Psychosis and bipolar
Psychotic and bipolar disorders are recorded far more often in autistic adults than in other adults. The size of the link is uncertain because the measures overlap and because autistic speech and reactions can be read as psychosis; this entry covers the rates, bipolar's distinct picture, how the two are told apart, and what happens after a diagnosis.
Autistic adults are told they have psychosis or bipolar disorder much more often than other adults. Doctors do not always know when autism just looks like these conditions. This entry explains the numbers, how the two are told apart, and what happens after a diagnosis.
How often the two occur together Established
Psychotic and bipolar disorders are diagnosed more often in autistic people than in other people. A meta-analysis of 53 studies, restricted to adults, pooled a rate of about 9% for psychosis (N = 63,657) and about 7.5% for bipolar disorder (N = 31,739), both far above general-population rates. It found males more often given a psychosis diagnosis and females more often given a bipolar one, but cautions that research on risk factors in autism is thin, so these sex patterns rest on few studies.1
An updated meta-analysis of adults reached the same conclusion. Pooling 63 papers covering about 6.9 million people, it found a strong association between autistic traits and psychosis-spectrum traits and put the odds of one diagnosed condition given the other at about seven times.2
A meta-analysis of 96 studies that counted only confirmed diagnoses, across all ages, put bipolar disorder at about 5% and a schizophrenia-spectrum disorder at about 4% of autistic people; clinic samples gave higher figures than population samples.3 In average-IQ adults, a smaller meta-analysis pooled a schizophrenia-spectrum rate close to 6%.4
Whole-population data agree on the direction. In Manitoba, Canada, where autism is identified from health and education records rather than by referral, autistic people were about ten times as likely as other Manitobans to have a recorded schizophrenia diagnosis (age- and sex-adjusted relative risk 9.8) and about twice as likely to have a mood or anxiety disorder (2.0). Those two counts cover ages 13 to 45, where the diagnoses were measured, within a cohort that spans ages 6 to 45. In the models, male sex was linked to a higher risk of a schizophrenia diagnosis (aRR 1.83), and among autistic people the five-year rate of a recorded schizophrenia diagnosis was 7.6% in males and 5.5% in females. The recorded rate is higher than the pooled figures above, and the authors caution that clinical recognition and diagnostic practice shape what gets recorded.5
Doctors find psychosis and bipolar disorder in autistic people more than in other people. One review looked at 53 studies of adults. It found psychosis in about 9 in 100 autistic adults. It found bipolar disorder in about 7 in 100. Both are much higher than in other people.1
Another review added up 63 papers. It covered about 6.9 million people. It found that autistic traits and psychosis traits often go together. If a person had one condition, they were about seven times as likely to have the other.2
A third review counted only conditions a doctor had named. It found bipolar disorder in about 5 in 100 autistic people. It found a schizophrenia disorder in about 4 in 100.3 Another review looked at adults with an average IQ. It found a schizophrenia disorder in about 6 in 100.4
In Manitoba, Canada, a whole country's records were checked. Autistic people there were about ten times as likely to have a schizophrenia diagnosis recorded. They were also about twice as likely to have a mood or anxiety problem.5
Two meta-analyses of the association, an all-ages meta-analysis of diagnosed rates and a whole-population register all point the same way. The designs and age groups differ, so the size of the gap varies. The direction is the same across them.
The Manitoba register gives a whole-population count of recorded schizophrenia in autistic people: about ten times the rate of a diagnosis than in other residents.5
Big reviews and a whole-country study agree. They studied different groups and ages, so the exact number is not the same in each one.
