Genetic Risk for Mental Illness Is Far Less Disorder-Specific Than Psychiatry Has Assumed

Psychiatry has been organized, for more than a century, around the idea that different mental illnesses are distinct entities. Different names, different diagnostic criteria, different treatments. The classification system most widely used globally - the Diagnostic and Statistical Manual of Mental Disorders - lists hundreds of specific conditions, each with its own profile.

Genetic research has complicated that picture for decades. Family studies, twin studies, and more recently large-scale genomic analyses have consistently shown that many psychiatric conditions share a substantial portion of their genetic architecture. Someone with schizophrenia has elevated genetic risk not just for schizophrenia but for bipolar disorder and sometimes for autism and depression as well. The boundaries that clinical psychiatry draws seem not to correspond to the boundaries that genetic inheritance respects.

A study led by Dr. Kenneth S. Kendler at Virginia Commonwealth University, published in Genomic Psychiatry, introduces a framework for quantifying exactly how pronounced this problem is - and finds that across nine major psychiatric disorders, the answer varies dramatically and consequentially.

Defining Genetic Specificity

The concept Kendler's team introduces is "genetic specificity": the percentage of a person's genetic vulnerability to a given disorder that points exclusively toward that specific condition, rather than contributing to risk for other psychiatric disorders as well. High genetic specificity means that when you inherit genetic risk for condition X, most of that risk is particular to X. Low specificity means the inherited risk is diffuse - it raises your chances of developing any one of several different disorders, but it does not specifically predict which one.

To measure this, the team analyzed familial data from more than two million individuals born in Sweden between 1950 and 1995. Sweden's universal health registries provide remarkably complete population-level psychiatric diagnosis records, making it one of the best sources on earth for this kind of large-scale familial analysis. The nine disorders studied included schizophrenia, bipolar disorder, major depressive disorder, anxiety disorders, autism spectrum disorder, attention deficit hyperactivity disorder, anorexia nervosa, obsessive-compulsive disorder, and alcohol use disorder.

A Wide Range, With Surprises

The results revealed enormous variation in genetic specificity across disorders. Autism spectrum disorder showed relatively high genetic specificity - roughly half of its genetic risk was estimated to be disorder-specific. Schizophrenia was similarly distinctive. Anorexia nervosa also showed relatively high specificity.

At the other end of the spectrum were major depressive disorder and generalized anxiety disorder. Their genetic specificity was estimated to be close to zero - nearly all of the genetic risk for these conditions appears to be shared with other psychiatric disorders rather than pointing specifically at depression or anxiety. The genetic architecture of depression, in this framing, is not really "the genetics of depression" - it is a generic vulnerability to psychiatric disorder that, in one person's life circumstances, manifests as depression.

These are not fixed constants but variable quantities shaped by clinical features of how each disorder is defined. The study also found that specificity changes depending on factors such as the severity of the disorder, the age at which it first appears, and whether it occurs alongside other conditions. A severe, early-onset case of a disorder tends to show higher genetic specificity than a mild, late-onset case - consistent with the clinical intuition that early and severe presentations represent a more biologically "pure" form of the condition.

What This Means for Psychiatric Classification

The findings add to a substantial body of evidence suggesting that the current psychiatric diagnostic system does not map cleanly onto biological reality. If two conditions share nearly all of their genetic risk, they may be better understood as variant expressions of a shared underlying vulnerability than as distinct diseases that happen to look similar.

The practical implications are significant. Drug development has historically been organized around diagnostic categories - researchers test whether a compound works for schizophrenia or for depression, not for "high genetic specificity psychiatric disorder." If the boundaries are biologically arbitrary, this approach may explain why so many psychiatric drug candidates succeed in early trials but fail in broader patient populations that include both high- and low-specificity presentations.

The study has limitations inherent to its design. Familial risk estimation in registry data captures genetic and environmental factors that run in families, but cannot fully separate the two. The Swedish population has particular demographic characteristics that may not generalize globally. And while the dataset is impressively large, the quality of psychiatric diagnosis coding in administrative records varies.