Music-based therapy helps non-verbal autistic children to utter speech for the first time

by vicky

Article sourced from:

For a few years now there has been a great deal of interest in Melodic Intonation Therapy, a singing-based intervention that has been extensively tested by Gottfried Schlaug’s Boston group. This specially adapted sing-song training has been shown to help people with non-fluent aphasia (usually after stroke) regain some ability to speak . Today’s blog is about a variant of this type of therapy that has been tested by one of the Schlaug lab members, Catherine Wan, which can aid speech production in non-verbal autistic children.

In June of this year I attended Music and Neurosciences IV conference in Edinburgh and this was where I first heard about Auditory Motor Mapping Training (AMMT) as an intervention to help non-verbal autistic children. I wrote about a talk given on the subject by Catherine Wan in a previous blog which you can read here . In the past few weeks her first paper has been published in PLoSONE.


Up to 25% of individuals with Autism Spectrum Disorders (ASD) have difficulties communicating with speech. Regaining some ability to speak is a good prognostic indicator for children with ASD however, therapeutic interventions have had limited success up until now.

Catherine’s study tested the effectiveness of AMMT, an intonation based therapy which teaches children associations between sounds and articulatory actions. The therapy aims to capitalise on the well documented sparing and even heightening of musical skills that is often reported in ASD children, alongside their enjoyment of musical activities ( see work by my colleague Pam Heaton). The aim of AMMT is to facilitate auditory-motor mappings by teaching children to speak at different pitches while playing tuned drums.

“The therapist introduces the target words or phrases by simultaneously intoning the words and tapping the drums tuned to the same pitches” (Wan et al. 2011: e25505)

The therapy is based on intensive repetition of the same words and phrases in a highly structured environment.

How might it work?

The theory is that AMMT will likely engage and possibly strengthen a multi-sensory frontoparietal network of regions and pathways in the brain that respond to visual, auditory and motor representations of the same action, and sensory motor feedback. This network is thought to be crucial in our ability to speak, which involves mapping sound (the words we want to say) onto motor actions (movements of our mouth, throat and tongue).


Six non-verbal ASD children between 5 and 9 years of age took part in the study. All had at least 18 months speech therapy prior to the study but had shown little progress. All were defined as ‘non-verbal’ meaning they had a complete absence of intelligible words.

During each session the child sat opposite the therapist. In between them were two drums, one tuned to C4 (261.626Hz) and one tuned to E flat (311.127Hz).

Between the ‘Hello song’ and the ‘Goodbye song’ the therapist introduced 15 bi-syllabic items such as ‘mommy’, ‘more please’ and ‘all done’.

Using Boardmaker pictures as visual cues, the therapist held the child’s two hands in theirs and encouraged them to play the drums while they spoke the words, thereby encouraging bi-manual movement.


The children’s vocal ability was tested with probe assessments using one set of trained items (the 15 test items) and one set of untrained items (15 similar items that were not presented during therapy). For each word/phrase the child’s best utterance was transcribed and scored by the number of consonants and vowels correctly produced.

Experimental control was taken as three baseline tests that were taken prior to therapy.

Within 15 sessions of AMMT all the children showed a good response to therapy and by 40 sessions all of them showed significant improvements in their speech production, which was maintained at 4 week and 8 week follow-ups. There were even improvements for the untrained items, showing that the approximate consonant-vowel combinations that the children had learned extended beyond the words presented in therapy.


This first paper on AMMT is billed as a ‘proof of concept’ paper to demonstrate that they therapy can have positive impacts on speech production, even in children who have not uttered a word for the first 9 years of their life. Although their speech output was still extremely limited compared to age matched children, the ability to utter a few everyday phrases is a crucial step forward in their development. The fact that their improvements in some cases extended into a small number of phrases that were not even part of the therapy sessions is also an extremely promising finding. This paper will provide the basis for future randomized control trials.

Paper: Wan CY, Bazen L, Baars R, Libenson A, Zipse L, et al. (2011) Auditory-Motor Mapping Training as an Intervention to Facilitate Speech Output in Non-Verbal Children with Autism: A Proof of Concept Study. PLoS ONE 6(9): e25505. doi:10.1371/journal.pone.0025505