The Manitoba register gives a whole-population count of recorded schizophrenia in autistic people: about ten times the rate of a diagnosis than in other residents.5
Why the estimates disagree Emerging
The size of the overlap depends on how it is measured. The updated meta-analysis warns that the association may partly reflect methodological problems: autistic people may read psychosis questionnaires differently from how they were designed, and clinicians may find the two conditions hard to separate. In that analysis the link was strongest for the negative psychosis traits, where the measures overlap most, and weakest for the positive ones.2
A national cohort shows how far a screen can overreach. In the Netherlands Autism Register, 34.1% of 1,686 autistic adults (mean age 46) scored above the cutoff on a brief psychosis screen, the Prodromal Questionnaire-16.6 At follow-up, only 5 of 361 people (1.4%) reported a new schizophrenia-spectrum diagnosis, all of them among those who had screened positive. The authors treat the screen as a first step that needs a full assessment behind it, not a diagnosis.6
Where autism and schizophrenia are compared directly, they overlap on negative symptoms and differ on positive ones. In adults with autism (53), schizophrenia (39) and typical development (40), tested with the standard instruments for each condition, the autism and schizophrenia groups scored alike on negative symptoms, while positive symptoms separated them: the autistic group showed more repetitive behavior and stereotyped language, the schizophrenia group more delusions and hallucinations. The autism assessment tool also flagged many schizophrenia patients as autistic when they did not meet the criteria.7
Adolescent studies point the same way. A systematic review of 13 studies, mostly in children and adolescents, pooled four of them and found that about 12% of young people at clinical high risk for psychosis had an autism diagnosis, though the pooled estimate was very imprecise, with a 95% confidence interval of 2 to 44, and the studies disagreed widely; the autistic participants averaged about 11 years old and the high-risk samples about 18. The review found no difference in how often the high-risk group went on to develop psychosis, with or without autism.8
How a study measures psychosis changes the answer. The review of 63 papers warns about two problems. Autistic people may read the questions differently from how they were meant. And doctors may find it hard to tell the two conditions apart. The link was strongest for the negative traits. Those are the ones that look most like autism.2
A short check can give the wrong answer. In one big study, doctors gave a short psychosis check to 1,686 autistic adults. About 34 in 100 scored above the cutoff.6 Later, only 5 of 361 people were said to have a new schizophrenia problem. The authors say the check is only a first step. It should not be used as a diagnosis on its own.6
When the two conditions are compared, they look alike in the negative symptoms and differ in the positive ones. In one study, autistic adults, adults with schizophrenia, and other adults all took the tests. The autism and schizophrenia groups scored the same on negative symptoms. Positive symptoms set them apart. The autistic group had more repeated movements and set phrases. The schizophrenia group had more delusions and hallucinations.7
Studies of young people find the same thing. A review looked at 13 studies. Most were in children and teenagers. It found autism in about 12 in 100 young people at high risk of psychosis. But the number was very rough, and the studies did not agree. The autistic young people were about 11 years old on average. The high-risk group was about 18.8 The review found no difference in how often the high-risk group later got psychosis.8
The pooled figures sit at the top of a range, and their own authors warn the measures may inflate them. The clearest evidence is one national cohort and small clinical comparisons.
In psychosis, "positive" symptoms are added experiences such as hallucinations and delusions. "Negative" symptoms are reduced or absent ones, such as flat expression, few words and social withdrawal. Some negative symptoms resemble autistic traits.
The co-occurrence figures in this entry come mostly from adults, because psychosis and bipolar disorder usually begin in adulthood. The studies of how the two look alike are often in children and adolescents, where the conditions are harder to tell apart. Each claim here names its sample.
The big numbers may be too high, and even their authors warn about this. The best evidence comes from one large study and a few small ones.
In psychosis, "positive" symptoms are things added, like hallucinations and delusions. "Negative" symptoms are things missing, like a flat voice, few words, and pulling away from people. Some negative symptoms look like autistic traits.
Most numbers here come from adults. Psychosis and bipolar disorder usually start in adulthood. The studies about telling them apart are often in children and teenagers. Every fact here says who was studied.
