There is no anxiety gene, only a long list of variants that nudge risk slightly.
For generations, anxiety has been understood primarily through the lens of experience and behavior, yet the biological substrate beneath it has remained largely unmapped. A landmark study drawing on genetic data from nearly 700,000 people has now identified 74 locations in the human genome associated with anxiety symptoms — 39 of them never previously linked to the condition — offering science its most detailed view yet of how anxiety is written, in small and distributed ways, into the human body itself. The work does not promise a cure, but it offers something that has long been absent: a biological map precise enough to guide the next generation of treatment toward something genuinely new.
- Anxiety disorders touch an estimated 359 million people worldwide, yet the field has gone decades without a fundamentally new treatment — most patients still rely on drugs conceived in the 1980s.
- Previous genetic studies were too small to hear the signal; this one, by doubling the dataset and measuring anxiety as a spectrum rather than a diagnosis, found 74 genomic locations where earlier efforts found roughly three dozen.
- Two genes — PCLO and SORCS3 — emerged as particularly strong candidates, both active at the neuronal synapse and already implicated in other psychiatric conditions, giving researchers concrete molecular footholds.
- A surprising discovery revealed genetic overlap between anxiety and heart disease, gut disorders, and migraine, offering a biological explanation for patterns clinicians have long observed but could not fully account for.
- Common genetic variation explains only about 6% of anxiety severity, meaning the genome is one piece of a much larger puzzle — but it is now a piece detailed enough to build drug targets from, and that changes what is possible.
For decades, anxiety has been treated as a problem of the mind — something to manage, talk through, or address with medications developed a generation ago. A new study suggests that framing has always been incomplete.
Researchers from King's College London and Australia's QIMR Berghofer Medical Research Institute analyzed genetic data from nearly 700,000 people and mapped 74 distinct locations in the human genome tied to anxiety symptoms — the largest collection ever identified in a single study, with 39 of those locations never previously connected to anxiety. The work, published in Nature Human Behaviour, doubled the size of the previous largest genetic study of the condition.
What made the difference was not just scale but method. Rather than sorting people into binary categories of anxious or not, the researchers measured anxiety symptom severity as a continuous spectrum. This kept in the analysis the vast majority of anxious people who have never received a formal diagnosis — which, given how rarely anxiety sufferers seek treatment, is most of them. The result was a far richer genetic signal.
Two genes stood out: PCLO and SORCS3, both involved in how neurons communicate at the synapse. The 74 regions collectively paint anxiety not as a single broken mechanism but as a distributed architecture of hundreds of small genetic nudges. The study also revealed meaningful genetic overlap between anxiety and heart disease, gut disorders, and migraine — offering a biological explanation for clinical patterns long observed but poorly understood.
The effect sizes remain modest. Common genetic variation accounts for roughly 6% of the variation in anxiety severity across the population, and no clinically useful individual risk test emerges from this data. Environment, trauma, and life history still do most of the work. But what the study does provide is a map — a list of biological coordinates that drug developers can now test against, replacing decades of trial-and-error pharmacology with something more targeted.
With 359 million people currently affected globally and projections exceeding 515 million by 2040, the need is not abstract. The path from genetic coordinates to new treatments will be slow — functional studies, animal models, early trials — but the coordinates themselves are what was missing. Anxiety now has a map nearly twice the size of anything that existed before.
For decades, anxiety has been treated as a problem of the mind—something to talk through, to manage with willpower, to medicate with drugs developed in the 1980s. A new study suggests that framing has always been incomplete. Researchers analyzing genetic data from nearly 700,000 people have mapped 74 distinct locations in the human genome tied to anxiety symptoms, the largest collection ever identified in a single study. Thirty-nine of those locations had never been connected to anxiety before.
The work, led by King's College London and Australia's QIMR Berghofer Medical Research Institute and published in June in Nature Human Behaviour, represents a doubling of the previous largest genetic study of anxiety. That scale matters enormously. Genetic effects on complex psychiatric traits are individually small—they don't announce themselves. Only with hundreds of thousands of people can researchers detect signals that would otherwise disappear into statistical noise. This study found so much more than earlier attempts not because the method was smarter, but because the dataset was finally large enough to hear what was always being whispered in the data.
What made this study different wasn't just size but approach. Instead of sorting people into two categories—anxiety disorder or no anxiety disorder—the researchers measured anxiety symptom severity as a continuous scale, capturing everything from ordinary stress to disabling clinical anxiety. That choice proved decisive. A strict diagnosis throws away most of the genetic signal, because it excludes everyone who is meaningfully anxious but never received a formal clinical label. Given how few people with anxiety ever seek treatment, that's the majority of anxious people. By treating anxiety as a dimension rather than a switch, the team kept that signal in the analysis. The payoff was direct: 74 genetic locations instead of the roughly three dozen found in earlier studies of similar or larger size.
