Gold Et Al-2006-Cochrane Database of Systematic Reviews

Cochrane
Library
Cochrane Database of Systematic Reviews
Music therapy for autistic spectrum disorder (Review)
Gold C, Wigram T, Elefant C
Gold C, Wigram T, Elefant C.

Music therapy for autistic spectrum disorder.
Cochrane Database of Systematic Reviews 2006, Issue 2. Art. No.: CD004381.
DOI: 10.1002/14651858.CD004381.pub2.
www.cochranelibrary.com
Music therapy for autistic spectrum disorder (Review)

Copyright © 2010 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Cochrane Trusted evidence.
Informed decisions.
Library Better health. Cochrane Database of Systematic Reviews
TABLE OF CONTENTS
HEADER......................................................................................................................................................................................................... 1
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
BACKGROUND.............................................................................................................................................................................................. 3
OBJECTIVES.................................................................................................................................................................................................. 4
METHODS..................................................................................................................................................................................................... 4
RESULTS........................................................................................................................................................................................................ 6
DISCUSSION.................................................................................................................................................................................................. 7
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 8
ACKNOWLEDGEMENTS................................................................................................................................................................................ 9
REFERENCES................................................................................................................................................................................................ 10
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 14
DATA AND ANALYSES.................................................................................................................................................................................... 18
Analysis 1.1. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 1 Communicative skills: gestural.............................. 18
Analysis 1.2. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 2 Communicative skills: verbal................................. 18
Analysis 1.3. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 3 Behavioural problems (end of therapy)................. 19
Analysis 1.4. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 4 Behavioural problems (using all measurements 19
from 2nd day on)..................................................................................................................................................................................
WHAT'S NEW................................................................................................................................................................................................. 19
HISTORY........................................................................................................................................................................................................ 19
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 20
DECLARATIONS OF INTEREST..................................................................................................................................................................... 20
SOURCES OF SUPPORT............................................................................................................................................................................... 20
INDEX TERMS............................................................................................................................................................................................... 20
Music therapy for autistic spectrum disorder (Review) i

Informed decisions.
[Intervention Review]
Music therapy for autistic spectrum disorder
Christian Gold1, Tony Wigram2, Cochavit Elefant3
1Grieg Academy Music Therapy Research Centre (GAMUT), Unifob Health, Bergen, Norway. 2Institute of Music and Music Therapy,
University of Aalborg, Aalborg, Denmark. 3Grieg Academy Department of Music, University of Bergen, Bergen, Norway
Contact address: Christian Gold, Grieg Academy Music Therapy Research Centre (GAMUT), Unifob Health, Lars Hilles gate 3, Bergen,
5015, Norway. christian.gold@grieg.uib.no.
Editorial group: Cochrane Developmental, Psychosocial and Learning Problems Group.

Publication status and date: Edited (no change to conclusions), published in Issue 1, 2010.
Citation: Gold C, Wigram T, Elefant C. Music therapy for autistic spectrum disorder. Cochrane Database of Systematic Reviews 2006, Issue
2. Art. No.: CD004381. DOI: 10.1002/14651858.CD004381.pub2.
ABSTRACT
Background
The central impairments of people with autistic spectrum disorder (ASD) include social interaction and communication. Music therapy
uses music and its elements to enable communication and expression, thus attempting to address some of the core problems of people
with ASD.
Objectives
To review the effects of music therapy for individuals with autistic spectrum disorders.
Search methods
The following databases were searched: CENTRAL, 2005, (Issue 3); Medline, (1966 to July 2004); Embase, (1980 to July 2004); LILACS, (1982
to July 2004); PsycINFO, (1872 to July 2004); CINAHL, (1982 to July 2004); ERIC, (1966 to July 2004); ASSIA, (1987 to July 2004); Sociofile,
(1963 to July 2004); Dissertation Abstracts International, (late 1960's to July 2004). These searches were supplemented by searching specific
sources for music therapy literature and manual searches of reference lists. Personal contacts to some investigators were made.
Selection criteria
All randomised controlled trials or controlled clinical trials comparing music therapy or music therapy added to standard care to "placebo"
therapy, no treatment or standard care.
Data collection and analysis

Studies were independently selected, quality assessed and data extracted by two authors. Continuous outcomes were synthesised using
a standardised mean difference (SMD) in order to enable a meta-analysis combining different scales, and to facilitate the interpretation of
effect sizes. Heterogeneity was assessed using the I2 statistic.
Main results
Three small studies were included (total n = 24). These examined the short-term effect of brief music therapy interventions (daily sessions
over one week) for autistic children. Music therapy was superior to "placebo" therapy with respect to verbal and gestural communicative
skills (verbal: 2 RCTs, n = 20, SMD 0.36 CI 0.15 to 0.57; gestural: 2 RCTs, n = 20, SMD 0.50 CI 0.22 to 0.79). Effects on behavioural problems
were not significant.
Music therapy for autistic spectrum disorder (Review) 1

Informed decisions.
Authors' conclusions
The included studies were of limited applicability to clinical practice. However, the findings indicate that music therapy may help children
with autistic spectrum disorder to improve their communicative skills. More research is needed to examine whether the effects of music
therapy are enduring, and to investigate the effects of music therapy in typical clinical practice.
PLAIN LANGUAGE SUMMARY
Music therapy for people with autistic spectrum disorder
People with autism spectrum disorders (ASD) have difficulties with communication, behaviour and/or social interaction. Music therapy
uses music and its elements to enable people to communicate and to express their feelings. In this way music therapy addresses some
of the core problems of people with ASD. This review set out to assess the evidence for the effectiveness of music therapy for individuals
with ASD.
Three small studies were included which examined the short-term effect of brief music therapy interventions for autistic children. Music
therapy was superior to "placebo" therapy with respect to verbal and gestural communicative skills, but it was uncertain whether there was
an effect on behavioural outcomes. The included studies were encouraging, but of limited applicability to clinical practice. More research
with better design, using larger samples, in more typical clinical settings is needed to strengthen the clinical applicability of the results and
to examine how enduring the effects of music therapy are. When applying the results of this review to practice, it is important to note that
the application of music therapy requires specialised academic and clinical training.

Informed decisions.
BACKGROUND The clinical picture varies because individuals have variable

