Psychiatry has treated depression as depression, schizophrenia as schizophrenia, and ADHD as ADHD – distinct conditions requiring distinct diagnoses, distinct medications, and distinct treatment pathways. Two landmark genetic studies published in 2025, each arriving from different directions, reached the same conclusion: the boundaries drawn between many of the most prevalent psychiatric conditions may have less to do with underlying biology than with the limits of how medicine has historically organized its diagnostic categories.
A person diagnosed with bipolar disorder is statistically more likely to also carry a diagnosis of anxiety. Someone with ADHD frequently co-presents with depression. Schizophrenia and bipolar disorder, long treated as categorically separate conditions requiring entirely different pharmaceutical approaches, share so much genetic territory that researchers have begun asking whether separating them was ever fully justified.
Two major research teams, using genomic data drawn from more than a million people, uncovered a psychiatric disorders genetic link that runs deep into human brain development, threading through shared molecular pathways that influence how neurons form, how proteins connect, and how risk accumulates across developmental stages.
The Eight-Disorder Study: What the Cell Research Found
In 2019, researchers at the Psychiatric Genomics Consortium, Harvard University, and the UNC School of Medicine identified 136 genomic “hot spots” associated with eight psychiatric disorders. Among those, 109 hot spots were shared across multiple disorders – a property scientists call “pleiotropic.” Knowing that shared genetic regions exist is a different matter from understanding what they actually do – and whether those shared variants function differently from variants that affect only a single condition.
A follow-up study published in early 2025 in the journal Cell addressed that gap. The study was led by Hyejung Won, PhD, associate professor in the Department of Genetics and the UNC Neuroscience Center, and Patrick Sullivan, MD, FRANZCP, the Yeargen Distinguished Professor of Psychiatry and Genetics. They delineated the functional consequences of genetic variants into two groups, finding that pleiotropic variants may be optimal targets for treatment due to their extended roles in development and sensitivity to change.
The Eight Disorders Involved
The research drew from genome-wide studies of anorexia nervosa, attention-deficit/hyperactivity disorder, autism spectrum disorder, bipolar disorder, major depression, obsessive-compulsive disorder, schizophrenia, and Tourette syndrome, analyzing data from 232,964 cases and 494,162 controls.
The MPRA Methodology
The team used a massively parallel reporter assay (MPRA), a high-throughput experimental technique that allows for the simultaneous testing of thousands of genetic variants to determine their impact on gene expression regulation. They performed this analysis in human neural progenitor cells, which allowed them to identify which variants have functional regulatory activity and to distinguish between those with pleiotropic effects and those with disorder-specific effects.
The researchers took almost 18,000 variations of the shared and unique genes involved and introduced them into precursor cells that become neurons, to examine their impact on gene expression during human development. This process identified 683 genetic variants that impacted gene regulation, which were then further explored in neurons from developing mice.
What Pleiotropic Variants Actually Do
The distinction between pleiotropic and disorder-specific variants proved to be more than taxonomic. Many of these shared variants remain active for extended periods, potentially influencing multiple developmental stages and offering new targets for treatments that could address several disorders at once.
“The proteins produced by these genes are also highly connected to other proteins,” explained University of North Carolina geneticist Hyejung Won. “Changes to these proteins in particular could ripple through the network, potentially causing widespread effects on the brain.”
Won noted that “pleiotropy was traditionally viewed as a challenge because it complicates the classification of psychiatric disorders.” The new research reframes that challenge as an opportunity: if shared variants are more functionally potent and more broadly active across development than disorder-specific ones, they may also be more tractable as therapeutic targets.
The Fourteen-Disorder Study: A Genetic Map at Scale
A second study, published in December 2025 in Nature and described by its authors as the most extensive cross-disorder psychiatric genetics analysis ever conducted, expanded the lens considerably. The study, led by researchers at University of Colorado Boulder, Harvard, and Mass General Brigham, examined DNA data from more than 1 million individuals diagnosed with at least one of 14 psychiatric disorders and 5 million with no diagnoses.