Two genes emerged with particularly strong support: PCLO and SORCS3, both already implicated in other psychiatric conditions and both active in how neurons communicate at the synapse. Many of the 74 regions sit in or near genes expressed heavily in brain tissue, painting a picture of anxiety not as a single broken mechanism but as a distributed genetic architecture—hundreds of small effects stacked on top of each other. There is no anxiety gene, only a long list of variants that each nudge risk slightly, in combination with everything else happening in a person's life.
One finding pointed outside the brain entirely. The researchers discovered significant genetic overlap between anxiety and seemingly unrelated physical conditions: heart disease, gut disorders, and migraine. This overlap doesn't mean anxiety causes those conditions or vice versa. It means some of the same underlying genetic variation raises risk for all of them simultaneously. For years, clinicians have observed that anxious patients disproportionately present with irritable bowel syndrome, migraine, and cardiovascular symptoms without a clean biological explanation. This study offers one: shared genetic architecture.
But the effect size deserves precision. Common genetic variation accounts for roughly 6% of the variation in anxiety symptom severity across the population. Polygenic risk scores built from this data explain up to about 2.9% of that variance individually. That's meaningful for research—it's exactly the kind of signal that lets scientists build better biological models and test drug targets—but it is nowhere near enough to predict who will develop anxiety at an individual level. Environment, life history, trauma, and everything outside the genome are still doing most of the work. Nobody is getting a genetic test for anxiety risk that means anything clinically useful yet, and this study doesn't change that.
What it does offer is a map. Anxiety disorders affected an estimated 359 million people globally in 2021, about 4.4% of the world's population, and that number is projected to climb past 515 million by 2040. In many high-income countries, fewer than one in three people with an anxiety disorder receives treatment. The treatments that do exist—mostly SSRIs and structured therapy, both developed decades ago—don't work for everyone who tries them. There has been very little genuinely new pharmacology for anxiety in a generation. A genetic map this detailed gives drug developers actual biological targets to test against, instead of the trial-and-error approach that has largely defined anxiety treatment since benzodiazepines and later SSRIs. The genes implicated here point toward specific synaptic and neuronal pathways that can, in principle, be probed with new compounds—the same basic strategy that has produced real advances in other conditions once their genetic architecture became clear enough to act on.
The next phase will be slower and less headline-friendly than the discovery itself: functional studies to work out what these genes actually do in neurons, animal models to test candidate pathways, and eventually, if any of it holds up, early drug trials years down the line. Genetics moves from a list of coordinates to an actual treatment slowly. But a list of coordinates is exactly what was missing, and now anxiety has one nearly twice the size of anything that existed before.
Citações Notáveis
Despite the public health impact of anxiety, progress in the understanding of its genetics lags behind other major mental health conditions.— Professor Thalia Eley, King's College London, senior author
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter that they measured severity instead of just diagnosis?
Because most anxious people never get diagnosed. If you only count people with a formal diagnosis, you're throwing away the genetic signal from millions of people who are genuinely anxious but never saw a doctor. By treating it as a spectrum, they kept all that information in the analysis.
So they found 74 locations. What do those locations actually do?
Most of them sit in or near genes that are active in the brain, especially at the synapses where neurons talk to each other. Two genes—PCLO and SORCS3—came through strongest. But the real picture is that there's no single lever to pull. It's hundreds of small effects, each one nudging risk a little bit.
The study mentions genetic overlap with heart disease and migraines. That seems strange.
It's not that anxiety causes those things or vice versa. It means some of the same genetic variation raises risk for all of them at once. Doctors have been seeing this pattern for years—anxious patients also get migraines and gut problems—but they didn't have a biological explanation for it. Now they do.
Can they use this to predict who will get anxiety?
Not yet. Genetic variation only accounts for about 6% of the variation in anxiety severity. Everything else—trauma, life history, environment—is still doing most of the work. This is useful for research and drug development, but it's not a predictive test.
What happens next?
Years of slower work. Functional studies to figure out what these genes actually do in neurons, animal models, eventually drug trials. It's not glamorous, but it's how you go from a map to an actual treatment.
Why has anxiety genetics lagged so far behind other psychiatric conditions?
Partly just funding and attention. Depression and schizophrenia had large research consortia working on them for over a decade. Anxiety is the most common mental illness on the planet, but it got less resources. This study is trying to close that gap.