levels of ability, from profound learning disability to a spiky
Description of the condition cognitive profile where superior skills are present in some areas
of functioning. Children and adults with autistic spectrum disorder
Autistic spectrum disorder (ASD) is the core disorder of
frequently pose considerable behavioural challenges to their
the pervasive developmental disorders as defined within the
parents and other family members (Diggle 2002). Prevalence
International Classification of Diseases and Related Health
estimates for autistic spectrum disorder range from 5 to 63 children
Problems, tenth edition (ICD-10) (WHO 1992) and the Diagnostic
per 10,000, with more recent estimates showing much higher
and Statistical Manual of Mental Disorders (DSM IV) (APA 1994).
prevalence rates than those from older studies (Chakrabarti 2001,
DSM-IV-TR (APA 2000) describes the diagnostic criteria of autism
Fombonne 1999). At the high functioning end of the autistic
to include three core elements. First it is a qualitative impairment
spectrum is a disorder known as Asperger Syndrome, with the
in social interaction, as manifested by at least two of the
same fundamental core impairments as autism, but also some
following: marked impairment in the use of multiple nonverbal
differences in language development, motor skills and originality of
behaviors such as eye-to-eye gaze, facial expression, body
thought (Asperger 1979).
postures, and gestures to regulate social interaction; failure to
develop peer relationships appropriate to developmental level; Description of the intervention
a lack of spontaneous seeking to share enjoyment, interests,
or achievements with other people (e.g.,by a lack of showing, Music therapy has been defined as "a systematic process of
bringing, or pointing out objects of interest) or a lack of social or intervention wherein the therapist helps the client to promote
emotional reciprocity. Second, it requires there to be qualitative health, using musical experiences and the relationships that
impairments in communication as manifested by at least one of develop through them as dynamic forces of change" (Bruscia
the following: delay in, or total lack of, the development of spoken 1998, p. 20). Central techniques in music therapy include free and
language (not accompanied by an attempt to compensate through structured improvisation, songs, and listening to music.
alternative modes of communication such as gesture or mime);
in individuals with adequate speech, marked impairment in the How the intervention might work
ability to initiate or sustain a conversation with others; stereotyped The processes that occur within musical improvisation may help
and repetitive use of language or idiosyncratic language; or a lack people with autism spectrum disorder to develop communicative
of varied, spontaneous make-believe play or social imitative play skills and their capacity for social interaction. Musical interaction
appropriate to developmental level. Finally there must be restricted in music therapy, in particular musical improvisation, is sometimes
repetitive and stereotyped patterns of behavior, interests, and understood and described as a kind of non-verbal and pre-
activities, as manifested by at least one of the following: verbal language which enables verbal people to access pre-
encompassing preoccupation with one or more stereotyped and verbal experiences, enables non-verbal people to interact
restricted patterns of interest that is abnormal either in intensity communicatively without words, and enables all to engage
or focus; apparently inflexible adherence to specific, nonfunctional on a more emotional, relationship-oriented level than may be
routines or rituals; stereotyped and repetitive motor mannerisms accessible through verbal language (Alvin 1991). Listening to music
(e.g. hand or finger flapping or twisting, or complex whole-body within music therapy also involves an interactive process that
movements); or persistent preoccupation with parts of objects. often includes selecting music that is meaningful for the person
(e.g. relating to an issue that the person is occupied with) and
The age of onset is also a factor, and delays or abnormal functioning
where possible reflecting on personal issues related to the music
in at least one of the following areas, with onset prior to age 3 years
or associations brought up by the music. For those with verbal
is necessary: (1) social interaction, (2) language as used in social
abilities, verbal reflection on the musical processes is often an
communication, or (3) symbolic or imaginative play. DSM-IV-TR also
important part of music therapy (Wigram 2002).
require that the disturbance is not better accounted for by Rett's
Disorder or Childhood Disintegrative Disorder. Music therapy for individuals with autistic spectrum disorders is
usually provided as individual therapy. A rationale for the use of
Individuals with autistic spectrum disorders have difficulties in all
music therapy for individuals with communication disorders is
aspects of communication. They are said to lack a "theory of mind",
based on the findings from infancy researchers such as Stern (Stern
sometimes called lacking a "sixth sense". They display impairments
1985, Stern 1989) and Trevarthen (Trevarthen 1999a) who describe
in social communication and social interaction and a restricted
sound dialogues between mothers and infants using "musical"
imagination and social repertoire, the latter characteristically
terms. When describing tonal qualities researchers use the terms
displayed as what seems to others to be obsessional behaviour
pitch, timbre and tonal movement, and when describing temporal
and rigidity in their own behaviour as well as in the behaviour they
qualities they speak of pulse, tempo, rhythm, and timing (Wigram
require from others in response to their own. The key construct
2002). Trevarthen (Trevarthen 1999b) describes the sensitivity of
is the "triad of impairment", which affects social interaction,
very young infants to the rhythmic and melodic dimensions of
language and communication, and behaviour and imagination
maternal speech, and to its emotional tone, as demonstrating that
(Wing 1997), that can be identified through examination of early
we are born ready to engage with the 'communicative musicality'
development and current presentation (Wing 2002). People with
of conversation, and this premise allows music to act as an
autistic spectrum disorder also present with a pervasive inability
effective medium for engaging in non-verbal social exchange for
to 'mind-read' (Baron-Cohen 1995), where a lack of perception
children and adults with autistic spectrum disorder. Necessary
and understanding of other people's feelings, beliefs or emotions
communicative behaviours, such as joint attention, eye contact
results in a consequential inability to respond appropriately. This
and turn-taking are characteristic events in shared, active music
has particular impact on social skills and interactions (Howlin
making.
1998).
Informed decisions.
Why it is important to do this review -Activity level

-Quality of life in both school and home environments
Clinical reports and pre-experimental studies suggest that music -Stress in the family
therapy may be an effective intervention for people with -Adverse events
autism spectrum disorder. For example, Edgerton examined Data sources could include non-standardised or standardised
the development of communicative skills in eleven autistic instruments (for a review of relevant standardised instruments see
children over the course of music therapy sessions, finding a (Ozonoff 2005), parent or teacher report, or school
continuous increase of communicative acts and responses in all records.
subjects (Edgerton 1994). Schumacher described qualitatively how
relationship patterns of autistic children change and develop Data from rating scales were included only if the instrument was
during long-term music therapy (Schumacher 1999a, Schumacher either a self-report or completed by an independent rater or relative
1999b). Two systematic reviews pertaining to the scope of this (not the therapist).
review have yielded conflicting results: One review (Whipple
2004) concluded that music therapy was effective for people with Search methods for identification of studies
autistic spectrum disorders; however, interventions and study
The following search terms were used:
designs were too heterogeneous to allow clinically meaningful and
#1 MUSIC
methodologically strong conclusions. The other review (Ball 2004)
#2 MUSIC THERAPY
concluded that effects were unknown, but failed to identify many
#3 musi*
possibly relevant studies. Thus, a more comprehensive systematic
#4 gim
review of controlled studies in this area is necessitated.
#5 ((guided imagery) near music)
OBJECTIVES #6 vibroacoustic
#7 vibro-acoustic
To review the effects of music therapy, or music therapy added to #8 (#1 or #2 or #3 or #4 or #5 or #6 or #7)
standard care, for individuals with autism spectrum disorders. #9 (asperger next syndrome)
#10 autis*
METHODS #11 kanner*
#12 (childhood near schizophren*)
Criteria for considering studies for this review #13 (speech near disorder*)
#14 (language near delay*)
Types of studies
#15 pdd
All relevant randomised controlled trials (RCTs) and controlled #16 CHILD DEVELOPMENT DISORDERS, PERVASIVE
clinical trials (CCTs). #17 (#9 or #10 or #11 or #12 or #13 or #14 or #15 or #16)
#18 (#8 and #17)
Types of participants
Where necessary, the search terms were modified to suit the
Individuals of any age who are diagnosed with a pervasive
requirements of the other databases searched. An optimal sensitive
developmental disorder, as defined in ICD-10 or DSM-IV, whether
search strategy for randomised controlled trials was also used
identified by a psychological assessment or a psychiatric diagnosis.
where necessary.
This includes childhood autism (F84.0 in ICD-10), atypical autism
(F84.1), Asperger's syndrome (F84.5), and pervasive developmental Relevant trials were identified through searching the following
disorder not otherwise specified (F84.9). Individuals with Rett's databases:
disorder (F84.2) or childhood disintegrative disorder (F84.3) are Cochrane Central Register of Controlled Trials (CENTRAL, 2005,
not included as they do not conventionally fall within the autistic Issue 3)
spectrum disorders, given their significantly different clinical Medline (1966 to July 2004)
course. Embase (1980 to July 2004)
LILACS (1982 to July 2004)
Types of interventions
PsycINFO (1872 to July 2004)
Music therapy (regular sessions of music therapy as defined above), CINAHL (1982 to July 2004)
delivered by a professional, compared with either "placebo" (the ERIC (1966 to July 2004)
concept of attention placebo in psychotherapy research is ASSIA (1987 to July 2004)
discussed in Kendall 2004), no-treatment or standard care control; Sociofile (1963 to July 2004)
or music therapy added to standard care compared with standard Dissertation Abstracts International (late 1960s to July 2004)
care (with or without "placebo").
The following specific sources for music therapy literature were also
Types of outcome measures searched:
musictherapyworld.net (accessed July 2004)
Measures of the following including:
Music Therapy Research CD ROM (AMTA 1999)
-Communicative and social skills, social interaction
Music Therapy World Info-CD ROM IV (Aldridge 2002)
-Quality of social interaction
References were also retrieved from related review articles (Ball
-Behavioural problems (e.g. stereotypic behaviour)
2004, Whipple 2004).
-Attention and concentration
-Cognitive ability
-Hyperacusis (hypersensitivity to sound)
Informed decisions.
Data collection and analysis 3.2 Loss to follow-up and intention to treat analysis
The included studies had complete data for all participants,
1. Selection of trials and therefore an intention-to-treat analysis was straightforward.
Two reviewers independently inspected the titles and abstracts (If studies with drop-outs had been included, we would have
identified from the search. Potentially relevant papers were examined the impact of drop-out rates using sensitivity analyses
obtained. Any disagreement was resolved through discussion and/ and imputing the negative outcome as appropriate.)
or consultation with a third reviewer.
4. Data analysis
2. Assessment of quality Data from washout periods in crossover studies were excluded
The authors assessed methodological quality independently, and from the analysis.
any disagreements were resolved by discussion. Assessment was
made of all included studies, to consider the following questions: 4.1 Binary data
Was the assignment to treatment groups truly random? No binary data were available from the included studies. We had
Was allocation adequately concealed? planned to use risk ratio and number needed to treat statistics with
How complete was follow up? 95% confidence intervals for binary outcomes.
How were the outcomes considered for people who withdrew?
Were those assessing outcomes blind to the treatment allocation? 4.2 Continuous data
For studies where outcomes were measured at several
Randomisation occasions during each treatment condition, we used only the last
Randomisation was judged as 'adequate' when computer- measurement at the end of treatment.
generated random numbers, a random numbers table, or coin- Where raw data were available, the distributions of values were
tossing were used to allocate participants to treatment conditions. visually checked in an attempt to detect skewness. Where skewness
Randomised as well as quasi-randomised trials were included in was found, we attempted to remove it by log-transformation. We
the review, as noted above. then examined how log-transformation influenced the effect size
estimate and used the more conservative estimate. The alternative
Allocation concealment of using change data which tend to be less skewed was not
The trials were divided into quality categories as defined in the considered because using change data may be less conservative
Cochrane Reviewers' Handbook (Alderson 2004), where: than using endpoint data.
We used standardised mean differences (SMD) for the analysis of
A: adequate allocation concealment; participants and researchers all continuous outcomes. When combining different scales for the
were unaware of participants' future allocation to condition until same outcome, it was necessary to standardise the effects in order
after decisions about eligibility were made and informed consent to make them comparable. When combining results on the same
was obtained; scale, either non-standardised weighted mean differences (WMDs)
or SMDs could have been used. We decided to use SMDs even in this
B: unclear concealment; allocation concealment measures were
case in order to facilitate the interpretation of effect sizes as small,
not described in detail;
medium or large according to the guidelines that are commonly
C: inadequate allocation concealment; allocation was not used in the behavioural sciences (Cohen 1988). It is noted that the
concealed from either participants before informed consent or choice of SMD versus WMD does not usually affect the significance
from researchers before decisions about inclusion were level of the results, and the authors cautiously assessed whether
made (this will always be the case for quasi-randomised studies). such was the case.
As this review aimed to include randomised and quasi-randomised
All SMDs (regardless of whether the study was a parallel or
studies, all three categories were eligible for inclusion. The rating
a crossover design) were standardised by the pooled standard
was only used as a descriptive measure of study quality.
deviation between participants, rather than the standard deviation
of the difference within participants. This is the standard procedure
Blindness of assessors
which enables comparisons of different scales and facilitates
It is not possible to blind either those who deliver music therapy or
interpretation of the magnitude of effects (Cohen 1988, Gold 2004).
those who receive it to the nature of the intervention. Blindness of
The calculation of the standard error then depended on the study
assessors, however, can be used and was assessed as a part of this
design. For parallel designs the standard error was calculated
review. Quality of blindness was determined primarily by whether
using the standard formulae for SMDs as implemented in RevMan
those who assessed and coded outcome measures were blind to
and described in the RevMan handbook. For crossover studies
condition, and the quality of blindness was categorised where:
we took into account the correlations within the participants as
Adequate = assessor blind to condition; recommended and described in the literature on meta-analysis of
Unclear = blinding of assessor not reported and information not crossover studies (Elbourne 2002).
available from researchers; and
4.3 Meta-Analysis
Inadequate = assessor not blind to condition.
A fixed effect model was used in all analyses. Fixed effect models
All of the above were included in the review.
have the advantage of being simpler and more easily interpretable.
3. Data management If a common effect size had not been tenable due to heterogeneity,
3.1 Data extraction a random effects model would have been considered. In addition
This was performed independently by two reviewers (CG, TW). to the fixed effect analyses, the authors also examined whether
When necessary the reviewers contacted the authors of trials to random effects analyses would have altered the results, and any
provide missing data. such difference would have been reported.
Informed decisions.
The length of all included studies was extremely short, varying from
5. Assessing heterogeneity one (Farmer 2003) to four weeks (Brownell 2002). The duration of
Consistency of results was assessed visually in the forest plot each treatment condition was only one week in each study. No later
and by examining I2 (Higgins 2002), a quantity which describes follow-up assessments were included in any of the studies.
approximately the proportion of variation in point estimates
that is due to heterogeneity rather than sampling error. We 4.2 Participants
supplemented this with a test of homogeneity to determine the The participants in the included studies were between 2 and 9
strength of evidence that the heterogeneity is genuine. In case of years of age, with 80-100% boys in each study. All participants had
heterogeneity, possible sources would have been investigated. received a diagnosis of autism. In one study (Buday 1995), levels
of retardation and severity of autistic traits were also specified.
6. Assessing bias Participants in this study ranged from mildly to severely mentally
Funnel plots were planned to be used to investigate any retarded (according to DSM III-R) and from mildly to moderately
relationship between effect size and study precision, but were not autistic (according to the Childhood Autism Rating Scale, CARS).
used in this review due to the small number of included studies.
4.3 Setting
7. Sensitivity analyses The participants received therapy either at home, at school, or at
Sensitivity analysis were planned to be conducted to determine the an outpatient therapy centre.
impact of study quality on outcome if studies of different quality
had been identified and included (for example studies with high 4.4 Study size
attrition rates). All three studies had extremely small sample sizes, varying from
four to ten participants per study. Crossover designs were used
8. Subgroup analyses in two of the studies (Brownell 2002 and Buday 1995) to partly
The impact of clients' age and intensity of therapy (number and compensate for the small sample size.
frequency of music therapy sessions) would have been examined in
subgroup analyses had heterogeneity been found. 4.5 Interventions
RESULTS 4.5.1 Music therapy