The work was led by the Psychiatric Genomics Consortium’s Cross-Disorder Working Group, co-chaired by Kenneth Kendler, M.D., a professor in the Department of Psychiatry at Virginia Commonwealth University’s School of Medicine, and Jordan Smoller, M.D., a professor in the Department of Psychiatry at Harvard Medical School.
Five Genetic Families of Mental Illness
The researchers found five genetic factors that explained the majority of genetic variance across the disorders: compulsive behaviors, schizophrenia and bipolar disorder, neurodevelopmental conditions, internalizing conditions such as depression and anxiety, and substance use conditions.
Using statistical methods that examined all 14 disorders simultaneously across the entire genome, researchers identified five major factors. Their analysis identified 428 genetic variants associated with more than one disorder, as well as 101 regions on chromosomes that were “hot spots” for these shared genetic variants.
Schizophrenia and Bipolar Disorder: Closer Than Anyone Thought
The close genetic overlap between schizophrenia and bipolar disorder was notable, the researchers said, given that these two disorders have historically been viewed as distinct. “Genetically, we saw that they are more similar than they are unique,” lead author Andrew Grotzinger, PhD, assistant professor of psychology and neuroscience at CU Boulder, said.
Patients with schizophrenia and patients with bipolar disorder are typically prescribed different medications, monitored through different clinical frameworks, and classified in different diagnostic categories in both the DSM-5 and the ICD-11. The genetic evidence now suggests those categorical walls may be overstated and possibly counterproductive to treatment planning.
Brain Development as the Common Thread
These genetic signatures were linked to hundreds of genetic regions as well as biological pathways, including some that affect brain cells involved in thinking and emotion. The findings suggest that shared genetic factors have a role early in brain development and could help to create a more biologically based way of understanding psychiatric conditions.
This developmental angle is consistent across both the Cell and Nature studies. The convergence suggests that a common biological window, the period during which neural circuits are being formed, may be the moment when genetic risk for multiple conditions is simultaneously activated or suppressed. Interventions targeting this window could have cross-disorder preventive or therapeutic effects.
Why People Are So Often Diagnosed With More Than One Condition
The data collected by both research teams helps explain one of the most clinically frustrating patterns in psychiatry: the frequency with which patients accumulate multiple diagnoses. Psychiatric conditions often occur together and share genetic influences, making diagnosis and treatment challenging, because diagnoses are currently based on symptoms rather than underlying biology.
Analysis of approximately one million people reveals that about half of Americans will meet the criteria for one or more psychiatric disorders at some point in their lives. If genetic architecture does not respect diagnostic borders, then co-occurring conditions may often be expressions of a single underlying biological vulnerability manifesting through different symptom clusters, rather than truly independent diseases that happen to coincide.
The Nature study’s framing of five genetic “families” addresses this directly. Rather than asking why so many people have two or three diagnoses, the framework suggests the more accurate question may be which genetic family or families a person belongs to, and what combination of environmental exposures then determines which specific symptom profile emerges.
What This Means for Treatment Development
The research group was “really interested in identifying groups of genes that influence multiple psychiatric disorders,” said Mary Marazita, Distinguished Professor of Oral and Craniofacial Sciences and professor of human genetics at the University of Pittsburgh. “It shows the promise of using these results to inform neurobiologically valid treatments and set targets for future development of treatments.”
The practical pathway runs in two directions. First, existing medications that target shared molecular mechanisms may already be more cross-diagnostically effective than current prescribing practices recognize. Second, new drug candidates designed specifically around the shared pleiotropic variants identified in the Cell study could theoretically be developed to address multiple conditions simultaneously, an approach that would represent a structural departure from the one-drug-per-diagnosis model that currently dominates psychiatric pharmacology.
