Frank
Anxiety is significantly more common among autistic and ADHD adults than in the general population, with estimates ranging from roughly 20% to over 50% depending on study design. However, standard diagnostic tools were built and validated on neurotypical populations, raising genuine questions about whether these measurements accurately capture neurodivergent experience.
Several mechanisms are supported by research, including intolerance of uncertainty, alexithymia, sensory processing differences, and executive function challenges. A substantial body of evidence also points to environmental and social factors — particularly masking and minority stress — suggesting that much neurodivergent anxiety may reflect a rational response to navigating a world not designed for them, rather than an intrinsic neurological defect.
A deep dive into why anxiety shadows so many neurodivergent lives — and why the obvious answer is the wrong one.
Ask almost any autistic or ADHD adult whether they’ve struggled with anxiety, and you’ll usually get a tired half-laugh before they answer. The connection feels too obvious to need a study. Of course we’re anxious. The question that actually deserves a deep dive isn’t whether neurodivergent people carry more anxiety. It’s why — and that question turns out to be far stranger, more contested, and more important than the easy story we tell ourselves.
The easy story goes like this: neurodivergent brains are anxious brains. Faulty wiring, a jumpy amygdala, a nervous system set permanently to “high.” It’s tidy. It fits the medical model. And it’s almost certainly an oversimplification that, taken at face value, points us toward the wrong solutions.
This piece walks through what the research really shows: how big the gap actually is, why we may not even be measuring it correctly, what mechanisms are genuinely implicated, and why the most honest answer might be that a lot of “neurodivergent anxiety” isn’t a glitch in the brain at all — it’s a rational response to a world built for someone else.
Part One: How much more anxiety, really?
Start with the numbers, because they’re both striking and slippery.
For autism, the most cited synthesis is Hollocks and colleagues’ 2019 meta-analysis in Psychological Medicine, which pooled prevalence across studies and landed on roughly 27% of autistic adults meeting criteria for an anxiety disorder at any given moment, and about 42% over a lifetime. A separate large umbrella review by Lai and colleagues, published in The Lancet Psychiatry the same year, put the pooled figure closer to 20% — lower, but with a crucial caveat: estimates from clinical samples ran far higher than those from community or population-based samples. Other reviews describe autistic anxiety as running at least four to five times the general-population baseline.
For ADHD, the picture is similarly heavy. The National Comorbidity Survey Replication found that around 47% of adults with ADHD also met criteria for an anxiety disorder — nearly half. Some reviews put the figure between 47% and 53%, and broader estimates suggest that anywhere from two-thirds to ninety percent of adults with ADHD carry at least one psychiatric comorbidity, with anxiety among the most common.
Hold those against a general-population reference point. In a given year, somewhere around one in five adults experiences an anxiety disorder; generalized anxiety disorder specifically sits near five to six percent. So the elevation is real and large. Neurodivergent people are not imagining a pattern.
But notice how quickly the numbers start to wobble. Twenty percent or forty-two percent? Community sample or clinic sample? Self-report or diagnostic interview? Each of those choices moves the figure dramatically, and that instability is the first clue that we’re not dealing with a simple fact waiting to be counted. We’re dealing with something the measuring instrument itself struggles to hold still.
Two more honest caveats before going further. First, almost all of this data is about autism and ADHD, because that’s where the research money has gone. Dyslexia, dyspraxia, dyscalculia, and the rest are far less studied, and confident claims about “neurodivergence” as a single category quietly smuggle in the assumption that what’s true for an autistic person is true for a dyslexic one. Often it isn’t. Second, most of these samples are drawn from people who sought a diagnosis or treatment in the first place — which means the calm, thriving, low-anxiety neurodivergent adult who never walked into a clinic is largely invisible in the data. The averages we cite are almost certainly skewed upward by who shows up to be counted.
Part Two: The problem nobody wants to talk about — are we even measuring anxiety?
Here’s the uncomfortable part, and the place where this stops being a comfortable blog post and starts being a real one.
The questionnaires clinicians use to measure anxiety — the GAD-7, the State-Trait Anxiety Inventory, the standard structured interviews — were built and validated on neurotypical people. Researchers in this field now openly acknowledge that these tools haven’t been properly validated for autistic populations, and that this is a live limitation hanging over the entire literature.
Why does that matter? Because anxiety in a neurodivergent person doesn’t always look like anxiety in the textbook. A landmark study by Kerns and colleagues found that up to 46% of autistic children showed anxiety presentations that didn’t fit neatly into standard diagnostic categories at all — fear of specific changes, unusual phobias, distress tied to sensory experience rather than to social evaluation. If your ruler was designed to measure a different shape, what exactly are your measurements capturing?
