Anxiety is still more often treated as a mood problem than a biology problem. A new study makes that harder to justify.
Researchers analyzing genetic data from nearly 700,000 people have identified 74 distinct locations in the human genome linked to anxiety symptoms — the largest number ever found in a single study — and 39 of those had never been connected to the condition before. It is the most detailed genetic map anxiety has ever had.
The study, led by researchers at King’s College London and Australia’s QIMR Berghofer Medical Research Institute, was published this June in Nature Human Behaviour. It drew on data from 693,869 people of European ancestry, pooled across dozens of cohorts, making it roughly double the size of the previous largest anxiety GWAS. Scale matters enormously in this kind of research — genetic effects on complex psychiatric traits tend to be small individually, and it takes very large sample sizes before those effects rise above statistical noise.
That’s the main reason this study found so much more than earlier attempts: it wasn’t a smarter method applied to old data, it was a much bigger dataset finally able to detect signals that were always there.
Why severity, not diagnosis
The study is a genome-wide association study, or GWAS — the standard tool for scanning entire genomes across huge populations to find spots where genetic variation lines up with a trait. What sets this one apart methodologically is the choice to measure anxiety symptom severity rather than a binary diagnosis. Instead of sorting participants into “has anxiety disorder” and “does not,” the team scored how severe anxiety symptoms were across the full population, capturing everything from ordinary stress responses to disabling clinical anxiety on one continuous scale.
That distinction matters more than it sounds. A strict diagnostic category throws away most of the genetic signal in a population, because it discards everyone who is meaningfully anxious but never received a formal clinical diagnosis — which, given how few people with anxiety ever seek treatment, is the majority of anxious people. By treating anxiety as a dimension rather than a switch, the researchers kept that signal in the analysis, and the payoff shows up directly in the numbers: 74 loci instead of the roughly three dozen found in earlier, diagnosis-based studies of similar or larger size.
“Despite the public health impact of anxiety, progress in the understanding of its genetics lags behind other major mental health conditions.”
That’s Professor Thalia Eley of King’s College London, the study’s senior author, framing why this gap needed closing. Anxiety disorders are the most common category of mental illness on the planet, yet — until now — they have had a noticeably thinner genetic literature than conditions like depression or schizophrenia, which have each had their own large, well-funded consortia mapping their genomes for over a decade. This study is, in part, an attempt to bring anxiety’s genetics up to the same resolution.
What turned up in the genome
Two genes came through with particularly strong support: PCLO and SORCS3, both already implicated in other psychiatric conditions and both active in the way neurons communicate with each other at the synapse. Many of the 74 regions sit in or near genes expressed heavily in brain tissue, reinforcing a picture where anxiety isn’t a single broken gene but a wide, distributed genetic architecture — hundreds of small effects stacked on top of each other rather than one lever anyone could flip. That’s consistent with what’s been found for nearly every major psychiatric trait studied at this scale: there is no “anxiety gene,” only a long list of variants that each nudge risk slightly, in combination with everything else going on in a person’s life.
One of the more striking findings sits outside the brain entirely. The researchers found significant genetic overlap between anxiety and a set of seemingly unrelated physical conditions — heart disease, gut disorders, and migraine among them. That overlap doesn’t mean anxiety causes those conditions or vice versa. It means some of the same underlying genetic variation raises risk for anxiety and for these physical conditions simultaneously, which helps explain a pattern clinicians have observed for years without a clean biological explanation: anxious patients disproportionately also present with irritable bowel syndrome, migraine, and cardiovascular symptoms. This study gives that clinical observation an actual genetic mechanism to investigate.
What it does and doesn’t mean
It’s worth being precise about the size of the effect here, because genetics headlines routinely overstate this kind of result. Common genetic variation accounts for roughly 6% of the variation in anxiety symptom severity across the population, and 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 eventually test drug targets — but it is nowhere near enough to predict, at an individual level, who will develop anxiety. Environment, life history, trauma, and everything else 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 this data does offer is a map, and the scale of the problem it’s mapping is enormous. 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 as the global burden of anxiety keeps rising faster than population growth alone would explain. Despite that scale, in many high-income countries, fewer than one in three people with an anxiety disorder receives treatment — and in lower-income settings the proportion is far smaller, and 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 the introduction of 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. That’s the real story in this study: not that anxiety has been “explained,” which it very much has not, but that it now has 74 more addresses researchers can go knock on, several of which point toward biology nobody had previously connected to anxiety at all.
The next phase of this research, as with most large GWAS findings, 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 it.