Treating alcohol dependence without accounting for underlying anxiety, or addressing depression while ignoring the shared genetic terrain that also raises substance use risk, is an incomplete approach.
A More Biologically Grounded Diagnostic System
The broader implication of both studies is that the diagnostic architecture in psychiatry, built primarily around symptom observation rather than biological measurement, may need to be reconstructed with genetic and neurobiological data at its foundation. Psychiatric disorders display high levels of comorbidity and genetic overlap, challenging current diagnostic boundaries.
A symptom-based system was the best available tool for most of psychiatry’s history. Genomic data at population scale is now available to support a more biologically grounded alternative. The Nature study’s five-factor model is not a finished taxonomy, but it represents the most well-evidenced framework yet proposed for reorganizing how mental illness is classified.
The Diversity Gap in the Research
Both studies carry a significant limitation that their authors acknowledge directly. The vast majority of genetic data in these studies came from people of European-like genetic ancestry, and extending this research to more diverse populations is both a scientific and an ethical priority.
The current research mainly focuses on people of European ancestry. Future research should include more diverse populations and explore how these insights can guide new treatments.
Genetic risk architecture can vary across ancestral populations, and findings derived predominantly from one ancestry group may not generalize reliably to others. Any clinical applications built on this research, including diagnostic tools or treatment protocols premised on shared genetic variants, will need to be validated across a representative range of human genetic diversity before they can be applied equitably at a population level.
Key Takeaways
Two studies published in 2025, one in Cell and one in Nature, have together produced the most detailed and extensive evidence to date that psychiatric disorders share deep genetic roots. The Cell study, led by researchers at the University of North Carolina at Chapel Hill, used a massively parallel reporter assay to examine nearly 18,000 genetic variants across eight conditions, including autism, ADHD, schizophrenia, bipolar disorder, major depression, OCD, anorexia nervosa, and Tourette syndrome. They found that the variants shared across multiple disorders are functionally more active across brain development than disorder-specific variants, making them prime candidates for cross-disorder therapeutic targeting.
The Nature study, drawing on data from more than one million individuals with diagnoses across 14 conditions, reorganized the landscape entirely by identifying five broad genetic families into which those disorders cluster: compulsive behaviors, schizophrenia/bipolar, neurodevelopmental, internalizing, and substance use.
These findings do not render existing diagnoses meaningless. They do suggest that the categories psychiatry has relied upon for decades are, at minimum, biologically permeable. The clinical reality patients live with, in which multiple diagnoses accumulate and medications borrowed from one condition are routinely prescribed for another, reflects an underlying genetic reality that the field is only now acquiring the tools to properly describe. The gap between what is known genetically and what is practiced clinically remains wide. Closing it is a project that will unfold across years of additional research, drug development trials, and diagnostic revision.
What the Biology Has Always Been Trying to Say
Psychiatry built its diagnostic system from what clinicians could observe in the room: a patient who couldn’t sleep, couldn’t concentrate, heard voices, restricted food, washed their hands forty times before leaving the house. That framework helped millions of people get treatment. It also produced a strange artifact: the person who arrived with five separate diagnoses, a medication list that looked like a pharmaceutical filing cabinet, and a sneaking suspicion that nobody had quite figured out what was actually wrong with them.
These two studies suggest, collectively, that the biology was never as divided as the diagnostic manual implied. The genes involved in schizophrenia are not strangers to the genes involved in bipolar disorder. The pathways active in ADHD intersect with those implicated in depression. The shared variants identified in the Cell study are not merely present across conditions; they are disproportionately active during the developmental windows when the brain is being built, which means the risk is being shaped long before any symptom appears.
None of this means that depression and schizophrenia are the same thing, or that individual diagnoses have no value. A purely categorical system, one condition, one label, one drug, is an approximation, and not always a close one. The science is now detailed enough to start building something more accurate. The clinical infrastructure to use it is the next problem.
AI Disclaimer: This article was created with the assistance of AI tools and reviewed by a human editor.