The deeper danger is conflation. Several things that get scored as “anxiety” on a standard scale might be something else entirely:
- An autistic person’s restlessness or need to leave a loud room could be sensory overload, not anxious anticipation.
- A reported sense of dread might be interoceptive confusion — a body sending unclear signals that the person can’t easily label (more on this below).
- An ADHD adult’s “social anxiety” might partly be working-memory overload: the genuine cognitive cost of tracking a fast group conversation, mislabeled as fear.
- Repetitive behavior that looks like a symptom might actually be the coping strategy that’s reducing the distress.
If a measure is partly capturing core neurodivergent traits and scoring them as pathology, then every correlation built on top of it inherits the error. You could end up confidently publishing that “trait X causes anxiety” when what you’ve really shown is that “trait X resembles the thing my anxiety scale was built to detect.” That’s not a minor footnote. It’s a crack running through the foundation, and a serious researcher named it as the single most likely way this whole line of inquiry goes wrong.
None of this means neurodivergent anxiety isn’t real. It means the construct is murkier than the clean prevalence statistics suggest, and humility is warranted before we start naming causes.
Part Three: The mechanisms that actually have evidence behind them
With that humility in place, the literature does point to several genuine candidate pathways. The important thing is that they tangle together, vary from person to person, and rarely act alone.
Intolerance of uncertainty
If there’s a single thread that runs deepest through the research on autistic anxiety, it’s intolerance of uncertainty (IU) — the tendency to find not-knowing genuinely distressing, to need predictability, and to struggle to act when an outcome is ambiguous. IU is what researchers call a transdiagnostic mechanism: it drives anxiety in the general population too. But it appears notably elevated in autistic people, and a well-supported model from South and Rodgers proposes it as a key link between sensory differences and anxiety.
The proposed chain is elegant and plausible: if your sensory system delivers the world unpredictably — if a sound that’s fine on Tuesday is unbearable on Wednesday — then the world becomes a place where you can’t reliably forecast your own experience. Chronic unpredictability breeds hypervigilance. Hypervigilance is exhausting. And a brain that can’t trust its forecasts stays braced for impact. Notice that in this model, the “anxiety” is downstream — a reasonable adaptation to genuine unpredictability, not a free-floating defect.
Alexithymia and the unreliable inner narrator
Layered onto IU is alexithymia: difficulty identifying and describing one’s own emotions, which is far more common in autistic people than in the general population. It’s often rooted in differences in interoception — the sense of what’s happening inside your own body, the signals that tell you you’re hungry, tired, or afraid.
Think about what anxiety actually is at the body level: a racing heart, a tight chest, shallow breath. Now imagine those signals arriving without clear labels. Research suggests that naming a feeling (“I am anxious”) is itself a regulating act — it measurably calms the brain’s alarm system. So a person who can’t easily read or name their internal state is denied one of the most basic tools for managing it. Worse, several studies trace a serial pathway — autistic traits feed alexithymia, alexithymia feeds intolerance of uncertainty, and that uncertainty feeds anxiety. The strategy that would help (labeling the feeling) is precisely the one that’s hardest to access. That’s not a character flaw. It’s a cruel structural bind.
Sensory processing and executive load
Sensory sensitivity shows up repeatedly, but the evidence increasingly suggests the anxiety often comes from anticipating overload rather than from the sensory input itself — the dread of the open-plan office, not just the office. Some autistic people even describe certain intense sensory input as anxiety-reducing, which breaks any simple “sensation in, anxiety out” model.
On the ADHD side, executive function differences — working-memory limits, planning difficulty, time blindness — create a steady drip of real-world stress: missed deadlines, forgotten commitments, the chronic low hum of being behind. Whether that causes anxiety or simply travels alongside it (both potentially downstream of the same underlying neurology, or of years of accumulated failure feedback) is genuinely unresolved.
Social and rejection mechanisms
Finally, social pathways. Many ADHD adults describe rejection sensitive dysphoria — an intense pain response to perceived rejection. It’s widely reported in the community but lacks formal diagnostic criteria, and it’s an open question whether it’s a distinct phenomenon or a vivid description of something already understood. In autism, elevated social anxiety may reflect realistic appraisal as much as distortion: if you’ve actually been misread, excluded, or punished for being different, anticipating it again isn’t irrational. It’s pattern recognition.
