Psychotropic Medicines and Autism: High Rates, Off-Label Use and Slow Change

Autistic people — and above all autistic people who also have a learning disability — are prescribed psychotropic medicines far more often than everyone else. These medicines help many people. What the data keep showing is how much of this prescribing sits outside what the medicine is licensed for, how often no matching diagnosis is recorded, and how long prescriptions run once they have started.

In ordinary life this seldom looks like a decision. It looks like a repeat prescription that began during a crisis and has renewed itself ever since, or a tablet offered within weeks when the communication assessment is years away. Each prescription may have been reasonable the day it was written. The question is what happened next.

Between a third and two-thirds of autistic people are prescribed a psychotropic medicine

The largest international review pooled 47 studies covering more than 300,000 autistic people, with data collected between 1976 and 2012. The share prescribed psychotropic medicines ranged from 2.7% to 80%, median 45.7% — 41.9% among children, 61.5% among adults. Antipsychotics were the class used most often, then ADHD medication and antidepressants, and psychotropic polypharmacy ran from 5.4% to 54% of people, median 23.0%^[1].

National figures sit across that range. Among 33,565 autistic children in US commercial insurance claims from 2001 to 2009, 64% filled at least one psychotropic prescription and 35% had two or more classes at once^[6]; in US insurance data for all ages in 2014, 64% of commercially insured and 69% of Medicaid-enrolled autistic people received psychotropic treatment^[7]. In UK primary care in 2015, 32.3% of 10,856 autistic people were prescribed a psychotropic medicine — 4.91 times the odds of a matched general-population group (95% CI 4.46 to 5.40) — and 9.8% had polypharmacy^[8]. In Germany, 33.0% of autistic patients aged 0–24 were prescribed psychotropics in 2009^[9]; in Denmark in 2017, 30% of autistic 6- to 17-year-olds filled two or more prescriptions from four drug classes, most often ADHD medication or melatonin^[10]; in Australia, 54.4% of 390 autistic children had at least one subsidised psychotropic prescription between 2002 and 2019, and 25.4% two or more classes within a year^[11].

None of this is, by itself, evidence of over-medication. Autistic people have high rates of the co-occurring conditions these medicines are meant for — ADHD, anxiety, depression, epilepsy, sleep problems — and treating a diagnosed condition is good healthcare. The concern is the gap between how much is prescribed and how much diagnosis, evidence and review sits behind it.

The gap is widest for autistic people who also have a learning disability

England's Public Health England study of GP records from April 2009 to March 2012 is still the most detailed picture. Among adults with learning disabilities, antipsychotics were prescribed on 17.0% of person-days and antidepressants on 16.9%; for adults recorded as autistic without a learning disability the figures were 8.2% and 17.0%. A relevant indication was recorded for only 41.9% of adults with learning disabilities prescribed antipsychotics. Compared with epidemiological estimates of mental illness in this population, the authors estimated that 13% were taking antipsychotics without a psychotic illness and 10% antidepressants without an affective illness — 30,000 to 35,000 adults once overlap is allowed for^[3].

A cohort of 33,016 adults with an intellectual disability at 571 UK practices between 1999 and 2013 showed the same shape: 21% had a record of mental illness, 25% of behaviour that challenges and 49% of psychotropic prescribing. Of those treated with antipsychotics by the end of follow-up, 71% had no record of severe mental illness, and new prescribing was higher for people with a record of behaviour that challenges (incidence rate ratio 2.08, 95% CI 1.90 to 2.27), autism (1.79) and dementia (1.42)^[2]. Behaviour that challenges is not a diagnosis: NHS England calls it closer to “a social construct that may identify a person's communication needs and the impact of environmental triggers”, and medicating it an unlicensed use^[12].

The disparity persists. In England in 2024–25, 13.8% of patients with a learning disability were prescribed antipsychotics against 0.9% of patients without one, and 22.4% antidepressants against 11.0%^[13]. A Finnish national cohort of 37,196 people with intellectual disabilities, matched to a comparison cohort, found 42.3% using psychotropics in 2019 against 15.0%, antipsychotics 28% against 3.3% and polypharmacy 18.2% against 3.6%^[14].