Bipolar disorder alongside autism Emerging
Bipolar disorder alongside autism has its own clinical picture, though it rests on one study so far. In 62 adults with bipolar disorder and autism without intellectual disability, compared with a sex-matched group who had bipolar disorder alone, the mood symptoms had started earlier, more were given an "other specified bipolar and related disorder" diagnosis, more had an anxiety disorder, and temperament scores for cyclothymia, anxiety and depression were higher. The group with autism also had lower overall functioning.9
The study also reported a high frequency of adverse reactions to antidepressants and antipsychotics in the group with autism. Its authors describe the combination as a distinct phenotype with poorer tolerability and call for treatment tailored to the person.9
The pooled co-occurrence of bipolar disorder in autistic adults is about 7.5%, higher in females; across all ages, about 5% of autistic people have a diagnosed bipolar disorder.1,3
Bipolar disorder with autism has its own pattern, but only one study has looked at it. It compared 62 adults who had both conditions with adults who had bipolar disorder alone. In the group with autism, the mood problems started earlier. More of them had anxiety. Their scores for mood swings and low mood were higher. Their everyday functioning was lower.9
The study also found more bad reactions to medicine. The group with autism often reacted badly to antidepressants and antipsychotics. The authors say this mix needs treatment made for the person.9
Bipolar disorder is found in about 7 in 100 autistic adults. It is more common in women. Across all ages, about 5 in 100 autistic people have it.1,3
One clinical study compares bipolar disorder with and without autism; the pooled co-occurrence figures come from the section above. The clinical picture is a single center's and is not yet replicated.
A condition of mood episodes that swing between depression and mania or hypomania, with periods of usual mood in between.
One study compares bipolar disorder with and without autism. It comes from a single place, and no one has repeated it.
A condition where mood swings between deep lows and highs called mania. In between, the mood is usual.
Telling autism and psychosis apart Emerging
Some of what is read as psychosis in autistic people is autistic communication itself. In adolescents and young adults aged 11 to 20, autistic participants showed a distinct pattern of language that separated them from the psychosis groups: delayed echolalia, pedantic speech, and unusual difficulty with irony and sarcasm. In the same study, about a fifth of the first-episode and high-risk participants scored above a questionnaire cutoff for a pervasive developmental disorder on parent report, though none reached the cutoff for full autism; the authors call for more thorough autism assessment in these groups.10
Where the two are tested side by side, the negative symptoms are the shared ground and the positive symptoms the difference. In a small older study of 14 autistic men and 14 men with schizophrenia, none of the men with schizophrenia met autism criteria, but half of the autistic men met criteria for schizophrenia, disorganized type, on the strength of negative symptoms such as reduced expression and withdrawal.11 A review of childhood-onset schizophrenia found that, in the two largest studies, it was preceded by or overlapped a pervasive developmental disorder in 30% to 50% of cases, and that the two conditions share some risk genes.12 Catatonia is a related confusion: it can be read as a worsening of autism or as psychosis, and it is covered in Catatonia.
Some things that look like psychosis are just how autistic people communicate. In one study, autistic teenagers and young adults had a certain way of talking. They repeated phrases they had heard before. They used formal, exact speech. They found jokes and sarcasm hard.10 In the same study, about 20 in 100 of the other two groups scored above a check. The check was for a developmental condition. None reached the level for full autism. The authors ask doctors to check for autism more carefully in these groups.10
When the two are tested side by side, the negative symptoms are shared and the positive ones differ. In one old study, 14 autistic men and 14 men with schizophrenia were tested. None of the men with schizophrenia had autism. But half of the autistic men had a kind of schizophrenia. The sign was negative symptoms, like a flat voice and pulling away.11 A review of childhood schizophrenia found it came with a developmental condition in 30 to 50 in 100 cases. The two conditions also share some genes.12 Catatonia can also be mistaken for autism getting worse or for psychosis. It has its own entry: Catatonia.
The clearest evidence is small clinical comparisons, one of them decades old and male-only. They show group differences in speech and symptoms, not how well a clinician tells the conditions apart in practice.
Repeating words or phrases, either straight away or after a delay. It is common in autistic people and in some other speech and language differences.
The evidence is small studies, and one is many years old and studied only men. They show how the groups differ. They do not show how well a doctor tells them apart.
