What If Psychiatric Diagnoses Are Describing Symptoms — Not Biology?

Hook

What if many psychiatric diagnoses aren’t separate diseases at all — but different surface expressions of shared biology?

A major cross-disorder genetics paper in Nature suggests that might be closer to the truth than psychiatry’s current labels imply.

Context: When Biology Doesn’t Respect Diagnostic Boundaries

The study analyzed genetic data across 14 psychiatric disorders using very large GWAS datasets. Instead of finding 14 cleanly separated genetic signatures, the researchers found the data clustered into five underlying genomic factors that explained most of the shared genetic liability across diagnoses.

This matters because psychiatry is built on a diagnostic structure that is reliable at describing what clinicians observe — but often weak at explaining why those symptom patterns happen.

The DSM Problem: Descriptive, Not Mechanistic

The DSM is a descriptive framework. It was designed to:

standardize diagnosis across clinicians
define syndromes using symptom checklists
support research, services, and insurance coding

But it was never built to be a map of biology.

So when genetics shows that conditions with separate DSM criteria share substantial risk architecture, that is not a minor technical detail — it’s a structural mismatch:

DSM categories describe outcomes
genetics starts to reveal mechanisms

Why Comorbidity Is the Rule (Not an Exception)

Clinically, many people carry multiple diagnoses over time: depression + anxiety; PTSD + depression; eating disorders + OCD traits; bipolar disorder with psychosis.

The DSM treats this as “comorbidity” — multiple co-occurring conditions.

But a cross-disorder genetic model suggests a different interpretation:

Many so-called distinct disorders may share upstream vulnerabilities, with symptoms diverging downstream.

That makes comorbidity less mysterious and more expected.

What This Could Mean for Treatment

If shared biology cuts across diagnoses, then borrowing treatment approaches across diagnostic lines becomes more than an improvisation — it becomes rational.

This is already happening in practice:

Trauma-focused approaches sometimes help subsets of “treatment-resistant” depression.
SSRIs span depression, anxiety, OCD, and PTSD-related symptom clusters.
Mood stabilizers and antipsychotics get used across a blurry mood–psychosis spectrum.

The genetics doesn’t “prove” any specific regimen — but it strengthens the argument that the treatment target should increasingly be the disrupted system, not just the label.

A useful reframe is:

Treatment resistance can mean “wrong target,” not “difficult patient.”

From Labels to Pathways

Instead of asking only:

“Which DSM box fits best?”

An emerging model asks:

“Which systems look dysregulated in this person?”

Examples include:

stress / threat circuitry
compulsive habit loops
excitatory–inhibitory signaling balance
neurodevelopmental wiring vulnerability
reward and addiction pathways

Different labels can reflect the same underlying system — and the same system can present differently depending on development, environment, and context.

Important Limits (No Hype)

This does not mean:

genetic testing can currently match people to medications
one drug will treat many disorders
diagnoses are obsolete

It does mean:

The DSM describes what we see.
Genetics starts to explain what’s going wrong.