NHS England's own 2025 summary is that people with a learning disability are “thought to be 15 times more likely and autistic people 7 times more likely to be prescribed an antipsychotic than the general population”^[12]. In an English primary-care analysis published as a conference abstract — not journal peer-reviewed, so provisional — antipsychotic prescribing between 2001 and 2023 fell from 44% to 22% among autistic adults with an intellectual disability and from 10% to 7% among autistic adults without one, while non-autistic adults moved from 1.1% to 1.2%; autistic adults with an intellectual disability had indications recorded least often and the longest courses^[15]. Antidepressant prescribing has risen faster for autistic adults than for others since 1997: in the UK in 2023, 30% of autistic adults were prescribed one, against 14.7% of non-autistic adults^[16]. The inequalities facing people with autism and a learning disability run wider than medicines, but medicines are the easiest part to count.

Very little of this prescribing is licensed for autism

Few medicines are licensed anywhere with autism named on the label, and licences differ by country.

Guidelines are correspondingly narrow. NICE tells clinicians not to use antipsychotics, antidepressants or anticonvulsants for the core features of autism, in children or adults; where antipsychotics are considered for behaviour that challenges, a specialist should identify a target behaviour, review effects at 3–4 weeks and stop treatment if there is no clinically important response at 6 weeks — a use NICE itself flags as off-label^[22][23]. For people with a learning disability it adds that antipsychotics should be considered only when other interventions have not worked or the risk to the person or others is very severe, offered only alongside psychological or other interventions, at the minimum effective dose, with documented reasons, a plan for stopping and a full review after 3 months and then at least every 6 months^[24].

Practice sits some way from that. In Sweden, among under-18s starting an antipsychotic between 2008 and 2021, autism was the most commonly recorded likely indication (15.4% of all first dispensings), and of the dispensings with an identifiable indication, 81% were off-label^[25]. In Australian general practice, off-label antipsychotic prescribing to children and young people rose from 69.8% in 2011 to 79.7% in 2017^[26]; Australia at least names the practice, treating medication given primarily to influence behaviour as a regulated “chemical restraint” that registered disability providers must have authorised^[27].

Off-label does not mean ineffective. The 2023 Cochrane review of medicines for irritability and aggression in autism — 131 trials, 7,014 participants — found that atypical antipsychotics probably reduce irritability in the short term compared with placebo (standardised mean difference −0.90, 95% CI −1.25 to −0.55; 12 studies, 973 participants; moderate-certainty evidence), with no clear evidence of an effect on aggression and no clear evidence that antidepressants affect irritability^[28]. For the people these medicines suit, the benefit is real. What the trials do not cover is years of use, or prescriptions with no target recorded at all.

The side effects are the reason review matters

Pooling nine placebo-controlled paediatric trials of 3 to 8 weeks, the US risperidone label records 32.6% of children and adolescents gaining 7% or more of their body weight, against 6.9% on placebo^[17]; the autism irritability trial found a mean gain of 2.7 kg in eight weeks against 0.8 kg^[29]. NHS England lists sedation, weight gain, raised lipids, diabetes risk, movement disorders and hormonal changes among effects that “can impact on a person's quality of life”^[12].

At population level the absolute risks are small, but real and lasting. A Dutch study of 89,991 young people aged 0–19 who started antipsychotics between 2006 and 2022 compared each person with themselves over time: drug-treated type 2 diabetes rose by 9.02 extra cases per 10,000 users (95% CI 5.99 to 12.06) at five years, roughly one per 1,100 people treated^[30]. Movement side effects are also commoner in people with an intellectual disability than in others taking the same drugs (adjusted incidence rate ratio 1.30, 95% CI 1.18 to 1.42)^[31]. This is part of the wider picture of physical health in autistic adults.

Monitoring is not optional in guidance. NICE's guideline on psychosis in children and young people, which its learning disability guideline points to for monitoring^[24], expects baseline checks of weight, height, blood pressure, glucose or HbA1c, lipids, prolactin and movement effects before an antipsychotic is started, then weight weekly for six weeks and bloods at 12 weeks and every six months^[32]. In England, everyone on the learning disability register aged 14 and over is entitled to an annual health check including a structured medication review^[12].

STOMP and STAMP: a decade of effort, and a slow curve

Concern about psychotropic medicines given to people with a learning disability was raised by parents in the serious case review into Winterbourne View hospital in 2012; the Public Health England study followed in 2015; and in June 2016 NHS England launched STOMP — stopping over-medication of people with a learning disability, autism or both^[33]. STAMP, supporting treatment and appropriate medication in paediatrics, followed in December 2018 with the Royal College of Paediatrics and Child Health^[34]. Both programmes stress that they are “not anti-medication”^[12].

