenia + Bipolar (SB) factor
  2. Internalizing factor (major depression, PTSD, anxiety)
  3. Compulsive factor (e.g., OCD, anorexia; with Tourette and anxiety loading more weakly)
  4. Neurodevelopmental factor (autism, ADHD; with Tourette loading more weakly)
  5. Substance-use disorders (SUD) factor (opioid, cannabis, alcohol, nicotine; with ADHD loading more weakly)

A key implication: within some clusters, the shared genetic signal is large and disorder-specific signal is relatively small.

Descriptive vs Mechanistic Psychiatry

DSM is descriptive

The DSM improves reliability by grouping symptom patterns into syndromes. It is designed to answer:

  • What does this look like?
  • How severe is it?
  • How long has it lasted?

Genetics is closer to mechanistic

Genetics begins to answer:

  • Which pathways are implicated?
  • Which cell types show enrichment?
  • Why do diagnostic boundaries blur?

This helps explain why DSM categories may not map neatly onto biology, and why comorbidity is common.

Implications for Diagnosis

These findings do not eliminate diagnoses. They suggest a complementary model:

  • Diagnosis remains useful for communication, services, and treatment planning.
  • Biology-informed dimensions may improve:
    • risk stratification
    • explanation of mixed presentations
    • identification of subtypes within a diagnosis

Over time, diagnostic systems may become more dimensional (e.g., internalizing vs compulsive vs psychotic-spectrum) rather than purely categorical.

Implications for Treatment

Why cross-diagnostic treatments can work

Some treatments already span disorders (e.g., SSRIs across depression/anxiety/OCD). Shared genetic and pathway signals provide a plausible biological rationale for why cross-diagnostic borrowing can be effective in subsets of patients.

Pathway-informed development and repurposing

If clusters implicate specific biology (e.g., excitatory neuron enrichment in SB; oligodendrocyte biology in internalizing), this may guide:

  • novel target discovery
  • repurposing strategies
  • mechanism-based trials that recruit by symptom domains or biomarkers rather than diagnosis alone

“Treatment resistance” reframed

A useful clinical frame is that non-response can reflect:

  • wrong target
  • wrong mechanism assumption rather than patient “complexity” alone.

Limitations

  • Large psychiatric GWAS are still uneven by ancestry; many analyses remain biased toward European-ancestry datasets.
  • Individual variants have small effects; genetics is probabilistic, not deterministic.
  • Mapping genes → pathways → symptoms remains incomplete.

FAQ

Q: Does this prove psychiatric diagnoses are wrong?
A: No. It shows diagnoses are descriptive syndromes that may not align cleanly with underlying biology.

Q: Can I use genetic testing to pick the best medication today?
A: Not reliably. Current findings inform research and future stratification more than individual prescribing.

Q: Why does this matter for patients?
A: It helps explain comorbidity, overlapping symptoms, and why treatments can work across labels in some people.

Further Reading