In all studies, music therapy was provided on a daily basis in an
Description of studies individual (one-to-one) setting. The duration of the music therapy
condition was only one week in all studies.
1. Excluded studies All studies used a highly structured approach to music therapy,
The search yielded a total of 312 references. 52 of these were which is not unexpected given the American origin of the studies
deemed possibly relevant and selected for closer inspection. Of (see Wigram 2002, Gold 2005, and the discussion section of this
these, 13 were excluded from the review because they concerned review). All studies used receptive techniques (listening to music);
an assessment rather than an intervention (e.g. assessing traits one study also used active techniques (Farmer 2003).
of people with ASD using music therapy techniques). 24 studies Songs sung by the music therapist were composed or chosen
were excluded because they did not have an appropriate design individually for the participants and were usually used with specific
(5 studies with ABAB or similar designs, i.e. studies comparing aims. For example, songs were based on a social story addressing a
different treatments that all participants received in the same central problem behaviour of the particular individual in treatment
order; 10 case series; 9 case studies). 7 studies employed an (Brownell 2002); they contained signs and words to be learned
intervention involving only listening to music (e.g. auditory (Buday 1995); or they were used to build a relationship and to
integration training), rather than music therapy. One study was provide a safe and understandable structure for the participants in
excluded because the outcome measure was unclear (see table of the study (Farmer 2003).
excluded studies).
Active music-making by the participants, which is often typical for
2. Awaiting assessment
music therapy in clinical practice (Wigram 2006), was reported in
Three studies (unpublished master theses) could not be retrieved
only one study (Farmer 2003). Participants were allowed to play
and are therefore still awaiting assessment (Laird 1997, O'Dell 1998,
guitar and drums. Playing instruments was partly used to reinforce
Wood 1991).
adjusted behaviour. The report did not specify whether and in what
3. Ongoing studies ways the therapist improvised or otherwise played music together
One relevant longer-term study of improvisational music therapy with the client.
was still ongoing when this review was written (Kim 2005).
4.5.2 "Placebo" therapy
4. Included studies All studies compared music therapy to some kind of "placebo"
Three short-term American studies comparing music therapy to activity to control for the non-specific effects of therapeutic
a "placebo"-type therapy met the criteria for the review (Brownell attention. Since in all studies music was considered as the
2002, Farmer 2003, Buday 1995; see table of included studies). Two specific ingredient of music therapy, the placebo conditions were
of the studies (Brownell 20002 and Buday 1995) were of crossover constructed to closely match the music therapy condition, only
design; the remaining included study (Farmer 2003) was parallel that music was not used. For example, a social story was read
group. Other characteristics of these studies are described below. instead of sung to the participants (Brownell 2002); rhythmic
speech was used instead of singing (Buday 1995); or the same play
4.1 Length of trials activities were offered without using songs or music instruments
(Farmer 2003). Therefore, the "placebo" conditions included many

Informed decisions.
elements similar to the music therapy conditions, including some 3. Loss to follow-up
potentially active ingredients (but none that are specific to music No dropouts were reported in any of the studies.
therapy). The concept of a "placebo" therapy designed to control
for the therapist's attention has been widely used in psychotherapy 4. Overall
research (Kendall 2004), but has also been criticised (Lambert 2004;
4.1 Performance bias
see the discussion section of this review).
Medication levels were not monitored in the included studies.
4.5.3 Other conditions However, due to the short duration of the studies, it appears
One study (Brownell 2002) also reported the outcomes during a unlikely to assume significant differential change in medication
baseline and a washout period with no intervention. These data might distort the results. All participants received the full therapy
were not used in this review. intervention as intended.
4.6 Outcome measures 4.2 Data reporting and analysis