What has changed? In England's GP data, the share of patients with a learning disability prescribed antipsychotics fell from 15.5% in 2017–18 to 13.8% in 2024–25, while the share of patients without a learning disability stayed at 0.9%^[13]. That is real movement. It is also, across seven years of data, 1.7 percentage points.

Public Health England's own evaluation found that after the launch, among adults with learning disabilities the antipsychotic trend turned from rising to falling, anxiolytics began to fall and antidepressants stopped rising — though for autistic children and young people the only change was a steeper rise in hypnotic prescribing, and the report stressed that “it is not possible to say whether or not these were the result of the programme”^[33]. The primary-care abstract reaches a similar verdict for autistic adults: the disparity narrowed “both before and after” STOMP began^[15]. Antidepressant prescribing to people with a learning disability has meanwhile edged up, from 21.7% in 2022–23 to 22.4% in 2024–25^[13].

Progress is slow partly because stopping is genuinely difficult and the evidence for doing it well is thin. A meta-analysis of 26 studies of focused psychotropic medication review — mostly in older people, people with dementia and adults with an intellectual disability — found review associated with less prescribing, but studies reporting clinical outcomes “failed to demonstrate benefit”, on evidence the authors rated poor^[35]. In two Dutch studies of supported withdrawal from 141 people with intellectual disabilities, about half did not complete it^[36], and the Royal College of Psychiatrists, which backs the STOMP pledge, notes that withdrawal led to deterioration in behaviour for some people, with no reliable way to tell in advance who^[34]. That argues for careful, supported, well-monitored review — not for leaving prescriptions unexamined, and not for anyone stopping a medicine on their own^[4].

What the evidence asks of us

• A recorded reason, every time. A relevant indication was recorded for fewer than half of adults with learning disabilities prescribed antipsychotics^[3], yet NICE already requires the rationale, expected duration and a strategy for stopping to be documented^[24].
• Reviews that reach people. Structured, person-centred medication reviews are the core of STOMP and STAMP, and in England should form part of the annual health check from age 14^[12]. Their coverage should be published, not assumed.
• Fund the alternatives. NICE says antipsychotics should be offered only alongside psychological or other interventions^[24]; where those are years away, that cannot hold in practice.
• Monitor physical health as guidance describes. Baseline and repeat checks of weight, glucose, lipids, prolactin and movement effects are already specified^[32], and the metabolic^[30] and movement^[31] evidence says why.
• Publish autism-specific prescribing data. We found no other country publishing annual figures comparable to England's^[13], and autistic people without a learning disability are close to invisible in routine statistics.
• Fund research into safe, supported reduction. The best syntheses show prescribing changing without demonstrated benefit to people's lives^[35], and half of the Dutch withdrawal attempts were not completed^[36].

None of this is an argument against medicines. They treat conditions that autistic people have in large numbers, and for some people they are the difference between a life that works and one that does not. The argument is for the ordinary standard of care everyone else expects: a clear reason written down, the smallest dose that helps, a date to look again, an honest account of the side effects, and a conversation the person is part of.

Language: this piece uses identity-first language (“autistic people”), the preference of most autistic-led organisations, and uses clinical terms such as “autism spectrum disorder”, “intellectual disability” or “behaviour that challenges” only when naming a study, dataset, licence or guideline category.

Why we're publishing this

Health Insurance UK is a commercial health-insurance resource, not a charity and not a campaign. We compiled this evidence because autistic people, families and autistic-led organisations pressing for better prescribing should not have to reassemble it from thirty scattered sources. Our part is to put the official numbers and the regulators' own words in one place.

How to read this data

Almost everything here comes from routine records — GP databases, insurance claims and national registers — not trials or surveys. A prescription record shows that a medicine was prescribed or dispensed, not that it was taken, and an absent diagnosis code is not proof that no diagnosis exists: the authors of the UK cohort study note that prescriptions without a record of severe mental illness “are not necessarily inappropriate”. Effect measures are reported as the studies give them: an odds ratio is not a risk ratio. One source, marked in the text, is a conference abstract without journal peer review; England's figures are experimental statistics covering about 55% of registered patients. The evidence is concentrated in high-income countries. For the wider evidence, see our autism research library.

Use this data

Free to cite with attribution to Health Insurance UK. Charities, campaigners and journalists are welcome to reuse these figures and charts. For the underlying figures as a spreadsheet, get in touch.

