Autism spectrum disorder (ASD) is a developmental impairment characterized by persistent deficits in social communication and interactions. The prevalence figures have increased rapidly in recent years due to the expansion of diagnostic criteria and increased public awareness. The clinical presentations of ASD vary to a large extent, and approximately 56% of children with ASD possess below average intellectual ability (IQ < 85). The intellectual, verbal, and social ability in this population may greatly influence intervention outcomes. The social development of children with ASD and comorbid intellectual disability (ID) is not well understood, and how children with ASD/ID respond to social skill interventions remains to be investigated.

Musical elements are a part of various behavioral interventions for ASD, however, the effects of music as interventions for ASD individuals have not been comprehensively examined in Hong Kong. The proposed study will address limited research evidence on music therapy as an intervention for social functioning in children with ASD with mild to borderline ID. Music therapy is a systematic process of intervention, wherein a therapist helps clients promote their health by using musical experience and relationships that develop through them. In particular, the investigators will examine whether using music therapy in social skill intervention provides additional benefits relative to non-musical intervention in a 12-week randomized controlled trial. Pre-treatment neural response of electroencephalograms (EEG) to social scenes will be used to predict the outcomes of social skill interventions, whereas EEG responses to music will be used to predict the effectiveness of musical social skill intervention. If correlation is found, then the long-term goal is to develop individualized intervention based on pre-treatment markers to maximize treatment efficacy.

Aims and hypotheses:

1. Is social skill intervention using music therapy more effective in enhancing social interaction than non-musical social skill training for children with ASD and co-occurring mild/borderline ID?
2. Participants with enhanced neural response to social scenes relative to baseline would be more responsive to social skill interventions.
3. Participants with enhanced neural response to preferred music relative to baseline would be more responsive to music therapy targeting social skill intervention.

Autism spectrum disorder (ASD) is a developmental impairment characterized by persistent deficits in social communication and interactions coupled with restricted and repetitive patterns of behaviors. The symptoms of ASD appear in early childhood. Given that no known cure or effective pharmacological intervention for ASD is available, behavioral interventions that target autism symptoms is the main strategy to support individuals with ASD to be integrated into school and the society. The choice to examine music in social skill intervention is motivated by accruing evidence of its potential for therapeutic use for various clinical populations. For ASD intervention, music has been explored as a medium to increase social skills in the field of music therapy for some years. Children with ASD and ID are often described as preoccupied in their own world and may be highly passive and unresponsive in social situations. The low verbal ability of these children is one limitation to their participation given that social interactions often involve verbal exchanges. Music may become a conduit for children whose natural affinity to social interactions is impaired. Group music making and improvisations encourage initiation and turn-taking behavior in a non-verbal situation and may be an alternative avenue for social practice. Children with ASD can learn to tolerate the presence of and physical contact with other people, distinguish between oneself and others, and practice social behavior in music therapy group activities.

Although these claims are supported by data, no conclusive evidence is available that music therapy is an effective social skill intervention for children with ASD. One hypothesis that has not been tested is that success in social skill interventions may depend on individual differences. Autistic symptoms and abilities vary widely among children with ASD, arising both from different developmental speeds and specific interests. Some individuals may possess deficits in the areas of restricted interests or repetitive behavior, which results in poor social skills; however, they are accepting of social interactions. The willingness to seek out social experiences is the foundation to social skill development because without initial interests in social behavior, the motivation to attend to the explicit training and modeling is low. In such cases of social avoidance, cultivation of social interests through other strategies, such as music making, may be effective. Narrowing down the types of children who may respond well to a particular type of intervention can facilitate timely and efficient treatment provision. In this study, the investigators will test quantitative electroencephalogram (qEEG) signals in response to music and social scenes as predictive measures of treatment outcomes.

Spontaneous electrophysiological neural activities at rest or when engaged in task can be measured by qEEG. Spectral analysis is used to decompose qEEG signals across several minutes into different frequency bands, which are associated with specific functions. The most consistent pattern of findings in the literature is that individuals with ASD show a U-shaped pattern of spectral power relative to controls without ASD; that is, excessive power is observed at low-frequency (delta, 1-3Hz and theta, 4-7Hz) and high-frequency (beta, 13-35Hz and gamma, >35Hz) bands, but reduced power is detected in middle-range frequency band (alpha, 8-12Hz). The pattern was found with wide topographic distribution, which suggested abnormalities across multiple brain regions. This finding was reported for children in different age groups and for individuals with or without comorbid ID. High-frequency bands have been associated with emotional responses and emotion recognition and may be linked to such deficits in individuals with ASD. Neural patterns that deviate from age-matched controls to a greater extent may suggest more severe behavioral symptoms.

Frontal alpha asymmetry (FAA) is typically investigated in relation to emotional response and motivations, both in clinical and normal populations. Left-lateralized or left-dominant brain activity has been linked to an approach system where an individual experiences positive emotions and motivations. In contrast, right-lateralized frontal activity may reflect negative emotions and intention to withdraw. Reduced alpha power in the left frontal area of the scalp in individuals with ASD has also been reported. The modulation of FAA has been demonstrated in neurofeedback training in individuals with ASD to activate their imitation behaviors; so this measure is hypothesized to predict and be responsive to social skill interventions. High-functioning individuals with ASD show intact emotional processing when listening to music, although they demonstrate distinct neural patterns relative to neurotypical adults, which is interpreted as increased cognitive load and physiological arousal. Individual preference for music varies within the target group with ASD/ID and may predict how they respond to social skill intervention using music therapy. Although self-report rating scales exist for reporting music experience and preference, qEEG measures may be superior because they index automatic responses to musical stimuli that the target population of children with ASD/ID may find difficult to understand or express.

EEG data can be collected with low-cost, commercially available products and take only 5-10 min to set up. Children with ASD and ID are only required to sit still for a few minutes because no explicit task response is needed. With the relative ease of data collection, this procedure may be included in routine clinical check-ups. The analyses of these data can help researchers and clinicians to understand how neural responses in children with ASD are related to the clinical presentation of ASD symptoms and their social deficits. Such data may be further compared with those of neurotypical children to elucidate the mechanisms of social functioning and to improve social skill interventions. With positive validation, these neural markers can then be used to prescribe the type of intervention that will most likely succeed.

Inclusion Criteria:

• A formal clinical diagnosis of ASD
• An assessed overall and verbal IQ of 50-84 by a certified clinician
• Children who report hypersensitivity to sounds may be included with consent but will be withdrawn immediately if adverse reactions are observed

Exclusion Criteria:

• Children who exhibit severe physical or sensory disabilities (e.g., deafness) that may limit their participation in either intervention
• Children with other neurodevelopmental, psychiatric, or neurological comorbidities or are on prescribed psychiatric medication will also be excluded from the study