By its definition, autism spectrum disorder (ASD) is characterised by restricted patterns of behaviour, interest, or activities and impairment in social communication and interaction in one or more settings.

People with ASD display a wide array of clinical presentations and severity. Furthermore, people with ASD may have many additional mental health conditions, as well as intellectual disability (ID) and language impairments.

In this scenario, rehabilitation implements psychosocial and behavioural therapies to change the way autistic people communicate and interact and to enhance their functional independence in activities of daily living.

Music therapy has been used for ASD since 1950; music interaction helps autistic people to develop communication skills and the capacity for social interaction. In the process of music interaction, participants engage in verbal, nonverbal and pre-verbal modes of communication.

This review investigated the short- and medium-term effects of music therapy for people with autism in individual or group settings.

This review is important for

People with ASD, their loved ones/caregivers, health professionals caring for this population, general practitioners, researchers, and policymakers.

Outcomes of this review

This is an update of Cochrane Systematic Reviews (CSRs) published in 2006 and 2014. The primary outcomes analyzed were a global improvement, socializing, verbal and nonverbal communication, quality of life in one or more settings, total autism symptom severity, and adverse events. The CSR included 16 new studies, for a total of 26 studies with a total number of 1,165 participants published and/or registered up to August 2021.

Results showed that, compared to placebo or standard care, music therapy probably increases overall improvement and does not increase adverse events; it is likely to ameliorate life quality and total autism symptom severity immediately after therapy. It is uncertain whether music therapy is effective to increase social interaction or verbal and nonverbal communication at the end of therapy.

These results must be interpreted with caution because (i) the population in these studies was mostly composed of children and young adults; therefore, the results are not generalisable to the adult population; (ii) the first studies included in the review were limited in their generalizability because they included only a few types of music therapies, although new included studies analyzed different types of therapies in different settings, therefore expanding the generalizability of the results; and (iii) the evidence was very low to moderate; therefore, further research is likely to change the current results. The evidence was deemed as such because of the (iv) wide confidence intervals and (v) small sample sizes. There was (vi) poor reporting of the procedures for randomization and allocation or, for some studies, there was an overall (vii) lack of randomization and/or concealment. There was a (viii) high risk of blinding bias due to the nature of the intervention itself.

Author’s conclusion of the review

Compared with placebo or standard care, music therapy probably produces an overall improvement by the end of therapy, including increased life quality and decreased symptom severity, probably without increasing adverse events. It is uncertain whether music therapy may promote social interaction and communication, both verbal and nonverbal.

Future recommendations

The authors recommended that future trials should be (a) pragmatic and should be aware of the different types of music therapies, as well as of the relevant outcomes for autistic patients. Future research should (b) address long-term assessment of outcomes; (c) minimize the risk of bias, adopting rigorous designs and using standards on (d) randomization, (e) allocation, and (f) blinding.

Furthermore, trials should (g) compare the different types of music therapy and settings, and (h) continue to probe the effectiveness of music therapy compared to other interventions or standard care.

In the end, since the study population in this review was composed mainly of children and young adults, future studies should also (i) include the adult population.

Robin Kuruvila Sentinella