Important: This article summarises published research for general information. It is not medical advice, and it does not recommend starting, stopping or changing any medicine. No one should stop or alter a prescribed medicine without talking to the prescriber first. If you have concerns about your own or someone else's health or medicines, please speak to a qualified health professional.

Sources

1. Jobski K, Höfer J, Hoffmann F, Bachmann C. Use of psychotropic drugs in patients with autism spectrum disorders: a systematic review. Acta Psychiatrica Scandinavica, 2017. europepmc.org/article/MED/27624381
2. Sheehan R, Hassiotis A, Walters K, Osborn D, Strydom A, Horsfall L. Mental illness, challenging behaviour, and psychotropic drug prescribing in people with intellectual disability: UK population based cohort study. BMJ, 2015. europepmc.org/article/MED/26330451
3. Glover G, Williams R, Branford D, Avery R, Chauhan U, Hoghton M, Bernard S. Prescribing of psychotropic drugs to people with learning disabilities and/or autism by general practitioners in England. Public Health England, 2015. knowledge.lancashire.ac.uk/id/eprint/17970
4. NHS. Common questions about risperidone. NHS.uk, reviewed 2025. nhs.uk/medicines/risperidone
5. NHS. Antidepressants — stopping or coming off antidepressants. NHS.uk, reviewed 2025. nhs.uk/medicines/antidepressants
6. Spencer D, Marshall J, Post B, Kulakodlu M, Newschaffer C, Dennen T, Azocar F, Jain A. Psychotropic medication use and polypharmacy in children with autism spectrum disorders. Pediatrics, 2013. europepmc.org/article/MED/24144704
7. Houghton R, Ong RC, Bolognani F. Psychiatric comorbidities and use of psychotropic medications in people with autism spectrum disorder in the United States. Autism Research, 2017. europepmc.org/article/MED/28963809
8. Houghton R, Liu C, Bolognani F. Psychiatric comorbidities and psychotropic medication use in autism: a matched cohort study with ADHD and general population comparator groups in the United Kingdom. Autism Research, 2018. europepmc.org/article/MED/30380202
9. Bachmann CJ, Manthey T, Kamp-Becker I, Glaeske G, Hoffmann F. Psychopharmacological treatment in children and adolescents with autism spectrum disorders in Germany. Research in Developmental Disabilities, 2013. europepmc.org/article/MED/23747941
10. Rasmussen L, Bilenberg N, Thomsen Ernst M, Abitz Boysen S, Pottegård A. Use of psychotropic drugs among children and adolescents with autism spectrum disorders in Denmark: a nationwide drug utilization study. Journal of Clinical Medicine, 2018. europepmc.org/article/MED/30308952
11. Baldes A, May T, Brignell A, Williams K. Patterns of psychotropic prescribing practices in autistic children and adolescents: an Australian perspective of two cohorts five years apart. Child Psychiatry & Human Development, 2024. europepmc.org/article/MED/38824199
12. NHS England. Stopping over medication of people with a learning disability and autistic people (STOMP) and supporting treatment and appropriate medication in paediatrics (STAMP). NHS England, 2025. england.nhs.uk/learning-disabilities/improving-health/stomp
13. NHS England Digital. Health and Care of People with Learning Disabilities, Experimental Statistics 2024 to 2025 (and earlier releases in the series). NHS England Digital, 2025. digital.nhs.uk/…/experimental-statistics-2024-to-2025
14. Virtanen V, Saastamoinen L, Arvio M, Vesala HT, Miettinen S, Teittinen A, Tolppanen AM. Prevalence of psychotropic use and psychotropic polypharmacy in a Finnish national cohort of persons with intellectual disabilities. Journal of Intellectual Disability Research, 2026. europepmc.org/article/MED/42454604
15. Sadik A, Pardiñas A, Khandaker G, Rai D, Madley-Dowd P. Antipsychotic prescribing trends for autistic adults with and without intellectual disability from 1997 to 2023: a population-based cohort study using English primary care records (conference abstract, not journal peer-reviewed). BJPsych Open, 2025. europepmc.org/article/PMC/PMC12242199
16. Sadik A, Khandaker GM, Pardiñas AF, Lundberg M, Lee BK, Magnusson C, Rai D, Madley-Dowd P. Antidepressant prescribing trends for adults with and without autism in the United Kingdom from 1997 to 2023. Autism Research, 2026. europepmc.org/article/MED/42658004