One study (Buday 1995) reported means, standard deviations,
4.6.1 Communicative skills: gestural and F test results for the outcomes described above. From these
Non-verbal (gestural) communicative skills were examined in statistics it was possible to calculate an SMD with a standard error
two studies (Buday 1995, Farmer 2003). Both studies addressed as appropriate for crossover studies (see Elbourne 2002 and the
the participants' behaviour within therapy sessions. Independent section on continuous data in the method section of this review).
observers counted the number of communicative gestures (e.g. For the other two studies, individual patient data were extracted
imitating a sign or motion) in the session. In one study (Buday from tables or graphs. We screened the data for skewness before
1995), the outcome consisted simply of the frequency count of data synthesis. The data of one study (Farmer 2003) showed a
appropriate gestures within a session. In the other study (Farmer skewed distribution. A log transformation would have removed the
2003) a completed gesture was given a score of 2 and an attempt skewness, but would also have increased the effect size estimate.
a score of 1, and the outcome consisted of the sum of these scores Therefore we decided to use the more conservative original scale.
for all attempted and completed gestures within a session. The
exact criteria for what was seen as a communicative gesture were Effects of interventions
different between the two studies. The measures used for this
COMPARISON 1: MUSIC THERAPY versus "PLACEBO" THERAPY
outcome were not published.
1. Communicative skills: gestural
4.6.2 Communicative skills: verbal
Skills in non-verbal, gestural communication were measured on
Communicative skills in verbal communication were addressed
continuous scales addressing observed behaviour. The results
in the same two studies (Buday 1995, Farmer 2003). Independent
showed a significant effect in favour of music therapy, suggesting
observers rated in-session behaviour in a similar way as in the
that improvement in gestural communicative skills was more likely
previous outcome only that it was the frequency of appropriate
to occur with music therapy than with a similar therapy not
verbal responses that was counted for this outcome. Again, the
including music (2 RCTs, n = 20, SMD 0.50 CI 0.22 to 0.79). Results
measures used were unpublished.
were consistent between the two studies (I2 = 0%; see figure 1.1).
4.6.3 Behavioural problems
2. Communicative skills: verbal
One study (Brownell 2002) addressed individually targeted
Observed skills in verbal communication were measured on
repetitive problem behaviour. Occurrence of behaviour was
continuous scales. The results showed a significant effect favouring
assessed outside therapy sessions. Independent observers
music therapy over the "placebo" intervention, suggesting that
(teachers) counted how often the targeted behaviour occurred in
improvement in verbal communicative skills was more likely to
the classroom. The frequency count was used as the outcome
occur with music therapy (2 RCTs, n = 20, SMD 0.36 CI 0.15 to 0.57).
measure. No published scale was used.
The results were consistent between studies (I2 = 0%; see figure 1.2).
Risk of bias in included studies 3. Behavioural problems
1. Randomisation The frequency of observed problem behaviour was measured on
Two studies (Buday 1995, Farmer 2003) stated explicitly that a continuous scale. Data were only available from one trial with
randomisation was used to assign participants to conditions. The four participants, and results were not significant when examining
remaining study (Brownell 2002) used the term "counterbalanced" only the last day in therapy (figure 1.3). However, when averaging
to describe an assignment that was either random or quasi- participants' behaviour over all days in therapy except the first
random, but intended to be random. Methods of randomisation one (figure 1.4), there was a significant effect suggesting that
and allocation concealment were not specified in the studies. music therapy may be slightly more beneficial than a similar verbal
therapy in reducing behaviour problems (1 RCT, n = 4, SMD -0.24 CI
2. Blindness and quality of outcome assessment -0.45 to -0.03).
There was one single-blind study, with blinded assessors (Buday
1995). It was not reported whether blindness was used in the other DISCUSSION
studies. All studies used independent raters to a6ssess outcomes.
All studies reported a high inter-rater reliability for the assessment Findings
of outcomes (Brownell 2002 inter-rater reliability 0.86 to 0.94;
Music therapy was compared to a "placebo" therapy which
Buday 1995: agreement rate 98%; Farmer 2003: agreement rate
attempted to control for all non-specific elements of music
91%).
therapy, such as the attention of a therapist. Outcomes assessed

Informed decisions.
included communicative skills and behavioural problems, and Control conditions

results were significant for communicative skills and reached The included studies used a dismantling strategy to isolate the
borderline significance for behavioural problems, suggesting a effect of the specific "ingredients" of music therapy by setting
beneficial effect of music therapy. up comparison conditions which were very similar to the music
therapy interventions, excluding only the music component. Any
The effect sizes found in the results of this review can be interpreted conclusion from such comparisons will therefore address the
in accordance with common guidelines for interventions in the effects of specific music therapy techniques, rather than the
behavioural sciences (Cohen 1988). The effect on non-verbal absolute effects of music therapy. This type of design is justified
(gestural) communicative skills reached a medium effect size (SMD when exploring the intervention strategies in which music therapy
0.50), which is seen as a clinically relevant magnitude when works. However, such comparison conditions may introduce some
comparing an active therapy condition to a "placebo" therapy. artificiality into the studies through selecting out and applying
The effects on verbal communicative skills were slightly smaller a single intervention strategy. This is not typically undertaken in
(SMD 0.36) and ranged between a small and a medium effect clinical treatment, although it does isolate a specific cause-effect
size. Considering that the "placebo" therapy possibly contained relationship. In the broader field of psychotherapy research, similar
not only "non-active" but also some of the "active ingredients" of constructions of "placebo" therapy to control for the therapist's
music therapy (see the discussion below), this can still be seen as attention and the non-specific elements have been broadly used
a clinically relevant magnitude. It is however interesting to note (Kendall 2004, pp. 20-21). However, recent research on the common
that non-verbal communicative skills, which may be more closely factors in psychotherapy raise the question of how adequate
related to the non-verbal communication within music therapy, it is conceptually, and also whether it is technically possible,
seemed to show greater change than verbal communicative skills. to separate the active from the non-active elements of therapy
However, it may also be that non-verbal communicative skills (Lambert 2004, pp. 150-152). In any case, the results of the included
are relatively easier to address than verbal communicative skills studies are likely to underestimate the true effects of music therapy,
especially in low-functioning children. Whether the effects of music because the control conditions contain a number of potentially
therapy on communicative skills are generalisable and persistent efficacious techniques which are also used in music therapy.
remains to be shown.
The data on behavioural problems were very limited. The outcome Duration and population
was only addressed in the smallest of the included studies. The Time-limited, intensive treatments such as those examined in the
effects were non-significant at the end of the short-term music included studies are relevant in acute care, and probably also in
therapy programme, and barely significant when averaging over the light of limited available resources. However, ASD is pervasive
the course of the therapy. Either way, the effect size was small. developmental disorder leading to a chronic condition which
Clearly, more data are needed to draw any conclusions about the requires sustained therapeutic intervention starting as early as
effects of music therapy on behaviour problems. possible. Music therapy is, in clinical reports for ASD, also described
as a longer-term intervention, and given the typical emergence of
Applicability of findings
entrenched and deteriorating behaviour, therapeutic intervention
Music therapy conditions
relies on consolidating progress over time. Conclusions about
The included studies were of limited generalisability to clinical
the potential benefits of longer-term music therapy cannot be
practice. Only a limited subset of the music therapy techniques
drawn from the available studies. With regards to the population
described in the clinical literature were used in the experimental
addressed, the applicability of the findings is limited to the age
treatment conditions. Receptive music therapy techniques with
groups included in the studies. No direct conclusions can be drawn
a high level of structuring predominated in the interventions;
about adults with ASD.
improvisational techniques were not mentioned. Approaches with
a high level of structure, including behaviourally oriented and
Strength of the evidence
directed interventions, are more often applied in the North
The limited information on randomisation methods and allocation
American context than in Europe, and the findings will therefore
concealment, the limited use of assessor blindness, the lack of
apply more easily to the former. However, improvisational
using standardised scales, and the small sample size limit the
techniques are widely used in many parts of the world (Edgerton
methodological strength of the evidence. However, there was no
1994, Kim 2006, Wigram 2006).
performance bias (co-intervention) or attrition bias (drop-out),
The included studies illustrate the value of structure, which and inter-rater reliability was high, and these factors contribute
is generally an essential element for children with ASD. Music positively to the strength of the evidence.
contains rhythmic, melodic, harmonic and dynamic structure
AUTHORS' CONCLUSIONS
which, when applied systematically and skilfully, can be effective
in engaging children with ASD. Intervention strategies employing Implications for practice
music improvisation are usually not pre-structured in the sense of a
fixed manual. Flexible but systematic treatment protocols for music The findings of this review indicate that music therapy may have
therapy are currently developing in clinical practice and research positive effects on the communicative skills of children with autistic
investigations in autism (Kim 2005, Wigram 2006) as well as in other spectrum disorder. Music therapy has been shown to be superior
fields (Rolvsjord 2005). One small RCT is currently investigating to similar forms of therapy where music was not used, and this
the effects of improvisational music therapy (Kim 2005). Further may be indicative of a specificity of the effect of music within music
such studies which are close to clinical practice will be needed to therapy. As only short-term effects have been examined, it remains
improve the applicability of findings. unknown how enduring the effects of music therapy on verbal and
non-verbal communicative skills are.