The honest summary: these mechanisms are real, partially overlapping, differently weighted in different people, and impossible to cleanly separate. Anyone selling you a single master cause is selling you something.
Part Four: The reframe — what if it’s not the brain, but the fit?
Here’s where the most interesting and least comfortable idea enters, and it reorganizes everything above.
What if a large share of “neurodivergent anxiety” isn’t generated by the neurology at all, but by the friction between that neurology and an environment that doesn’t accommodate it? Not a faulty alarm system — an alarm system responding correctly to a genuinely hostile environment.
Two lines of evidence give this real weight.
Masking. Many neurodivergent people spend their days camouflaging — consciously suppressing natural behaviors, scripting conversations, performing a neurotypical version of themselves to fit in or stay safe. A growing body of research, including a 2024 meta-analysis, links camouflaging to higher anxiety, depression, and burnout, and to lower wellbeing. The proposed mechanism is intuitive: masking is sustained, exhausting cognitive labor, and the long-term cost is depletion, a corroded sense of identity, and chronic stress. (A crucial caveat the careful studies insist on: the association is consistent, but the causal direction isn’t nailed down, and some people report upsides to masking too, like keeping a job. The picture is real but not simple.)
The radical implication is that masking effort — not trait severity — might be a primary driver. Two people with identical neurology could differ enormously in anxiety depending on how much suppression their environment demands of them. If that’s right, then a lot of what we’ve been measuring as “how autistic someone is” might really be “how hard they’re working to hide it.”
Minority stress. Researchers Botha and Frost extended the minority stress model — originally developed to explain why marginalized groups carry elevated mental-health burdens — to the autistic population. The framework distinguishes distal stressors (real, external discrimination, exclusion, bullying) from proximal ones (internalized stigma, the constant anticipation of rejection, hiding who you are). Chronic exposure to this kind of stress reliably produces anxiety and depression in any group it touches. There’s no reason neurodivergent people would be exempt, and considerable reason to think they’re heavily exposed.
Put masking and minority stress together and a different explanation emerges. Maybe neurodivergent people aren’t anxious because their brains are broken. Maybe many are anxious because they spend their lives translating themselves for a world that penalizes the untranslated version — and that’s exhausting, isolating, and frightening in ways that would make almost anyone anxious. The anxiety, in this frame, is less a symptom of the person and more a readout of the mismatch between person and environment.
This connects to a broader idea sometimes called the double empathy problem: the communication breakdown between neurotypical and neurodivergent people isn’t a one-way deficit in the neurodivergent person, but a two-way gap. If the misunderstanding runs both directions, then so does the responsibility for fixing it — and “fix the anxious person” stops being the obvious intervention.
Part Five: Why we can’t just settle this
It would be satisfying to end with a verdict. The science won’t let us, for reasons worth naming plainly:
Causation runs in circles. In neurodevelopment, the arrows loop. Anxiety leads to avoidance, avoidance produces skill gaps and missed experiences, those produce more rejection, which produces more anxiety. A snapshot study can’t tell you where the loop started, and even long-term studies struggle, because the loop has usually been turning since childhood.
The samples are skewed. As noted, most data comes from people who reached a clinic — disproportionately those whose distress brought them there. That inflates the apparent link between being neurodivergent and being anxious.
The heterogeneity is brutal. “Autism” and “ADHD” each contain enormous internal variety, they frequently co-occur, and within any one person the path to anxiety might run through sensory channels, social ones, executive ones, or trauma. A finding true for 45% of a sample can get written up as the mechanism and then fail to replicate everywhere else.
Biology and environment won’t separate. The same person can show high anxiety in a hostile setting and almost none in an accommodating one. Measure them only in the hostile setting and you’ll attribute the environment’s effect to their neurology — and that error has teeth.
That last point is why this isn’t merely an academic puzzle. If we wrongly conclude that neurodivergent anxiety is fixed and internal, that conclusion can be weaponized: to deny a student accommodations (“the anxiety is just how they’re wired, so changing the classroom won’t help”), to justify trying to train away protective behaviors, or to reinforce the idea that the neurodivergent person is the problem to be solved. Confident-but-wrong is more dangerous here than openly uncertain.
So what do we actually do with this?
Refusing a single tidy cause is not the same as having nothing useful to say. A few things follow with reasonable confidence.