Repeating words or phrases. Some people repeat them right away. Others repeat them later. It is common in autistic people.
What happens after a diagnosis Emerging
Outcomes differ for autistic people. In a retrospective cohort of 224 autistic and 7,185 non-autistic people after a first episode of psychosis, the autistic group used more community services, including crisis lines and police mental-health detentions, were more likely to be admitted to a psychiatric hospital (adjusted hazard ratio 1.42), and stayed longer when admitted (median 112 days against 49). A learning disability shaped the pattern, lowering community service use while lengthening admissions. That study is a preprint and has not been peer reviewed.13
People with prominent autistic features had a harder course. In 328 people treated for a first episode of psychosis, those with marked autistic features (about 12% of the sample, identified by a rating scale rather than a diagnosis) had more medication side effects at the start and were about 3.6 times less likely to be in remission after two years.14
Medication is part of the picture. In US Medicaid data, antipsychotics were prescribed to 45% of autistic young people with an intellectual disability and 22% of autistic youth without one, far more than in a comparison group; off-label use was highest in the group with an intellectual disability.15 Among autistic adults, antipsychotic use has been linked to a higher likelihood of diabetes in one survey, on top of already high rates of weight, cholesterol and blood-pressure problems.16 The prescribing rates and side effects are covered in full in Psychiatric medication in autistic adults.
Care differs for autistic people. One study looked back at 224 autistic people and 7,185 other people after a first episode of psychosis. The autistic group used more crisis services. They were more likely to go into a psychiatric hospital. When they did, they stayed longer: about 112 days, against 49 days for others. Having a learning disability also changed the pattern. That study has not yet been checked by other experts.13
Autistic features can mean a harder time. In one study, 328 people were treated for a first episode of psychosis. About 12 in 100 had strong autistic features. They had more side effects from medicine. They were about 3.6 times less likely to get better after two years.14
Medicine is part of this. In US records, antipsychotics were given to about 45 in 100 autistic young people with a learning disability. They were given to about 22 in 100 autistic young people without one. Far fewer other young people got them.15 In one survey, autistic adults who took antipsychotics were more likely to have diabetes.16 There is more detail in Psychiatric medication in autistic adults.
One cohort is a preprint that has not been peer reviewed, and the rest are single studies. They concern service use and treatment tolerance, not whether any treatment works for autistic adults with psychosis.
One study is not yet checked by other experts. The others are single studies. They are about care and medicine, not about which treatment works.
What would change it Emerging
Most of the sources point to the same improvements. Screening should be a first step, not a diagnosis: the Prodromal Questionnaire-16 was tested in autistic adults and flagged a third of them, so its authors propose a two-step assessment that confirms a screen with a full clinical evaluation.6 Telling the conditions apart should lean on positive symptoms, which separate them better than negative ones.7 Measurement should name its sample and its tool, because trait questionnaires read psychosis into autistic people more readily than the conditions warrant.2
Screening tools made for this population are not yet settled. No psychosis or bipolar screening tool has been validated as adequate specifically in autistic adults as of October 2026. The one study that tested a screen in autistic adults, the PQ-16, found it usable as a first step, but it could not judge how well the screen predicts on only five incident cases.6 For diagnosis, the average-IQ meta-analysis notes that misdiagnosis between autism and schizophrenia-spectrum conditions is likely, especially in adults who were not identified in childhood.4
The studies point to the same fixes. A check should be a first step toward a full assessment, not a diagnosis by itself.6 Doctors should focus on the positive symptoms, which tell the two conditions apart better.7 Studies should say who they studied and what test they used.2
There is still no good check made for this group. No psychosis or bipolar check has been shown to work well in autistic adults as of October 2026. The one study that tried the PQ-16 check found it useful as a first step. But it could not test how well the check predicts. Only five people got a diagnosis.6 The average-IQ study says doctors should watch for mixing the two conditions up. This matters most for adults who were not checked as children.4
These are recommendations drawn from the studies and meta-analyses above. No study found as of October 2026 has tested whether acting on them improves outcomes.