Informed decisions.
When applying the results of this review to practice, it is important had very small sample sizes and included no discussion of test
to note that the application of music therapy requires an academic power, possibly indicating that the issue has been ignored (see
and clinical training in music therapy. Trained music therapists Gold 2004 for a discussion of the issue of test power in music
are available in many countries. Training courses in music therapy therapy research). Limited sample size is common in research
teach not only the clinical music therapy techniques as described on ASD, and crossover designs seem an obvious choice because
in the background of this review, but also aim at developing the they help to improve test power. However, crossover designs
therapist's personality and clinical sensitivity, which is necessary entail other problems. First, longer-term follow-up assessments
to apply music therapy responsibly. Academic training courses in are not possible with such designs. Second, crossover designs
music therapy exist in many countries, and information is usually are only adequate for interventions whose effects are thought to
available through the professional associations. be short-acting (Elbourne 2002), but for music therapy it is not
unreasonable to assume lasting effects if children learn new modes
Implications for research of behaviour as a result of therapy. In this case, crossover designs
would be inappropriate due to carryover effects which would lead
Relation of studies to clinical practice
to the effect being underestimated. Large parallel group trials
More research is needed to examine the effects of music
are necessary to resolve this dilemma. An informed discussion is
therapy over a longer term. Future studies should involve
needed about what designs are considered appropriate in research
therapy conditions that are close to clinical practice, especially in
on music therapy for ASD.
terms of frequency, duration, and therapy techniques. This may
include further efficacy studies as well as effectiveness studies.
ACKNOWLEDGEMENTS
Standardised and published tools should be used to evaluate the
outcome of music therapy. We would like to thank Jane Dennis and Jo Abbott of the Cochrane
Developmental, Psychosocial and Learning Problems Group for
Sample size, power, and the choice of research design their help throughout the process, and Lisa Tjosvold of the
Future research on music therapy for people with ASD will need to Cochrane Child Health Field for her help in retrieving additional
pay close attention to sample size and power. All included studies unpublished studies.

Informed decisions.
REFERENCES
References to studies included in this review Bruscia 1982 {published data only}
Brownell 2002 {published data only} Bruscia KE. Music in the assessment and treatment of echolalia.
Music Therapy 1982;2(1):25-41.
Brownell MD. Musically adapted social stories to modify
behaviors in students with autism: four case studies. Journal of Carroll 1983 {published data only}
Music Therapy 2002;39(2):117-144. [MEDLINE: PBS Record: 280]
Carroll JG. The use of musical verbal stimuli in teaching low-
Buday 1995 {published data only} functioning autistic children (doctoral thesis). University of
Mississippi, 1983.
Buday EM. The effects of signed and spoken words taught with
music on sign and speech imitation by children with autism.. Chilcote-Doner 1982 {published data only}
Journal of Music Therapy 1995;32(3):189-202. [MEDLINE: PBS
Chilcote-Doner SE. The effect of contingent vs. non-contingent
Record: 1220; 0022-2917]
presentation of rhythmic asynchonous stimulation on the
Farmer 2003 {published data only} stereotyped behavior of children with autism (doctoral thesis).
Vanderbilt University, 1982.
Farmer KJ. The effect of music vs. nonmusic paired with
gestures on spontaneous verbal and nonverbal communication Clauss 1994 {published data only}
skills of children with autism between the ages 1-5 (Master's
Clauss EL. Effects of music on attention and self-stimulatory
thesis). Tallahassee, FL: Florida State University (School of
behaviors in autistic people (doctoral thesis). Hofstra University,
Music), 2003.
1994. [MEDLINE: PBS Record: 1340; 0419-4217]
Dawson 1998 {published data only}

References to studies excluded from this review
Dawson G, Meltzoff AN, Osterling J, Rinaldi J, Brown E.
Applebaum 1979 {published data only} Children with autism fail to orient to naturally occurring
Applebaum E, Egel AL, Koegel RL, Imhoff B. Measuring social stimuli. Journal of Autism and Developmental Disorders
musical abilities of autistic children. Journal of Autism and 1998;28(6):479-485. [MEDLINE: PBS Record: 980; 0162-3257]
Developmental Disorders 1979;9(3):279-285. [MEDLINE: PBS
Record: 2180; 0162-3257] Diez 1989 {published data only}
Diez Cuervo A, Garcia de Leon M, Gonzalez Sanz L. Estudio de
Bettison 1996 {published data only} funciones cognitivas en ninos autistas mediante cartografia
Bettison S. The long-term effects of auditory training on electrica cerebral (mapping) [Brain mapping study of cognitive
children with autism. Journal of Autism and Developmental functions in autistic children]. Archivos de Neurobiologia
Disorders 1996;26(3):361-374. [MEDLINE: PBS Record: 190] 1989;52(1):9-16. [MEDLINE: PBS Record: 1700; 0004-0576]
Blackstock 1978 {published data only} Edelson 1999 {published data only}
Blackstock EG. Cerebral asymmetry and the development Edelson SM, Arin D, Bauman M, Lukas SE, Rudy JH, Sholar M,
of early infantile autism. Journal of Autism and Childhood Rimland B. Auditory integration training: A double-blind study
Schizophrenia 1978;8(3):339-353. [MEDLINE: PBS Record: 2190; of behavioral and electrophysiological effects in people with
0021-9185] autism. Focus on Autism and Other Developmental Disabilities
1999;14(2):73-81. [MEDLINE: PBS Record: 880; 1088-3576]
Bonnel 2003 {published data only}
Bonnel A, Mottron L, Peretz I, Trudel M, Gallun E, Bonnel AM. Edgerton 1994 {published data only}
Enhanced pitch sensitivity in individuals with autism: a Edgerton CL. The effect of improvisational music therapy on
signal detection analysis. Journal of Cognitive Neuroscience the communicative behaviors of autistic children. Journal of
2003;15(2):226-35. [MEDLINE: PBS Record: 270] Music Therapy 1994;31(1):31-62. [MEDLINE: PBS Record: 1420;
0022-2917]
Brown 1994 {published data only}
Brown SM. Autism and music therapy - is change possible, and Frissell 2001 {published data only}
why music?. Journal of British Music Therapy 1994;8:15-25. Frissell JK. Musical preferences and reproduction abilities
in children with autism. Dissertation AbstractsInternational:
Brown 2003 {published data only} Section B: The Sciences and Engineering 2001;62(3-B):1574.
Brown WA, Cammuso K, Sachs H, Winklosky B, Mullane J, [MEDLINE: PBS Record: 760; 0419-4217]
Bernier R, Svenson S, Arin D, Rosen-Sheidley B, Folstein SE.
Autism-related language, personality, and cognition in people Goldstein 1964 {published data only}
with absolute pitch: results of a preliminary study. Journal of Goldstein C. Music and creative arts therapy for an autistic child.
Autism & Developmental Disorders 2003;33(2):163-7; discussion Journal of Music Therapy 1964;1(4):135-138. [MEDLINE: PBS
169. [MEDLINE: PBS Record: 250] Record: 2460; 0022-2917]

Informed decisions.
Gore 2002 {published data only} Ma 2001 {published data only}

