Two psychiatrists examining the same patient in different clinics sometimes arrive at different diagnoses. One sees schizophrenia; the other sees bipolar disorder with psychotic features. For decades, clinicians assumed those divergent calls reflected disagreement about symptoms. A landmark study published at the end of 2025 suggests the disagreement may run far deeper – all the way to the genome.
The research, led by the international PGC Cross-Disorder Working Group, examined DNA data from more than 1 million individuals diagnosed with at least one of 14 psychiatric disorders and 5 million individuals with no diagnoses. The scale alone is without precedent in psychiatry. But the scale isn’t what’s generating the most discussion. The finding that is reshaping how researchers think about mental illness is far more precise: “Genetically, we saw that they are more similar than they are unique,” said corresponding author Andrew Grotzinger, assistant professor of psychology and neuroscience at CU Boulder.
That assessment was directed at schizophrenia and bipolar disorder specifically – two conditions that have been treated as categorically distinct for most of psychiatric history, each with separate diagnostic codes, separate medications, and separate clinical pathways. The genetic data now tells a different story. For disorders for which diagnostic separation has been most debated, such as schizophrenia and bipolar disorder, genomic methods have revealed that the majority of genetic signal is shared. The magnitude of that overlap, and what it could mean for how we diagnose and treat mental illness, is what this study forces psychiatry to confront.
The Architecture of the Study
The most recent effort from the PGC’s Cross-Disorder Working Group, published in a 2025 study in Nature, described the convergent and divergent risk pathways across 14 psychiatric disorders. The 14 disorders studied spanned both childhood- and adult-onset conditions, including schizophrenia, bipolar disorder, and major depressive disorder.
The work was co-chaired by Kenneth Kendler, M.D., a professor and eminent scholar at Virginia Commonwealth University’s Institute for Psychiatric and Behavioral Genetics, and Jordan Smoller, M.D., professor of psychiatry at Harvard Medical School and director of the Center for Precision Psychiatry at Massachusetts General Hospital.
The analytical approach was designed to do something previous studies couldn’t: rather than compare disorders against each other pairwise, the team modeled the entire genetic architecture across all 14 conditions simultaneously. This allowed researchers to identify patterns that individual pairwise comparisons would miss – specifically, whether groups of disorders cluster together around shared genetic signals. The team identified five “genomic factors” that explained the majority of the genetic signal in the disorders and connected different subsets of conditions.
They found that five underlying “genomic factors” involving 238 genetic variants made up the majority of the genetic differences between those with a particular disorder and those without it. In other words, out of the thousands of genetic variants in the human genome, 238 carry most of the biological risk for the full spectrum of psychiatric illness examined.
Five Genetic Families, Not Fourteen Separate Disorders
The five factors reflected: disorders with compulsive features such as anorexia nervosa, Tourette disorder, and OCD; schizophrenia and bipolar disorder; neurodevelopmental conditions, including autism and ADHD; internalizing conditions, including depression, anxiety, and PTSD; and substance use conditions.
This reorganization is clinically significant. For anyone who has wondered why a diagnosis of depression so often comes packaged with anxiety – or why ADHD frequently co-occurs with autism – the answer, according to this research, is partly genetic. Instead of existing in isolation, these conditions fall into five overlapping families, helping explain why depression, anxiety, PTSD, bipolar disorder, schizophrenia, and substance use disorders so frequently occur together.
For most people diagnosed with a psychiatric disorder, that diagnosis is rarely the only one they will receive. Many go on to develop a second or even third condition, which complicates how mental illnesses are classified and treated. The five-factor model provides a genetic explanation for that pattern – co-occurrence isn’t random misfortune or diagnostic error. Much of it is the same biological signal manifesting across multiple phenotypes.
The Schizophrenia-Bipolar Connection
The most arresting single finding in the schizophrenia and bipolar genetic analysis concerns how much these two conditions share at the genomic level. Grotzinger noted: “Right now, we diagnose psychiatric disorders based on what we see in the room, and many people will be diagnosed with multiple disorders. That can be hard to treat and disheartening for patients. This work provides the best evidence yet that there may be things that we are currently giving different names to that are actually driven by the same biological processes.”
The specific number attached to that claim is striking on its own: schizophrenia and bipolar disorder showed substantial overlap, sharing roughly 66-70% of their genetic markers. To put that in context, the two disorders are currently treated as separate categories in the Diagnostic and Statistical Manual of Mental Disorders (DSM), the primary diagnostic framework used by clinicians across North America. A patient presenting with psychotic symptoms gets assessed for schizophrenia. A patient with mood cycling gets assessed for bipolar disorder. The treatments that follow – antipsychotics for the former, mood stabilizers for the latter – are often quite different, even though the genetic architecture driving much of the risk is largely the same.
Grotzinger said it is too early to begin combining diagnoses based on the findings, but as researchers work to update the Diagnostic and Statistical Manual of Mental Disorders, he hopes the new study will be considered.
What Excitatory Neurons Have to Do With It
The study didn’t just identify statistical overlap – it traced that overlap to specific biology in the brain. For the schizophrenia and bipolar disorder factor, the implicated pathways include genes expressed in excitatory neurons, while genes expressed in brain cells called oligodendrocytes were more prominent for the internalizing factor.
Excitatory neurons are the brain cells responsible for transmitting signals that activate other neurons – they’re the “accelerators” of neural communication. The fact that variants linked to both schizophrenia and bipolar disorder are concentrated in genes expressed by these cells suggests that both conditions involve disruptions to the same neural signaling infrastructure, even if the resulting symptoms look different on the surface.