These are ideas from the studies above. No study found as of October 2026 has tested whether they help.
In the community Community
Autistic writers describe a pattern in which their traits are read as psychosis or as bipolar disorder. Megan Griffith writes that when she first sought therapy she was diagnosed with bipolar disorder, which she says is "actually a very common misdiagnosis for both ADHD and autism"; she doubted it because her "mania" was "too mild, even for hypomania" and her cycles "too quick, even for ultra rapid cycling bipolar".c1 Daniel Tammet, writing about another autistic person's diagnosis, notes that a psychologist "talked of childhood schizophrenia (as the autistic spectrum was still called then in some parts)".c2
The clinician guide by Henderson, Wayland and White gives the mechanism. It writes that some aspects of autism "can make autistics look psychotic when they are not", that hallucinations are not a feature of autism though experiences resembling them are more common, and that unusual language such as invented words can add to the impression of psychosis.c3 The guide describes a case in which a psychologist concluded an adolescent had "a burgeoning (or possibly active) psychotic disorder" and the person was admitted, where staff saw no evidence of psychosis and raised autism instead.c3
Steve Silberman, writing the history, notes that "the terms autism, childhood schizophrenia, and childhood psychosis were used virtually interchangeably", which confounded research for decades and left children later recognized as autistic with a schizophrenia label.c4
Autistic writers say their traits are often read as psychosis or bipolar disorder. Megan Griffith writes that a doctor first told her she had bipolar disorder. She says this is "actually a very common misdiagnosis for both ADHD and autism". She did not believe it. Her "mania" was "too mild, even for hypomania". Her cycles were "too quick, even for ultra rapid cycling bipolar".c1 Daniel Tammet writes about another autistic person. A doctor had called it "childhood schizophrenia (as the autistic spectrum was still called then in some parts)".c2
A guide for doctors gives the reason. It says some parts of autism "can make autistics look psychotic when they are not". It says hallucinations are not part of autism. But autistic people are more likely to have experiences that look like them. Odd language, like made-up words, can also look like psychosis.c3 The guide tells of a young person a doctor thought was psychotic. The young person went into a hospital. The staff there saw no psychosis and asked about autism instead.c3
The writer Steve Silberman adds the history. He says "the terms autism, childhood schizophrenia, and childhood psychosis were used virtually interchangeably". That confused research for many years. Children we now call autistic were often told they had schizophrenia.c4
How autistic writers describe being read as psychotic or bipolar when they are not. This is not a research finding.
This is how autistic writers describe being seen as psychotic or bipolar. It is not a research finding.
Sources
Research
- 1Varcin KJ, Herniman SE, Lin A, et al. Occurrence of psychosis and bipolar disorder in adults with autism: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews, 2022. https://doi.org/10.1016/j.neubiorev.2022.104543
- 2Miles MR, Golm D, Palmer-Cooper E. The association between autism and psychosis and the tools used to measure it: An updated systematic review and meta-analysis. British Journal of Clinical Psychology, 2026. https://doi.org/10.1111/bjc.70020
- 3Lai MC, Kassee C, Besney R, et al. Prevalence of co-occurring mental health diagnoses in the autism population: a systematic review and meta-analysis. The Lancet Psychiatry, 2019. https://doi.org/10.1016/S2215-0366(19)30289-5