* Gore AK. The effect of music versus emphasised speech on Ma YC, Nagler J, Lee MHM, et al. Impact of music therapy
echolalic autistic children (master's thesis). Tallahassee: Florida on the communication skills of toddlers with pervasive
State University, 2002. developmental disorder. Annals of the New York Academy of
Sciences 2001;930:445-447.
Gore AK. The effect of music versus emphasised speech on
echolalic autistic children. Annual conference of the American Mahlberg 1973 {published data only}
Music Therapy Association. Atlanta, GA, 2002. Mahlberg M. Music therapy in the treatment of an autistic child.
Journal of Music Therapy 1973;10(4):189-193. [MEDLINE: PBS
Griggs 1997 {published data only}
Record: 2370; 0022-2917]
Griggs Drane ER, Wheeler JJ. The use of functional assessment
procedures and individualized schedules in the treatment of Miller 1979 {published data only}
autism: Recommendations for music therapists. Music Therapy Miller SB, Toca JM. Adapted melodic intonation therapy: a case
Perspectives 1997;15(2):87-93. [MEDLINE: PBS Record: 1050; study of an experimental language program for an autistic child.
0734-6875] J Clin Psychiatry 1979;40(4):201-3..
Hadsell 1988 {published data only} Mottron 2000 {published data only}
Hadsell NA, Coleman KA. Rett syndrome: A challenge for music Mottron L, Peretz I, Menard E. Local and global processing of
therapists. Music Therapy Perspectives 1988;5(2):52-56. music in high-functioning persons with autism: Beyond central
coherence?. Journal of Child Psychology and Psychiatry and
Hairston 1990 {published data only}
Allied Disciplines 2000;41(8):1057-1065. [MEDLINE: PBS Record:
* Hairston MP. Analyses of responses of mentally retarded 780; 0021-9630]
autistic and mentally retarded nonautistic children to
art therapy and music therapy. Journal of Music Therapy Mudford 2000 {published data only}
1990;27(3):137-150. [MEDLINE: PBS Record: 1640; 0022-2917] Mudford OC, Cross BA, Breen S, Cullen C, Reeves D, Gould J,
Douglas J. Auditory integration training for children with
Heaton 1999 {published data only}
autism: no behavioral benefits detected. American Journal of
Heaton P, Hermelin B, Pring L. Can children with autistic Mental Retardation 2000;105(2):118-29. [MEDLINE: PBS Record:
spectrum disorders perceive affect in music? An experimental 320]
investigation. Psychological Medicine 1999;29(6):1405-10.
[MEDLINE: PBS Record: 330] O'Connell 1974 {published data only}
O'Connell TS. The musical life of an autistic boy.. Journal
Heaton 2003 {published data only}
of Autism and Childhood Schizophrenia 1974;4(3):223-229.
Heaton P. Pitch memory, labelling and disembedding in autism. [MEDLINE: PBS Record: 2310; 0021-9185]
Journal of Child Psychology and Psychiatry and Allied Disciplines
2003;44(4):543-51. [MEDLINE: PBS Record: 260] O'Loughlin 2000 {published data only}
O'Loughlin R. Facilitating prelinguistic communication skills
Kolko 1980 {published data only}
of attention by integrating a music stimulus within typical
Kolko DJ, Anderson L, Campbell M. Sensory preference and language intervention with autistic children (doctoral thesis).
overselective responding in autistic children. Journal of Autism University of Toledo, 2000.
and Developmental Disorders 1980;10(3):259-271. [MEDLINE:
PBS Record: 2120; 0162-3257] Pasiali 2004 {published data only}
Pasiali V. The use of prescriptive songs in a home-based
Krauss 1982 {published data only}
environment to promote social skills acquisition by children
Krauss T, Galloway H. Melodic intonation therapy with with autism: Three case studies. Music Therapy Perspectives
language delayed apraxic children. Journal of Music Therapy 2004;22(1):11-20.
Pasiali V. The use of prescriptive therapeutic songs to promote
Lee 2004 {published data only} social skills acquisition by children with autism: Three case
Lee L-YL. Using musical improvisation to effect linguistic studies. Annual conference of the American Music Therapy
and behavioral changes in a cohort of Taiwanese autistic Association. Atlanta, GA, 2002.
children. 6th annual conference of the American Music Therapy
Association. Austin, Texas, 2004. Rao 2001 {published data only}
Rao PA. Social interactional changes in mothers of autistic
Litchman 1976 {published data only} children following an acoustic intervention (doctoral thesis).
Litchman MD. The use of music in establishing a learning College Park: University of Maryland, 2001.
environment for language instruction with autistic children
(doctoral thesis). Buffalo: State University of New York, 1976. Saperston 1973 {published data only}
[MEDLINE: PBS Record: 2250; 0419-4217] Saperston B. The use of music in establishing communication
with an autistic mentally retarded child. Journal of Music

Informed decisions.
Therapy 1973;10(4):184-188. [MEDLINE: PBS Record: 2360; characteristics and other related pervasive developmental
0022-2917] disorders (master's thesis). Tallahassee: Florida State University,
1997.
Schmidt 1976 {published data only}
Schmidt DC, Franklin R, Edwards JS. Reinforcement of O'Dell 1998 {published data only}
autistic children's responses to music. Psychological Reports O'Dell AW. Effects of paired auditory and deep pressure
1976;39(2):571-577. [MEDLINE: PBS Record: 2290; 0033-2941] stimulation on the stereotypical behaviors of children with
autism (master's thesis). Tallahassee: Florida State University,
Starr 1998 {published data only} 1998.
Starr E, Zenker K. Understanding autism in the context of music
therapy: Bridging theory and practice. Canadian Journal of Wood 1991 {published data only}
Music Therapy 1998;6(1):1-19. [MEDLINE: PBS Record: 970; Wood SR. A study of the effects of music on attending behavior
1199-1054] of children with autistic-like syndrome (master's thesis). San
Jose State University, 1991.
Staum 1984 {published data only}
Staum MJ, Flowers PJ. The use of simulated training and music
lessons in teaching appropriate shopping skills to an autistic References to ongoing studies
child. Music Therapy Perspectives 1984;1(3):14-17. Kim 2005 {unpublished data only}
Stevens 1969 {published data only} * Kim, J. Joint attention and attunement in improvisational
music therapy with the autistic child. Aalborg University 2005.
Stevens E, Clark F. Music therapy in the treatment of autistic
children. Journal of Music Therapy 1969;6(4):98-104. [MEDLINE:
PBS Record: 2430; 0022-2917] Additional references
Thaut 1987 {published data only} Alderson 2004
Thaut MH. Visual versus auditory (musical) stimulus preferences Alderson P, Green S, Higgins JPT. Cochrane Reviewers'
in autistic children: a pilot study. Journal of Autism and Handbook 4.2.2 [updated December 2003]. The Cochrane
Developmental Disorders 1987;17(3):425-32. [MEDLINE: PBS Library. Chichester, UK: John Wiley & Sons, Ltd, 2004.
Record: 390]
Aldridge 2002
Thaut 1988 {published data only} Aldridge D, Fachner J. Music Therapy World Info-CD ROM IV.
Thaut MH. Measuring musical responsiveness in autistic Witten-Herdecke: University Witten-Herdecke, 2002.
children: a comparative analysis of improvised musical
tone sequences of autistic, normal, and mentally retarded Alvin 1991
individuals. Journal of Autism and Developmental Disorders Alvin J, Warwick A. Music therapy for the autistic child. 2nd
1988;18(4):561-71. [MEDLINE: PBS Record: 380] Edition. Oxford: Oxford University Press, 1991.
Toolan 1994 {published data only} AMTA 1999
Toolan PG, Coleman SY. Music therapy, a description of American Music Therapy Association. Music therapy research:
process: Engagement and avoidance in five people with Quantitative and qualitative foundations CD-ROM I. 1964-1998.
learning disabilities. Journal of Intellectual Disability Research Silver Spring, MD: American Music Therapy Association, 1999.
APA 1994
Watson 1979 {published data only}
American Psychiatric Association. Diagnostic and Statistical
Watson D. Music as reinforcement in increasing spontaneous Manual of Mental Disorders (DSM-IV). 4th Edition. Washington,
speech among autistic children. Missouri Journal of Research in DC: American Psychiatric Association, 1994.
Music Education 1979;4(3):8-20.
APA 2000
Wimpory 1995 {published data only}
American Psychiatric Association. Diagnostic and Statistical
Wimpory D, Chadwick P, Nash S. Brief report: Musical Manual of Mental Disorders (DSM-IV-TR). 4th Edition.
interaction therapy for children with autism: An evaluative Washington, DC: American Psychiatric Association, 2000.
case study with two-year follow-up. Journal of Autism and
Developmental Disorders 1995;25(5):541-552. [MEDLINE: PBS Asperger 1979
Record: 1200; 0162-3257] Asperger H. Problems of infantile autism. Communication
1979;13(3):45-52.
References to studies awaiting assessment Ball 2004
Laird 1997 {published data only} Ball CM. Music therapy for children with autistic spectrum
Laird PD. The effect of music on cognitive/communicative disorder. Bazian Ltd 2004, issue November 11.
skills with students diagnosed with autism, autistic like

Informed decisions.
Baron-Cohen 1995 Ozonoff 2005

Baron-Cohen S. Mindblindedness: An Essay on Autism and Ozonoff S, Goodlin-Jones BL, Solomon M. Evidence-based
Theory of Mind. Cambridge, Massachusetts: The MIT Press, assessment of autism spectrum disorders in children and
1995. adolescents. J Clin Child Adolesc Psychol 2005;34(3):523-40.
Bruscia 1998 Rolvsjord 2005