Match the help to the mechanism. Generic anxiety treatment — and standard protocols do appear less effective in autistic people — wastes time when the real driver is sensory, or executive, or the sheer cost of masking. Anxiety rooted in uncertainty responds to predictability and structure, not to being argued out of “irrational” worry. Anxiety rooted in interoceptive confusion may respond to building body-awareness rather than challenging thoughts. Anxiety rooted in executive overload needs scaffolding, not reassurance. The point is to ask which anxiety before reaching for the default.
Change environments, not just people. If a meaningful share of the anxiety is mismatch and masking, then redesigning the environment — sensory-considered spaces, predictable structures, places where a person doesn’t have to perform a costume of normalcy — is a legitimate and underused intervention. And accommodations built for neurodivergent people tend to help everyone: clearer expectations and calmer spaces are good for the whole room.
Hold the science loosely. Treat prevalence numbers as ranges, not facts. Treat single-mechanism claims with suspicion. And resist the pull of the deficit story, not because it’s politically inconvenient, but because the evidence genuinely doesn’t support its certainty.
The honest answer to “why do neurodivergent people have so many anxiety issues” is layered, and the layers matter. Yes, there are real neurological differences — in how uncertainty is tolerated, how the body’s signals are read, how sensory input lands. And also, woven all the way through, there is a world that asks a great many people to spend their days suppressing themselves, anticipating rejection, and bracing for an environment that wasn’t built with them in mind.
The anxious-brain story isn’t wrong so much as it’s incomplete in a way that quietly blames the person. The fuller story is harder to hold, because it implicates all of us in the conditions that produce the worry. But it’s also, in the end, more hopeful — because mismatches can be redesigned, environments can be changed, and a person who’s allowed to stop translating themselves might finally get to find out how much of the anxiety was ever theirs to carry alone.
A note on this piece: this is a synthesis of published research, written for general readers. The science here is genuinely unsettled, the prevalence figures vary widely by method, and none of this is a substitute for individual clinical guidance. If you recognize yourself in any of it, that recognition is worth taking seriously — and worth taking to someone qualified to help you with your specific situation.
References and further reading
- Botha, M., & Frost, D. M. (2020). Extending the Minority Stress Model to Understand Mental Health Problems Experienced by the Autistic Population. Society and Mental Health, 10(1), 20–34.
- Hollocks, M. J., Lerh, J. W., Magiati, I., Meiser-Stedman, R., & Brugha, T. S. (2019). Anxiety and depression in adults with autism spectrum disorder: a systematic review and meta-analysis. Psychological Medicine, 49(4), 559–572.
- Hull, L., Petrides, K. V., & Mandy, W. (2021). Is social camouflaging associated with anxiety and depression in autistic adults? Molecular Autism, 12, 13.
- Kerns, C. M., Kendall, P. C., Berry, L., et al. (2014). Traditional and atypical presentations of anxiety in youth with autism spectrum disorder. Journal of Autism and Developmental Disorders, 44(11), 2851–2861.
- Kessler, R. C., Adler, L., Barkley, R., et al. (2006). The prevalence and correlates of adult ADHD in the United States: results from the National Comorbidity Survey Replication. American Journal of Psychiatry, 163(4), 716–723.
- Lai, M.-C., Kassee, C., Besney, R., et al. (2019). Prevalence of co-occurring mental health diagnoses in the autism population: a systematic review and meta-analysis. The Lancet Psychiatry, 6(10), 819–829.
- Maisel, M. E., Stephenson, K. G., South, M., Rodgers, J., Freeston, M. H., & Gaigg, S. B. (2016). Modeling the cognitive mechanisms linking autism symptoms and anxiety in adults. Journal of Abnormal Psychology, 125(5), 692–703.
- Milton, D. E. M. (2012). On the ontological status of autism: the “double empathy problem.” Disability & Society, 27(6), 883–887.
- Moore, H. L., Brice, S., Powell, L., et al. (2021). The mediating effects of alexithymia, intolerance of uncertainty, and anxiety on the relationship between sensory processing differences and restricted and repetitive behaviours in autistic adults. Journal of Autism and Developmental Disorders.
- South, M., & Rodgers, J. (2017). Sensory, emotional and cognitive contributions to anxiety in autism spectrum disorders. Frontiers in Human Neuroscience, 11, 20.
- Systematic review and meta-analysis of mental health outcomes associated with camouflaging in autistic people (2024). Clinical Psychology Review.
Prevalence figures throughout are drawn from clinical and community samples and vary substantially by study design, diagnostic criteria, and population. No figure should be read as a universal fact about all neurodivergent people.
