Autism is defined by obvious shortcomings in a person’s social interactions and communications, with a lack of understanding of non-verbal language and the social rules of communication or responding appropriately to others, and inappropriate use of expressive language. People with autism may isolate themselves from others, not take part in socialising activities and be unable to play or use their imagination. Typical behaviour patterns include repetitive motor activity, obsessive and ritualistic activities and an exaggerated focus on parts of objects. Poor social functioning, communication and lack of flexibility of thought and behaviour are present from early childhood. They may also have problems with the behaviour they expect from others in response to their own.

We present some of the evidence from Cochrane systematic reviews about complementary and alternative treatments related to epilepsy. This evidence comes from carefully researched reviews of information about clinical trials done to evaluate medical treatments. Studies are only included in these reviews if they meet pre-defined criteria.

People with autism have shortcomings in their social interactions and communications, with a lack of understanding of non-verbal language and the social rules of communication, and inappropriate use of expressive language. Typical behaviour patterns include repetitive motor activity, ritualistic activities and an exaggerated focus on parts of objects. Behaviours that are present in early childhood persist when they are adults.

There are Cochrane reviews available for four possible therapies.


Some individuals are sensitive to small protein parts (peptides) found in gluten and casein, which are in wheat and dairy products. The inability to adequately process these substances might contribute to autism and disruptive behaviours.

If this is true, it is possible that diets free of gluten or casein, or both, might reduce the symptoms associated with autism and help improve a person’s behaviour, cognitive and social functioning.

What the synthesised research says

One small trial involving 20 children investigated the effects of a combined gluten and casein-free diet on the characteristics associated with autism compared with a standard diet. No clear differences could be seen between the cognitive skills, linguistic ability and motor ability of the children on the different diets.

How it was tested

The researchers made a thorough search of the medical literature and found only one relevant trial that was from Norway. This trial had 20 participants. These were children aged 5 to 10 years old who were matched in pairs by severity of autistic symptoms and randomly assigned within each pair to a combined gluten and casein-free diet or standard diet for 12 months. The trial followed a child’s verbal communication and body language, reaction when spoken to or with physical contact, behaviour, sharing of emotions and interaction with other children.

Side effects and general cautions

A gluten and casein-free diet has costs, financially and in terms of inconvenience. It also sets limitations on foods of choice for the affected family member and, indirectly, others in the family. More high-quality research is needed in this area.


Millward C, Ferriter M, Calver S, Connell-Jones G. Gluten- and casein-free diets for autistic spectrum disorder. The Cochrane Database of Systematic Reviews 2004, Issue 2. Art. No.: CD003498.pub2. DOI: 10.1002/14651858.CD003498.pub2.


Vitamin B6 (pyroxidine) was first used with children diagnosed with autism when speech and language improvement was observed in some children as a result of large doses of vitamin B6. Interpersonal skills and body function may also improve. The use of mega-vitamin intervention began in the 1950s with the treatment of schizophrenic patients. Vitamin B6 is given with magnesium to counter several undesirable side effects of vitamin B6, including irritability, sensitivity to sound and bedwetting.

What the synthesised research says

Findings for this review are inconclusive.

Trial authors of one small trial involving 10 participants did not find any clear benefits with taking vitamin B6 with magnesium compared with a non-active treatment (placebo) on measures of social interaction, communication, compulsivity, impulsivity, or hyperactivity, over eight weeks. Another study provided insufficient data to conduct an analysis.

The latest trial involved a group of eight children with pervasive developmental disorders or Asperger syndrome. Vitamin B6 supplement may have improved the verbal INTELLIGENCE/IQ score of those children taking it for four weeks, compared with a non-active supplement (placebo).

How it was tested

The researchers made a thorough search of the medical literature and found three relevant trials with a total of 33 participants with autism. Twenty-three boys and 10 girls took part. They were aged between three and 18 years old. Two trials were conducted in the USA and one in Japan, between 1993 and 2002. The earliest trial did not give sufficient information to draw any conclusions.

The treatment studied was combined vitamin B6 with magnesium, taken as a tablet or capsule, or vitamin B6 alone.

Side effects, limitations of the trials

The doses of vitamin B6 and magnesium varied between trials and the age range of children in the trials was large.

It has been reported that a long-term administration of pyridoxine may induce adverse effects such as a loss of feeling in fingers and toes (sensory peripheral neuropathy), but no such effects were reported in these relatively short-term trials.

The children in the trial that looked at Asperger syndrome and pervasive developmental disorders had different developmental histories and diagnosis characteristics. They were functionally less severely affected with different abilities than children with autism.


Nye C, Brice A. Combined vitamin B6-magnesium treatment in autism spectrum disorder. The Cochrane Database of Systematic Reviews 2005, Issue 4. Art. No.: CD003497.pub2. DOI: 10.1002/14651858.CD003497.pub2.


Adults and children with autistic spectrum disorders are limited in their verbal and nonverbal communication, have behaviour problems and difficulties with socialising. They may respond inappropriately to sounds, for example by covering their ears, screaming, or appearing to be deaf and have abnormal responses to other sensory stimuli including visual perception, temperature or pain.