The researchers also found that disorders with shared genetic risk often followed similar biological patterns. For the internalizing cluster – which includes major depression, anxiety, and PTSD – oligodendrocytes (the cells that wrap nerve fibers in a protective coating called myelin, allowing electrical signals to travel efficiently) were the relevant cell type. This distinction suggests that different genetic families not only group different disorders together, but also implicate different brain systems in their pathology.
Depression, Anxiety, and PTSD: 90% Shared Risk
The schizophrenia-bipolar finding draws the most headlines, but the internalizing cluster contains a number equally hard to dismiss. To better understand those genetic influences, researchers analyzed data from more than 6 million people. Among the patterns that emerged: major depression, anxiety, and PTSD shared approximately 90% of their genetic risk – conditions that fall into the same overlapping family, helping explain why they so frequently occur together.
For clinicians, that figure challenges the way internalizing disorders are currently treated as distinct entities. A person receiving a diagnosis of generalized anxiety disorder and then developing major depression years later is not experiencing two separate biological events. Genetically, they may have had a single elevated risk all along, expressed first as one condition and later as another, shaped by life experience, stress, and circumstance.
The gut-brain axis is one of several biological systems researchers are examining for how environmental factors may interact with this shared genetic risk – turning elevated susceptibility into active disorder.
Why Multiple Diagnoses Are the Rule, Not the Exception
The current DSM-based system classifies disorders by their observable symptoms: mood, behavior, cognition, perception. The genetic analysis suggests that classification by symptom may not align with the underlying biology. Two patients with very different symptom profiles could share substantial genetic risk. Two patients who look clinically similar could, in fact, have quite different genetic drivers. A diagnostic system built purely on what the clinician observes in the room will sometimes group together what biology separates, and separate what biology groups.
The five-factor structure also sheds light on a clinical pattern that has long puzzled patients and their doctors: why a single psychiatric diagnosis so rarely stays singular. For most people diagnosed with one condition, a second or third diagnosis follows at some point, not because their health is deteriorating in isolation, but because the same underlying genetic signals are broad enough to express across multiple phenotypes. The five-factor model gives that pattern a biological name.
The findings also provide key insight into the biological pathways and gene expression in brain cell types that may underlie certain conditions, said co-corresponding author Jordan Smoller, director of the Psychiatric and Neurodevelopmental Genetics Unit in the Massachusetts General Hospital Center for Genomic Medicine and director of the Center for Precision Psychiatry in the Mass General Brigham Department of Psychiatry.
What This Could Mean for Treatment
The treatment implications carry the most immediate practical weight. Currently, a patient diagnosed with both depression and generalized anxiety disorder may be prescribed an antidepressant, a separate anti-anxiety medication, and referred to cognitive behavioral therapy for each condition. A patient with comorbid ADHD and autism might see a developmental pediatrician, a psychiatrist, and a behavioral specialist – each treating one piece of the clinical picture.
Grotzinger stated: “By identifying what is shared across these disorders, we can hopefully come up with ways to target them in a different way that doesn’t require four separate pills or four separate psychotherapy interventions.”
That isn’t a near-term clinical promise – the researchers are clear about that. Translating a genetic finding into a new drug or a revised treatment protocol takes years of further work. But the directional logic is sound. If five biological pathways account for most of the genetic variance across 14 disorders, then treatments designed to target those pathways could, in principle, address multiple conditions simultaneously rather than one at a time.
According to Jordan Smoller, “These findings provide valuable clues for advancing our understanding and treatment of mental illness with greater precision.”
The research also has implications for drug repurposing. Medications already approved for one psychiatric condition could, in principle, be systematically evaluated against the shared genetic pathways rather than tested condition by condition. The five-factor model gives researchers a more organized biological target than the current symptom-category system provides.
Key Takeaways
The 2025 Nature study from the PGC represents the most comprehensive genetic mapping of psychiatric illness ever conducted. Its core finding – that most psychiatric risk concentrates into five genetic groupings rather than fourteen distinct biological categories – challenges a diagnostic system that has been largely symptom-based since the DSM was first published. The findings could inform new, more precise ways to diagnose mental illness and develop therapies to treat more than one condition at once.
For patients, the near-term practical takeaway is this: if you carry one psychiatric diagnosis, your risk of developing a second or third is not a sign of worsening health or clinical failure. It reflects the structure of shared genetic risk the study has now documented across millions of people. Discussing your full psychiatric history – including family history of any mental health condition, not just the specific ones you’ve been diagnosed with – gives clinicians the clearest picture of your biological risk profile. For researchers and policymakers, the more important message is that the five-factor model identified by the PGC’s Cross-Disorder Working Group provides a genuinely new framework for organizing psychiatric genetics research and, eventually, treatment development. The goal isn’t to collapse all mental illness into five categories – the clinical differences between schizophrenia and bipolar disorder remain real and clinically meaningful. The goal is to ensure that the biological architecture underneath those clinical distinctions informs how we develop the next generation of interventions.
Disclaimer: The information provided here is for educational and informational purposes only and is not a substitute for professional psychological, psychiatric, or mental health advice, diagnosis, or treatment. Always seek the guidance of a licensed mental health professional, therapist, psychologist, or psychiatrist with any questions or concerns about your emotional well-being or mental health conditions. Never ignore professional advice or delay seeking support because of something you have read here.
AI Disclaimer: This article was created with the assistance of AI tools and reviewed by a human editor.
Read More: Why People With Schizophrenia Hear Angry Voices in the U.S.—But Kinder Ones in India and Africa