- 4Lugo Marín J, Alviani Rodríguez-Franco M, Mahtani Chugani V, et al. Prevalence of Schizophrenia Spectrum Disorders in Average-IQ Adults with Autism Spectrum Disorders: A Meta-analysis. Journal of Autism and Developmental Disorders, 2018. https://doi.org/10.1007/s10803-017-3328-5
- 5Singal D, Enns JE, Friesen K, et al. Mental health disorders and suicide risk among autistic children, adolescents and adults: a population-based study using linked administrative data. The Lancet Regional Health - Americas, 2026. https://doi.org/10.1016/j.lana.2026.101505
- 6Basic D, Begeer S, Ziermans T. Evaluating the Prodromal Questionnaire-16 for Screening and Prediction of Psychosis in Autistic Adults: Cross-Sectional and Longitudinal Findings From a National Cohort. Schizophrenia Bulletin, 2026. https://doi.org/10.1093/schbul/sbag160
- 7Trevisan DA, Foss-Feig JH, Naples AJ, et al. Autism Spectrum Disorder and Schizophrenia Are Better Differentiated by Positive Symptoms Than Negative Symptoms. Frontiers in Psychiatry, 2020. https://doi.org/10.3389/fpsyt.2020.00548
- 8Vaquerizo-Serrano J, Salazar de Pablo G, Singh J, Santosh P. Autism Spectrum Disorder and Clinical High Risk for Psychosis: A Systematic Review and Meta-analysis. Journal of Autism and Developmental Disorders, 2022. https://doi.org/10.1007/s10803-021-05046-0
- 9Barbuti M, Amadori S, De Rosa U, Perugi G. Clinical characterization of bipolar disorder in adults with autism without intellectual disability: a cross-sectional study. Journal of Affective Disorders, 2026. https://doi.org/10.1016/j.jad.2026.121527
- 10Solomon M, Olsen E, Niendam T, et al. From lumping to splitting and back again: atypical social and language development in individuals with clinical-high-risk for psychosis, first episode schizophrenia, and autism spectrum disorders. Schizophrenia Research, 2011. https://doi.org/10.1016/j.schres.2011.03.005
- 11Konstantareas MM, Hewitt T. Autistic disorder and schizophrenia: diagnostic overlaps. Journal of Autism and Developmental Disorders, 2001. https://doi.org/10.1023/a:1005605528309
- 12Rapoport J, Chavez A, Greenstein D, Addington A, Gogtay N. Autism spectrum disorders and childhood-onset schizophrenia: clinical and biological contributions to a relation revisited. Journal of the American Academy of Child and Adolescent Psychiatry, 2009. https://doi.org/10.1097/chi.0b013e31818b1c63
- 13Ward JH, Lewis JR, Weir EM, Ford TJ, Cardinal RN. Comparing outcomes following a first episode of psychosis in autistic and non-autistic people: a clinical retrospective cohort study. medRxiv, 2026 (preprint, not peer reviewed). https://doi.org/10.64898/2026.06.01.26354576
- 14Bioque M, Pérez-Ramos A, Llorca-Bofí V, et al. Clinical Implications of Autistic Features in Patients With a First Episode of Psychosis. Acta Psychiatrica Scandinavica, 2026. https://doi.org/10.1111/acps.70079
- 15Benevides TW, Rast J, Lee B, et al. Antipsychotic Drug Prescriptions for Transition-Age Youth on the Autism Spectrum. Journal of Autism and Developmental Disorders, 2026. https://doi.org/10.1007/s10803-026-07233-3
- 16Bishop L, Charlton RA, McLean KJ, et al. Cardiovascular disease risk factors in autistic adults: The impact of sleep quality and antipsychotic medication use. Autism Research, 2023. https://doi.org/10.1002/aur.2872
Community sources
- c1Griffith M. Welcome to AuDHD: How to Survive (and Thrive) as an Adult with Autism and ADHD. Jessica Kingsley Publishers, 2026. Part "Making Sense of Who You Are", chapter "My journey to AuDHD". Megan Griffith's site
- c2Tammet D. Nine Minds: Inner Lives on the Spectrum. The Experiment, 2025. Chapter "Danny".
- c3Henderson D, Wayland S, White J. Is This Autism? A Companion Guide for Diagnosing. Routledge, 2023. Part "Connecting culture", sections "In summary" and "Overlap and differences".
- c4Silberman S. NeuroTribes: The Legacy of Autism and the Future of Neurodiversity. Avery, 2015. Part "The Invention of Toxic Parenting". Steve Silberman's site