Education and behavioural methods form the basis of treatment. Medications are often given to control their hyperactive, aggressive and obsessive-compulsive behaviours. Alternative therapies include diet, vitamins, and sound therapy.

Sound therapy involves listening to electronically modified music for varying periods of time as a way of improving hearing processing defects and a person’s concentration. The music is modified by filtering out selected frequencies and modulation by randomly alternating high and low sounds. A therapist may modify treatment to suit the individual.

Auditory integration training was developed by Berard in France in 1982 as a way of ‘re-educating’ the hearing process and improving behaviour and learning problems. He applied it to a variety of disorders including autism, depression, hyperactivity and learning difficulties. The training involves using headphones to listen to electronically modified music over a total 10 hours, two half-hour daily sessions over 10 days. The peak frequencies that an individual is ‘hypersensitive’ to are dampened out. Different intensities are used as well as alternating pitches, with care the listeners are not exposed to music loud enough to cause hearing loss.

The Tomatis sound therapy, developed by Dr Alfred Tomatis, uses electronically modified human voice as well as music. This is delivered via an ‘Electronic Ear’. The programs are individualised with different durations of therapy and breaks between treatments.

Samonas sound therapy is an adaptation of the Tomatis method and was developed by Steinbach. It utilises filtered music, voice and sounds of nature that are heard through headphones. Therapy takes place under a therapist's supervision and at home.

What the synthesised research says

No clear evidence was available for any beneficial effect of auditory integration therapy for people with autism.

No trials were found that assessed sound therapies other than auditory integration therapy.

How it was tested

The researchers made a thorough search of the medical literature and found six relevant trials. A total of 171 individuals with autism spectrum disorders, aged between three and 39 years, were randomly assigned to receive auditory integration training or non-active treatment. This involved listening to unmodified music through headphones or in the room but still wearing headphones.

Three trials included adults as well as children and were reported on between 1993 and 2000. The number of participants in the trials ranged from ten to 80. 14 months was the maximum duration of follow-up (used in one study only).

Three studies, two with children only, did not demonstrate any benefit of therapy over control conditions.

Three small trials reported some improvement in behaviour checklists after three months; these checklist measures were difficult to interpret in daily life. Staff members, teachers or parents collected the information.

Side effects, limitations of the trials

Three trials reported on possible adverse effects of the sound therapy. They reported minor physical complaints with both the active and non-active therapies. The treatment can be expensive. Any use of the intervention must be recommended only as ‘experimental’.

Many different outcome measures were used in the trials, which made it difficult to synthesis the information.


Sinha Y, Silove N, Wheeler D, Williams K. Auditory integration training and other sound therapies for autism spectrum disorders. The Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD003681.pub2. DOI: 10.1002/14651858.CD003681.pub2.


Autism refers to a range of developmental disorders described as autism spectrum disorders. These are life-long disorders. People with autism have difficulties in communicating, interacting socially and using their imagination. They have difficulties responding appropriately to others and can be rigid or obsessive in their behaviour. They may also have problems with the behaviour they expect from others in response to theirs. 

Music therapists use music to help with communicating and expressing feelings. Music therapy can be used long term, starting as early as possible in a child’s life. The therapist uses free and structured music making (improvisation), singing and listening to music as part of the sessions, which are usually one to one.

What the synthesised research says

Attending one-on-one daily sessions of music therapy over a week improved verbal communication and the use of gestures in communicating with the therapist. These benefits were shown in two trials involving a total of 20 children only, which is too few to have any clarity around the usefulness of music therapy.

The one trial that looked at the effects of music therapy on repetitive behaviour in the classroom was too small, with four children, to show any clear evidence of benefit.

How it was tested

The review authors searched the literature thoroughly and found three relevant trials. These were small and involved a total of only 24 children, mainly boys. The children, aged two to nine years, were randomized to attend daily music therapy sessions for a week or matched sessions that had the same activities without music as non-active treatment (placebo). These activities included reading a story and rhythmic speech instead of singing, or play activities without music (songs and instruments).

All the trials were carried out in North America. Music therapy was given in a very structured way compared to other, more improvisational approaches to music therapy. It involved listening to songs that contained a message with specific aims such as addressing behaviour, signs and words to be learned or to build a relationship; in one trial the children played guitar and drums too.

Side effects, limitations of the trials

The treatment can be expensive. Any use of the intervention must be recommended only as ‘experimental’. Music therapy was not harmful for children with autism.

The treatment that music therapy was compared against differed only in that music was not used. The children received the same amount of attention from the therapist and the children undertook the same structured activities. This means that they were likely to have received some benefit from the sessions. Neither medication use nor long-term benefits of music therapy were considered in these small trials.

Children with Rett’s disorder were not included in the trials.


C Gold, T Wigram, C Elefant. Music therapy for autistic spectrum disorder. The Cochrane Database of Systematic Reviews 2006, Issue 2. Art. No.: CD004381.pub2. DOI: 10.1002/14651858.CD004381.pub2.