17. US Food and Drug Administration. RISPERDAL (risperidone) prescribing information, revised January 2025. FDA, 2025. accessdata.fda.gov/drugsatfda_docs/label/2025
18. US Food and Drug Administration. ABILIFY (aripiprazole) prescribing information, revised January 2025. FDA, 2025. accessdata.fda.gov/drugsatfda_docs/label/2025
19. Summary of Product Characteristics: Risperidone 0.5mg film-coated tablets, section 4.1. electronic medicines compendium, revised 2021. medicines.org.uk/emc/product/3957/smpc
20. European Medicines Agency. Abilify (aripiprazole): EPAR — therapeutic indications. EMA, updated 2026. ema.europa.eu/en/medicines/human/EPAR/abilify
21. European Medicines Agency. Slenyto (melatonin): EPAR — therapeutic indications. EMA, updated 2026. ema.europa.eu/en/medicines/human/EPAR/slenyto
22. National Institute for Health and Care Excellence. Autism spectrum disorder in under 19s: support and management (CG170), recommendations. NICE, 2013, updated 2021. nice.org.uk/guidance/cg170
23. National Institute for Health and Care Excellence. Autism spectrum disorder in adults: diagnosis and management (CG142), recommendations. NICE, 2012, updated 2021. nice.org.uk/guidance/cg142
24. National Institute for Health and Care Excellence. Challenging behaviour and learning disabilities: prevention and interventions (NG11), recommendations. NICE, 2015. nice.org.uk/guidance/ng11
25. S Kamgang KH, Pasternak B, Reutfors J, Pradhan R, Wintzell V. Trends in antipsychotic drug utilisation and indication in the paediatric population: a nationwide cohort study in Sweden. BMJ Mental Health, 2026. europepmc.org/article/MED/42303441
26. Klau J, Gonzalez-Chica D, Raven M, Jureidini J. Antipsychotic prescribing patterns in children and adolescents attending Australian general practice in 2011 and 2017. JCPP Advances, 2024. europepmc.org/article/MED/38486961
27. NSW Department of Communities and Justice. Restrictive practices guidance: chemical restraint. NSW Government, updated 2024. dcj.nsw.gov.au/…/restrictive-practices-guidance-chemical-restraint
28. Iffland M, Livingstone N, Jorgensen M, Hazell P, Gillies D. Pharmacological intervention for irritability, aggression, and self-injury in autism spectrum disorder (ASD). Cochrane Database of Systematic Reviews, 2023. europepmc.org/article/MED/37811711
29. McCracken JT, McGough J, Shah B, et al. Risperidone in children with autism and serious behavioral problems. New England Journal of Medicine, 2002. europepmc.org/article/MED/12151468
30. Gangapersad RN, García-Gómez P, Koch BCP, Moghani V, Dierckx B. Antipsychotic exposure and pharmacologically treated type 2 diabetes in Dutch youths. JAMA Network Open, 2026. europepmc.org/article/MED/42507442
31. Sheehan R, Horsfall L, Strydom A, Osborn D, Walters K, Hassiotis A. Movement side effects of antipsychotic drugs in adults with and without intellectual disability: UK population-based cohort study. BMJ Open, 2017. europepmc.org/article/MED/28775195
32. National Institute for Health and Care Excellence. Psychosis and schizophrenia in children and young people (CG155), recommendations. NICE, 2013, updated 2016. nice.org.uk/guidance/cg155
33. Mehta H, Glover G. Psychotropic drugs and people with learning disabilities or autism: executive summary. Public Health England, 2019. gov.uk/…/psychotropic-drugs-and-people-with-learning-disabilities-or-autism
34. Royal College of Psychiatrists. Position statement PS05/21: Stopping the over-medication of people with intellectual disability, autism or both (STOMP) and supporting treatment and appropriate medication in paediatrics (STAMP). RCPsych, 2021. rcpsych.ac.uk/…/position-statement-ps0521-stomp-stamp.pdf
35. Sheehan R, Strydom A, Brown E, Marston L, Hassiotis A. Association of focused medication review with optimization of psychotropic drug prescribing: a systematic review and meta-analysis. JAMA Network Open, 2018. europepmc.org/article/MED/30646263
36. Weijgertze-Lanser J, Wissing MBG, Elbers RG, Jonker J, de Kuijper GM, Maes-Festen DAM. Off-label antipsychotic withdrawal in people with intellectual disabilities: development and internal validation of a prediction model. Journal of Intellectual Disability Research, 2025. europepmc.org/article/MED/40908653