Anxiety

Anxiety disorders are among the most common mental-health diagnoses in autistic people (see Mental health for the wider picture). A meta-analysis of studies of autistic adults, 30 of which measured anxiety, estimated that about 27% had an anxiety disorder at the time of study (13 studies) and 42% at some point in life (8 studies). Most of the studies drew on clinics, the spread between them was very large, and the authors note that the use of questionnaires and the presence of intellectual disability may shift the estimates.¹ At the time of study, social anxiety disorder was the most common kind, at about 29%, followed by obsessive-compulsive disorder (OCD) at 24%, generalized anxiety disorder at 18%, and panic disorder or agoraphobia at 15%; the last two rested on four studies each, and the figure for specific phobias on only two. Each figure pools a different set of studies, which is why one kind can exceed the overall rate.¹ Rates did not differ significantly between adults with and without intellectual disability, though all nine studies of current social anxiety had left out people with intellectual disability.¹

Where people come from changes the number. A 2019 meta-analysis across all ages, counting only confirmed diagnoses, put anxiety disorders at about 20%, against about 7% in published general-population figures, a comparison the authors made informally after the main analysis. Studies of whole populations or registers gave about 15%, and studies of clinic or community samples about 26%; more recently published studies reported lower rates, and age, gender, intellectual functioning and country were also linked to the estimates. It counted OCD separately, at 9%.² In a Stockholm register, 20.1% of 4,049 autistic young adults aged 18 to 27 had an anxiety diagnosis, against 8.7% of others, and the increase was greater in those without intellectual disability. Their non-autistic brothers and sisters also had more anxiety diagnoses than the general population, but the autistic adults still had about 2.5 times the odds of their own siblings, so shared family factors did not explain the link.³

Children show a different pattern. A 2011 meta-analysis of 31 studies of 2,121 autistic children and adolescents found about 40% with at least one anxiety disorder or an impairing level of anxiety, in studies that used questionnaires as well as interviews; specific phobia was the most common, at about 30%, followed by OCD, social anxiety and generalized anxiety at 15% to 17% each.⁴ A 2022 meta-analysis limited to population-based studies of autistic children and adolescents found about 11%, a reminder of how much clinic samples raise the figures.⁵

Some anxiety in autistic people does not match the usual diagnostic categories. In a study of 59 autistic young people aged 7 to 17, 17% had only anxiety of the kinds the diagnostic manual describes, 15% only anxiety that did not fit those kinds, and 31% both.⁶ The anxiety that did not fit included worries about routine, novelty or special interests (22%), unusual specific fears not explained by general sound sensitivity, such as of a baby crying, coughing or balloons (12%), and fearfulness in social situations without concern about rejection (8.5%).⁶

The overlap cuts both ways. The authors of the 2019 meta-analysis note that rates may be raised when sensory problems and avoidance are read as generalized anxiety, and lowered when autism hides the signs of another condition, which is called diagnostic overshadowing; their example is depression.²

Intolerance of uncertainty, reacting badly to not knowing what will happen, goes closely with anxiety in autistic people. A 2020 meta-analysis of 10 studies, mostly of children and young people, found a large correlation between the two (r = 0.62), which the authors judged similar to that in non-autistic people.⁷ Routines and sameness covers this link and the programs aimed at it. In 176 autistic adults, statistical models placed intolerance of uncertainty between sensitivity to sensory input and anxiety, in a study that measured everything once.⁸

Sensory over-responsivity, strong reactions to ordinary sounds, touch or light, may come first. In 149 autistic toddlers measured a year apart, it predicted later anxiety, while anxiety did not predict later over-responsivity.⁹ Asked directly, 246 autistic adults rated their sensory over-reactivity more as a cause of their anxiety, and their sensory seeking more as an effect of it.¹⁰

Difficulty naming one's own feelings, called alexithymia, is another thread. In 179 autistic adolescents and adults, difficulty identifying feelings went with anxiety at the time, and in the 76 followed for one to two years it predicted later anxiety.¹¹

