Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 May 26; 334(7603): 1069–1070.
PMCID: PMC1877953

Antipsychotic drugs in children with autism

Susan Morgan, medical assessor1 and Eric Taylor, professor of child and adolescent psychiatry2

Inadequacies in care should not be masked by the indiscriminate use of symptom controlling drugs

The core problems of autism—those involving social interaction, communication, and restricted and repetitive activities—can be compounded by behavioural problems, including severe tempers, aggression, and irritability.1 Severe aggression places a special burden on carers; it is more common in people with marked intellectual retardation and is related to poor daily living skills and impaired communication. Currently, no drugs are available to treat the underlying autistic condition. Specialised educational programmes, behaviour therapy, and environmental changes can improve aggressive behaviour,1 but if they fail drug treatments should be considered.2 Behavioural problems related to depression or attention deficit can be addressed by relevant therapy; but if the problem of aggression is unresponsive to these manoeuvres the need for symptom control arises. Major tranquillisers in particular have been used off-label, but their place has been uncertain because of doubts about safety and (until recently) efficacy.

Two well conducted double blind placebo controlled studies have compared placebo and the atypical antipsychotic risperidone in children with autism and behavioural problems. One study3 included 101 children with diagnosed autism. The other4 included 79 children with the broader category of “pervasive childhood developmental disorder,” the largest subset (n=55) having autism. Both studies showed that risperidone significantly improved a mixture of behavioural problems, including aggression measured on the irritability subscale of the aberrant behaviour checklist.5 This was accompanied by a global improvement as measured by the clinical global impression of change.

Adverse events were also reported such as somnolence (risperidone 67% v placebo 23%), extrapyramidal symptoms (risperidone 29% v placebo 10%), weight gain (risperidone 5% v placebo 0%), and raised prolactin concentrations (risperidone 43% v placebo 2%). Although prolactin concentrations tend to decrease with time, even while continuing risperidone, they are still higher than at baseline in longer term open label studies.6 The effect of this on growth (including bone mineral density) and sexual maturation is not known.

The licence holder (Janssen-Cilag) for risperidone applied to the UK licensing authority, the Medicines and Healthcare Regulatory Agency (MHRA), to include irritability in autism as a licensed indication. Although efficacy had been demonstrated, there was concern about the potential misuse of this drug as a form of long term chemical control, especially of the most intellectually disabled children, who may be the most likely to have adverse effects. Thus, the Committee on Safety of Medicines sought the views of experts in child psychiatry, paediatric endocrinology, and pharmacokinetics and of medical centres that specialise in treating autism. The National Autistic Society was consulted, as well as parents of autistic children, who were asked about their child's perception of the drug as well as their own. The overwhelming view of both experts and service users was that, if used appropriately, antipsychotic drugs had a positive role in the management of aggression associated with autism. This resulted in the offer of a conditional approval limiting the use of risperidone to the symptomatic treatment of severe aggression and violence in children with autism.

The conditions included, as part of a risk management plan, safety monitoring through a newly established registry of children on treatment, with regular written reports to the MHRA. However, after considering the conditions of approval and following discussions with the MHRA, the company withdrew its application. Thus, an opportunity to establish a safer mechanism for screening and monitoring autistic children on risperidone was lost. In addition, there is a risk that a wider indication for behavioural disturbances in children with mental retardation may be gained in the United Kingdom via European procedures.

The Food and Drug Administration in the United States has taken a different view and has licensed risperidone for unrestricted use for irritability in children and adolescents with autism.7 No limitations were put on the severity of the symptoms that might warrant risperidone and no restrictions were put in place, such as limiting prescription to experts in the field, although undertaking three further studies (two in animals and one clinical) was a condition of approval.

How should clinicians react? We consider that off-label use is still justified when other approaches fail or are unfeasible, and when underlying causes of aggression such as any physical condition or illness that causes distress to the child, adverse upbringing, or hyperkinetic disorder have been considered. These conditions are not contraindications to antipsychotic drugs, but attention to them may make medication unnecessary.

The assessment report on the use of risperidone in autistic children is available on the MHRA website.6 It recommends a conservative approach—that this type of drug should be prescribed by experts in the treatment of autism who are prepared to undertake careful diagnosis, appropriate screening, and monitoring.

Diagnosis should distinguish between aggression and other seriously challenging behaviours (which may justify an antipsychotic agent) and lesser levels of “irritability” (which may not). Screening should uncover any physical problems causing behavioural problems, such as seizures, or remediable problems in the care environment, such as a lack of special measures to promote appropriate communication. Ideally, monitoring should include a pretreatment baseline period, and growth (height, weight, sexual maturation, evidence of gynaecomastia), behavioural change (somnolence, paradoxical exacerbation of behavioural problems), extrapyramidal symptoms, bowel habit, and blood pressure should be monitored. Routine invasive monitoring, such as blood testing, is not a condition of prescribing as it is often unacceptable to affected children, but if a child is more than 10 centile points above the expected weight, fasting blood glucose, lipids, and prolactin concentrations should be measured if possible. Urinary glucose testing may be done if blood tests are not practical.

Children with autism are among the most vulnerable in our society, and as such should not be deprived either of effective medication or of precautions to optimise safety. Historically, society has not offered these children the highest standards of care, and it is vitally important that inadequacies in care provision are not masked by the indiscriminate use of symptom controlling drugs.


Competing interests: ET was an external expert for the Committee on Safety of Medicines. SM was the medical assessor of the use of risperidone in autism.

Provenance and peer review: Commissioned; not externally peer reviewed.


1. Howlin P. The effectiveness of interventions for children with autism. J Neural Transm Suppl 2005;69:101-19. [PubMed]
2. Lord C, Bailey A. Autism spectrum disorders. In: Rutter M, Taylor E, eds. Child and adolescent psychiatry, 4th ed. Oxford: Blackwell Science, 2002:654.
3. McCracken JT, McGough J, Shah B, CroninP, Hong D, Aman MG, et al; Research Units on Pediatric Psychopharmacology Autism Network. Risperidone in children with autism and serious behavioral problems. N Engl J Med 2002;347:314-21. [PubMed]
4. Shea S, Turgay A, Carroll A, Schulz M, Orlik M, Smith I, et al. Risperidone in the treatment of disruptive behavioral symptoms in children with autistic and other pervasive developmental disorders. Pediatrics 2004;114:1447-8.
5. Aman M, Singh NN, Stewart AW, Field CJ. The aberrant behaviour checklist: a behaviour rating scale for the assessment of treatment effects. Am J Ment Defic 1985;89:485-91. [PubMed]
6. Medicines and Healthcare Regulatory Agency. Medicines for children.

Articles from The BMJ are provided here courtesy of BMJ Group