Bruscia KE. Defining music therapy. 2nd Edition. Gilsum, NH: Rolvsjord R, Gold C, Stige B. Research rigour and therapeutic
Barcelona Publishers, 1998. flexibility: Rationale for a therapy manual developed for a
randomised controlled trial. Nordic Journal of Music Therapy
Chakrabarti 2001 2005;14(1):15-32.
Chakrabarti S, Fombonne E. Pervasive developmental disorders
in preschool children. JAMA 2001;285(24):3093-9. Schumacher 1999a
Schumacher K. Musiktherapie und Säuglingsforschung.
Cohen 1988 Frankfurt am Main: Peter Lang, 1999.
Cohen J. Statistical power analysis for the behavioral sciences.
2nd Edition. Hillsdale, NJ: Lawrence Erlbaum, 1988. Schumacher 1999b
Schumacher K, Calvet-Kruppa C. The AQR - an analysis system
Diggle 2002 to evaluate the quality of relationship during music therapy.
Diggle T, McConachie H R, Randle VRL. Parent-mediated early Nordic Journal of Music Therapy 1999;8(2):188-191.
intervention for young children with autism spectrum disorder
(Cochrane Review). Cochrane Database of Systematic Reviews Stern 1985
2003, Issue 1. Stern DN. The interpersonal world of the infant: A view from
psychoanalysis and developmental psychology. New York: Basic
Elbourne 2002 Books, 1985.
Elbourne DR, Altman DA, Higgins JPT, Curtin F, Worthington HV,
Vail A. Meta-analyses involving cross-over trials: Methodological Stern 1989
issues. International Journal of Epidemiology 2002;31:140-149. Stern DN. The representation of relational patterns:
Developmental considerations. In: Sameroff AJ, Emde RN
Fombonne 1999 editor(s). Relationship disturbances in early childhood. New
Fombonne E. The epidemiology of autism: A review. York: Basic Books, 1989.
Psychological Medicine 1999;29(4):769-786.
Trevarthen 1999a
Gold 2004 Trevarthen C. Musicality and the intrinsic motive
Gold C. The use of effect sizes in music therapy research. Music pulse: Evidence from human psychobiology and infant
Therapy Perspectives 2004;22(2):91-95. communication. Musicae Scientiae 1999;Special issue
1999-2000:155-215.
Gold 2005
Gold C, Heldal TO, Dahle T, Wigram T. Music therapy for Trevarthen 1999b
schizophrenia and schizophrenia-like illnesses. Cochrane Trevarthen C. How music heals. In: Wigram T, De Backer
Database of Systematic Reviews 2005, Issue 2. J editor(s). Clinical Applications of Music Therapy in
Developmental Disability, Paediatrics and Neurology. London:
Higgins 2002 Jessica Kingsley Publishers, 1999:7-10.
Higgins JPT, Thompson SG. Quantifying heterogeneity in a
meta-analysis. Statistics in Medicine 2002;21:1539-1558. Whipple 2004
Whipple J. Music in intervention for children and adolescents
Howlin 1998 with autism: A meta-analysis. Journal of Music Therapy
Howlin P. Children with Autism and Asperger Syndrome: A 2004;41(2):90-106.
Guide for Practitioners and Carers. Chichester, New York: Wiley,
1998. WHO 1992
World Health Organization. International Statistical
Kendall 2004 Classification of Diseases and Related Health Problems. 10th
Kendall P, Holmbeck G, Verduin T. Methodology, design, and Edition. Geneva: World Health Organization, 1992.
evaluation in psychotherapy research. In: Lambert MJ editor(s).
Bergin and Garfield's Handbook of Psychotherapy and Behavior Wigram 2002
Change. 5th Edition. New York, NY: Wiley, 2004:16-43. Wigram T. Indications in music therapy: Evidence from
assessment that can identify the expectations of music therapy
Lambert 2004 as a treatment for autistic spectrum disorder (ASD). Meeting the
Lambert MJ, Ogles BM. The efficacy and effectiveness of challenge of evidence based practice. British Journal of Music
psychotherapy. In: Lambert MJ editor(s). Bergin and Garfield's Therapy 2002;16(1):5-28.
Handbook of Psychotherapy and Behavior Change. 5th Edition.
New York, NY: Wiley, 2004:139-193.
Informed decisions.
Wigram 2006 developmental disorders. 2nd Edition. New York: John Wiley,
Wigram T, Gold C. Music therapy in the assessment and 1997:148-172.
treatment of autistic spectrum disorder: Clinical application
Wing 2002
and research evidence. Child: Care, Health and Development In
press. Wing L, Leekam SR, Libby SJ, Gould J, Larcombe M. The
diagnostic interview for social and communication disorders:
Wing 1997 background, inter-rater reliability and clinical use. Journal of
Wing L. Syndromes of autism and atypical development. In: Child Psychology and Psychiatry 2002;43:307-325.
Cohen D, Volkmar F editor(s). Handook of autism and pervasive
* Indicates the major publication for the study
CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Brownell 2002
Methods Allocation: quasi-randomised, possibly randomised ('counterbalanced')
Blindness: independent assessor (teacher), blinding not reported.
Duration: 4 weeks.
Design: crossover.
Participants Diagnosis: autism.

N=4.
Age: range 6-9.
Sex: 4 M, 0 F.
Setting: elementary school.
Interventions 1. structured receptive MT (songs with social stories). 5 individual daily sessions. N=4.
2. structured receptive "story therapy" (reading of social stories). 5 individual daily sessions. N=4.
3. no intervention. 2 x 5 days. N=4.
Outcomes Repetitive behaviours outside therapy sessions (in classroom). Inter-rater reliability 0.86 to 0.94
Notes
Risk of bias
Bias Authors' judgement Support for judgement
Allocation concealment? Unclear risk B - Unclear
Buday 1995
Methods Allocation: randomised.
Blindness: assessor blinded to the nature of the hypothesis and to treatment condition.
Duration: 2 weeks.
Design: crossover.
Participants Diagnosis: autism

N=10.
Age: range 4-9.
Sex: 8 M, 2 F.
Setting: public school.
Interventions 1. structured receptive MT (songs used to teach signs). 5 individual sessions. N=10.

Informed decisions.
Buday 1995 (Continued)

2. "rhythm therapy" (rhythmic speech used to teach signs). 5 individual sessions. N=10.
Outcomes Imitating behaviour in sessions (rating of a video recording with sound turned off to ensure blinding of
raters; inter-rater agreement 98%).
a) sign imitation
b) speech imitation
Notes
Risk of bias
Farmer 2003
Methods Allocation: randomised.
Blindness: not known.
Duration: 5 days.
Design: parallel group.
Participants Diagnosis: autism.

N=10.
Age: range 2-5.
Sex: 9 M, 1 F.
Setting: homes and therapy centres.
Interventions 1. MT sessions (combined active and receptive: guitar playing, songs. N=5
2. Placebo (no music) sessions. N=5.
Mostly individual sessions of 20 minutes
Outcomes responses within sessions (inter-rater agreement 91%).

a) verbal responses
b) gestural responses
Notes
Risk of bias
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Applebaum 1979 not intervention study (assessment)
Bettison 1996 not MT (AIT/only music listening)

Informed decisions.
Blackstock 1978 not intervention study (assessment)
Bonnel 2003 not intervention study (assessment)
Brown 1994 not RCT/CCT (case series)
Brown 2003 not intervention study (assessment)
Bruscia 1982 not RCT/CCT (case study)
Carroll 1983 not MT (only sung instructions)
Chilcote-Doner 1982 not MT (rhythmic strobe and drumbeat)
Clauss 1994 not RCT/CCT (case series, ABACA design)
Dawson 1998 not intervention study (assessment)
Diez 1989 not intervention study (assessment)
Edelson 1999 not MT (AIT/only music listening)
Edgerton 1994 not RCT/CCT (case series)
Frissell 2001 not intervention study (assessment)
Goldstein 1964 not RCT/CCT (case study)
Gore 2002 not usable (unclear outcome measure)
Griggs 1997 not RCT/CCT (case study), not intervention study (assessment)
Hadsell 1988 not RCT/CCT (case series), not ASD (Rett syndrome)
Hairston 1990 not RCT/CCT (case series)
Heaton 1999 not intervention study (assessment)
Heaton 2003 not intervention study (assessment)
Kolko 1980 not intervention study (assessment)
Krauss 1982 not RCT/CCT (case series), not ASD (apraxia, language delay)
Lee 2004 not RCT/CCT (case series)
Litchman 1976 not MT (listening to recorded nursery rhymes)
Ma 2001 not RCT/CCT (case series)
Mahlberg 1973 not RCT/CCT (case study)
Miller 1979 not RCT/CCT (case study)
Mottron 2000 not intervention study (assessment)

Informed decisions.
Mudford 2000 not MT (AIT/only music listening)
O'Connell 1974 not RCT/CCT (case study)
O'Loughlin 2000 not RCT/CCT - includes 3 case series where all received the same treatment (no. 1, 3, 4) and 1 case
series with an ABA design (no. 2).
Pasiali 2004 not RCT/CCT (case series, ABAB design)
Rao 2001 not MT (headphones with vs. without music)
Saperston 1973 not RCT/CCT (case study)
Schmidt 1976 not RCT/CCT (case series, AB design)
Starr 1998 not RCT/CCT (case series)
Staum 1984 not RCT/CCT (case study)
Stevens 1969 not RCT/CCT (case series)
Thaut 1987 not intervention study (assessment)
Thaut 1988 not intervention study (assessment)
Toolan 1994 not RCT/CCT (case series)
Watson 1979 not RCT/CCT (case series, ABCA design)
Wimpory 1995 not RCT/CCT (case study)
Characteristics of ongoing studies [ordered by study ID]
Kim 2005
Trial name or title Joint attention and attunement in improvisational music therapy with the autistic child
Methods
Participants Diagnosis: Autism (diagnostic consents by two child psychiatrists).