Social anxiety disorder is the most common anxiety diagnosis in autistic adults.¹ Some argue that part of it tracks a real threat, though the studies below measured victimization, not anxiety, so the link between the two is a reasonable idea rather than a finding. A 2023 meta-analysis of 34 studies estimated that about 44% of autistic people had experienced some form of victimization, and about 47% bullying; the studies mixed children and adults, and their estimates varied widely.¹³ In 111 autistic people the authors describe as high-functioning, everyday discrimination, internalized stigma and concealment went with poorer mental health even after allowing for general stress, which the authors read through the minority stress model.¹⁴

Camouflaging, hiding autistic traits in company, goes with anxiety, though less than often assumed. In 305 autistic adults without intellectual disability, most of them women, it was linked with generalized and social anxiety, but it added only about 4% to what age and autistic traits already explained, and the link did not differ between men and women. The study measured everything once.¹²

The common anxiety questionnaires hold up reasonably well in autistic adults without intellectual disability. The Hospital Anxiety and Depression Scale showed the same structure in 151 autistic older adolescents and young adults as in the general population,²⁰ and the GAD-7 appeared to measure anxiety in the same way in 477 autistic and 429 non-autistic college students aged 18 to 29.²¹ A 2026 systematic review of 27 studies of autistic adults without intellectual disability pooled the three questionnaires with enough studies and found them internally consistent, with an alpha of 0.91 for the GAD-7; internal consistency shows that the items hang together, not that they measure anxiety accurately.²³

The problems lie in particular items and uses. When questions about obsessive-compulsive symptoms were reworded more clearly for 50 autistic adults, their scores fell, which suggests the original items pick up autistic experiences such as routines; clarified social anxiety questions changed scores little. Non-autistic students showed a similar but usually smaller drop.²² A 2014 review of tools used in anxiety trials with autistic children found almost no evidence on whether they detect change during treatment.¹⁶ A 2026 review found that only about 8% of studies using anxiety or mood measures with autistic adults reported adapting them.²⁴

Autism-specific scales add items on uncertainty and the senses. The Anxiety Scale for Children with Autism Spectrum Disorder was developed in 170 children aged 8 to 16,¹⁷ and the Anxiety Scale for Autism-Adults was developed with autistic adults and tested in 551 of them.¹⁸ In 342 autistic and 316 non-autistic adults, the adult scale did not measure anxiety the same way in both groups, so its authors advise against using it to compare them directly; within the autistic group, women scored higher than men.¹⁹ Diagnosis can also go the other way: in 1,211 autistic adults on a Dutch research register, most of them highly educated, about a quarter reported an earlier psychiatric diagnosis they considered a misdiagnosis, and anxiety disorders were the second most common, after personality disorders. Women reported a perceived misdiagnosis more often than men, about 32% against 17%, largely because more women had an earlier psychiatric diagnosis; among those who had one, the rates were similar.¹⁵

Cognitive behavioral therapy (CBT) reduces anxiety in autistic children in the short term, by the ratings of clinicians and parents. A 2013 meta-analysis of 8 randomized trials with 469 children found large effects on clinician and parent ratings and a smaller one on the children's own reports; most comparison groups were waiting lists or usual care.²⁵ A 2021 meta-analysis of 19 trials with 833 young people found the same gradient, with effects largest on clinician ratings and smallest on self-report, and found no evidence that the benefits lasted at follow-up, though few trials measured it. It also found larger effects in younger children and in individual therapy, and concerns about reporting bias in most trials.²⁶ In a comparison across studies, effects were larger when parents took part.²⁷

Adapting the therapy for autism may add to it. In a trial of 167 autistic children aged 7 to 13, about one in five of them girls, a version of CBT adapted for autism reduced anxiety more than standard CBT, and both did better than usual care.²⁸

For adults the evidence is much thinner. A 2023 network meta-analysis of 71 randomized trials found no evidence of differences in final anxiety scores between any treatment and its comparison across the 13 adult studies, at very low certainty, though in direct comparisons some forms of CBT gave lower anxiety scores than no added treatment, at low certainty. Its authors recommend that autistic people have access to the same treatments as everyone else.²⁹ A 2025 network meta-analysis pooled 67 trials with 3,604 participants of all ages, mostly men without intellectual disability, and ranked mindfulness-based approaches highest for anxiety and CBT for depression. Its adult subgroup showed a larger anxiety effect than its child and adolescent subgroup, and its authors favor mindfulness for adult anxiety; it rated the anxiety comparisons low to moderate certainty. That grade covers its mixed-age network rather than adults alone, so the adult-specific rating stays at very low certainty.³⁹ Single trials are small. In 46 autistic adolescents and adults with OCD, CBT for OCD was not significantly better than anxiety management on OCD symptoms, and both groups improved.³⁰ In a pilot trial of 52 adults, group CBT adapted for autism did no better than a waiting list on the main anxiety measure.³¹ In a single-center trial of 121 adults, training in noticing body signals reduced anxiety more than an active comparison program at three months, by a small amount; 31% against 16% met the trial's criteria for recovery. Only 59% of the training group and 42% of the comparison group gave data at that point, so the main result rests on statistically filled-in data.³² Mental health covers mindfulness trials.

In England, the NICE guideline on autistic adults recommends offering the treatment advised for the specific anxiety disorder, adapted: a concrete, structured approach, written and visual information, more emphasis on changing behavior than thoughts, plain language without excess metaphor, regular breaks, and use of the person's special interests.³³

Medication for anxiety in autistic people rests on little direct evidence. A 2013 Cochrane review of 9 trials of SSRI antidepressants found no evidence of benefit in children and emerging evidence of harm, and only limited evidence in adults; the trials mostly measured core autism features and obsessive-compulsive behavior rather than anxiety, though one small adult study reported improvement in anxiety.³⁴ A 2021 meta-analysis of 15 trials of antidepressant and anti-anxiety medicines found no significant benefit in pooled data from 9 of them, on outcomes that measured obsessive-compulsive behavior and overall improvement.³⁵ A 2026 meta-analysis of 7 trials in children judged the effect of SSRIs on anxiety very uncertain and found slightly more side effects.³⁶

A UK and Australian trial, STRATA, is testing the SSRI sertraline against placebo for anxiety in autistic adults.³⁷ No study reporting its main results was found as of October 2026. In interviews with 62 of its participants, people in both the sertraline and placebo groups described their anxiety improving.³⁸

Many writers describe anxiety as the background of autistic life rather than an occasional state. Luke Beardon, an autism academic who writes for autistic adults, argues that there is "in the main a qualitative difference" between autistic anxiety and other people's, with more frequent, longer and more intense episodes, and suspects that many autistic adults have been anxious for so long that they do not know it.^c1 Orion Kelly, who lives with generalized anxiety disorder, writes: "I call it the caveman anxiety because every moment of my life I feel like a caveman on constant lookout for predators."^c3 Jacqueline, writing in their 50s in Alex Ratcliffe's collection of autistic adults' accounts, recalls being asked at 27 to remember a time free of fear or anxiety and finding none, and writes that being identified as autistic in their 40s made the ongoing anxiety easier to live with.^c4

Writers argue that the cause lies largely outside the person. Beardon sums up the view as a "golden equation", "Autism + Environment = Outcome", and writes that "it is the environment that is the biggest influencer over your anxiety – not you"; being autistic, Beardon stresses, "is not the same as being anxious".^c1 Nick Walker, who consulted on an MDMA trial for social anxiety in autistic adults, describes the anxiety it treated as "a symptom not of autism but of traumatic oppression" (see Psychedelics and cannabis).^c2 In a clinicians' guide, Allie S. recalls that a childhood therapist assumed social anxiety: "I had to explain that I didn't have serious social anxiety as a child, though I do now, and it was all sensory."^c5

Nonspeaking writers describe anxiety in the body, and what changed once they could communicate. Kaegan Smith, who began spelling on a letterboard at 18, writes: "When my body reacts to fear in ways that betray me, I use my communication method to conquer those fears."^c6 Justice Killebrew writes that anxious thoughts remained after he began to communicate by typing, and that a therapist helps him with exercises to manage them.^c7