Exclusion criteria: diagnostic comorbidity; previous experience of music therapy or play therapy.
N=15.
Age: range 3-7.
Interventions 1. Music therapy: 12 individual sessions over 3 months.

2. Free play: 12 sessions over 3 months.
Psychiatric medication allowed to change during trial period.
Crossover RCT.
Outcomes Pre/post (at baseline, crossover, and end of trial):

1. Mother-child free play interaction at home (MPI profile). Pre/post
2. Early Social Communication Scale (ESCS, Mundy et al, 1986, 2003),
3. Pervasive developmental disorder behavior inventory (PDDBI-C, Cohen & Subhalter, 1999)
Within sessions:

Informed decisions.
Kim 2005 (Continued)

4. Treatment measurement: Every session videotaped. Treatment manual used.
Starting date 2003
Contact information Jinah Kim, Aalborg University, Denmark. Email: kim_jinah@hotmail.com
Notes
DATA AND ANALYSES
Comparison 1. Music therapy vs. "placebo" therapy
Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants
1 Communicative skills: gestural 2 SMD (Fixed, 95% CI) 0.50 [0.22, 0.79]
2 Communicative skills: verbal 2 SMD (Fixed, 95% CI) 0.36 [0.15, 0.57]
3 Behavioural problems (end of therapy) 1 SMD (Fixed, 95% CI) -0.14 [-0.42, 0.14]
4 Behavioural problems (using all measure- 1 SMD (Fixed, 95% CI) -0.24 [-0.45, -0.03]
ments from 2nd day on)
Analysis 1.1. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 1 Communicative skills: gestural.
Study or subgroup Treatment Control SMD Std. Mean Difference Weight Std. Mean Difference
N N (SE) IV, Fixed, 95% CI IV, Fixed, 95% CI
Buday 1995 10 10 0.5 (0.15) 95.77% 0.48[0.18,0.77]
Farmer 2003 5 5 1.2 (0.716) 4.23% 1.17[-0.24,2.57]
Total (95% CI) 100% 0.5[0.22,0.79]

Heterogeneity: Tau2=0; Chi2=0.89, df=1(P=0.34); I2=0%
Test for overall effect: Z=3.43(P=0)
Favours "placebo" -4 -2 0 2 4 Favours MT
Analysis 1.2. Comparison 1 Music therapy vs. "placebo" therapy, Outcome 2 Communicative skills: verbal.
Buday 1995 10 10 0.3 (0.11) 97.42% 0.35[0.13,0.56]
Farmer 2003 5 5 0.8 (0.674) 2.58% 0.81[-0.51,2.13]
Total (95% CI) 100% 0.36[0.15,0.57]

Informed decisions.
Heterogeneity: Tau2=0; Chi2=0.45, df=1(P=0.5); I2=0%
Test for overall effect: Z=3.32(P=0)
Analysis 1.3. Comparison 1 Music therapy vs. "placebo"

therapy, Outcome 3 Behavioural problems (end of therapy).
Brownell 2002 4 4 -0.1 (0.141) 100% -0.14[-0.42,0.14]
Total (95% CI) 100% -0.14[-0.42,0.14]

Heterogeneity: Not applicable
Test for overall effect: Z=1(P=0.32)
Favours MT -1 -0.5 0 0.5 1 Favours "placebo"
Analysis 1.4. Comparison 1 Music therapy vs. "placebo" therapy, Outcome

4 Behavioural problems (using all measurements from 2nd day on).
Brownell 2002 4 4 -0.2 (0.109) 100% -0.24[-0.45,-0.03]
Total (95% CI) 100% -0.24[-0.45,-0.03]

Heterogeneity: Not applicable
Test for overall effect: Z=2.2(P=0.03)
Favours MT -1 -0.5 0 0.5 1 Favours "placebo"
WHAT'S NEW
Date Event Description
5 November 2009 Amended Minor edit in background.
HISTORY
Protocol first published: Issue 3, 2003
Review first published: Issue 2, 2006
10 November 2008 Amended Converted to new review format.

Informed decisions.
21 February 2006 Amended Minor update
29 January 2006 New citation required and conclusions Substantive amendment

have changed
CONTRIBUTIONS OF AUTHORS
CG - designed the protocol, co-ordinated the reviewing, searched for studies, extracted and analysed data, and wrote the report.
TW - extracted and analysed data, helped with writing the protocol and the report.
CE - helped with data extraction and analysis and with writing the report.
DECLARATIONS OF INTEREST
The authors of this review are clinically trained music therapists.
SOURCES OF SUPPORT
Internal sources
• Sogn og Fjordane University College, Norway.
• Aalborg University, Denmark.
External sources
• The Research Council of Norway, Norway.
INDEX TERMS
Medical Subject Headings (MeSH)

Autistic Disorder [*rehabilitation]; Child Development Disorders, Pervasive [*rehabilitation]; Communication; Music Therapy
[*methods]; Randomized Controlled Trials as Topic; Time Factors
MeSH check words

Child; Humans


Gold Et Al-2006-Cochrane Database of Systematic Reviews

Uploaded by

Copyright:

Available Formats

Gold Et Al-2006-Cochrane Database of Systematic Reviews

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gold Et Al-2006-Cochrane Database of Systematic Reviews

Uploaded by

Copyright:

Available Formats

Cochrane

Music therapy for autistic spectrum disorder (Review)

Gold C, Wigram T, Elefant C

Gold C, Wigram T, Elefant C.

Music therapy for autistic spectrum disorder (Review)

Music therapy for autistic spectrum disorder (Review) i

Music therapy for autistic spectrum disorder

Christian Gold1, Tony Wigram2, Cochavit Elefant3

Editorial group: Cochrane Developmental, Psychosocial and Learning Problems Group.

Data collection and analysis

Music therapy for autistic spectrum disorder (Review) 1

PLAIN LANGUAGE SUMMARY

Music therapy for people with autistic spectrum disorder

Music therapy for autistic spectrum disorder (Review) 2

BACKGROUND The clinical picture varies because individuals have variable

Why it is important to do this review -Activity level

RESULTS 4.5.1 Music therapy

Music therapy for autistic spectrum disorder (Review) 6

4.6 Outcome measures 4.2 Data reporting and analysis

Music therapy for autistic spectrum disorder (Review) 7

included communicative skills and behavioural problems, and Control conditions

Music therapy for autistic spectrum disorder (Review) 8

Music therapy for autistic spectrum disorder (Review) 9

Dawson 1998 {published data only}

Music therapy for autistic spectrum disorder (Review) 10

Gore 2002 {published data only} Ma 2001 {published data only}

Music therapy for autistic spectrum disorder (Review) 11

Music therapy for autistic spectrum disorder (Review) 12

Baron-Cohen 1995 Ozonoff 2005

Bruscia 1998 Rolvsjord 2005

Characteristics of included studies [ordered by study ID]

Participants Diagnosis: autism.

Bias Authors' judgement Support for judgement

Allocation concealment? Unclear risk B - Unclear

Participants Diagnosis: autism

Music therapy for autistic spectrum disorder (Review) 14

Buday 1995 (Continued)

Bias Authors' judgement Support for judgement

Allocation concealment? Unclear risk B - Unclear

Participants Diagnosis: autism.

Outcomes responses within sessions (inter-rater agreement 91%).

Bias Authors' judgement Support for judgement

Allocation concealment? Unclear risk B - Unclear

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion

Applebaum 1979 not intervention study (assessment)

Bettison 1996 not MT (AIT/only music listening)

Music therapy for autistic spectrum disorder (Review) 15

Study Reason for exclusion

Blackstock 1978 not intervention study (assessment)

Bonnel 2003 not intervention study (assessment)

Brown 1994 not RCT/CCT (case series)

Brown 2003 not intervention study (assessment)

Bruscia 1982 not RCT/CCT (case study)

Carroll 1983 not MT (only sung instructions)

Chilcote-Doner 1982 not MT (rhythmic strobe and drumbeat)

Clauss 1994 not RCT/CCT (case series, ABACA design)

Dawson 1998 not intervention study (assessment)

Diez 1989 not intervention study (assessment)

Edelson 1999 not MT (AIT/only music listening)