Pitch Direction Perception Predicts the Ability to Detect Local Pitch Structure in Autism and Typical Development

Background:  Most individuals with Autism Spectrum Disorders (ASD) have atypical sensory perception, particularly in the auditory domain. Some studies have reported enhanced low-level pitch discrimination (Bonnel, 2010) and higher-level processing of local pitch structure on global-local tasks (Heaton, 2005). However, auditory perception in ASD is understudied and results are inconsistent. It is unclear how low and higher-levels of auditory perception are related in ASD or typical development (TD), nor how these skills may change with development.

Objectives:  The aims of this research were to test: 1) for performance differences between children with ASD and TD controls on both low-level pitch processing (pitch direction) and higher-level pitch processing (global-local pitch) tasks, 2) the relationship between low and higher-level auditory perception, and 3) the effect of age on task performance. 

Methods:  Participants were 17 children with high-functioning ASD and 19 TD children who had no neurological or psychiatric history. Participants were recruited from the NeuroDevNet ASD Project (Zwaigenbaum et al, 2011) and groups were matched on age (mean age 13.3 years, range 9-18 years) and IQ (mean IQ 113.5, SD 13.6). In the low-level pitch direction task (Foster et al., 2014), participants heard two tones of different frequencies on each trial and judged whether the pitch rose or fell. In a higher-level global-local task (Ouimet et al., 2012), stimuli consisted of 9-tone melodies, each containing three triplet (3-tone) sequences. Participants discriminated between ascending and descending pitch direction at the global (across triplets) and/or local (within triplets) level. The relationship between performance (percent accuracy) on these two tasks and the effect of age was assessed using linear regression.

Results:  There were no significant group differences between ASD and TD in terms of overall performance accuracy on either of the pitch tasks. Overall accuracy on the low-level pitch discrimination task predicted overall accuracy on the higher-level global-local pitch task across all participants (P<0.001), but was particularly driven by performance on local pitch judgements (P<0.001; see Figure 1). The TD and ASD groups both showed a similar positive relationship between performance profiles on the low versus high-level pitch tasks. However, a trend suggests that there is a tighter coupling between low and high-level pitch task performance in ASD versus TD. There was also a trend toward a positive main effect of age on the pitch direction task performance only (and not the global-local task) across all participants.

Conclusions:  The present findings suggest that: 1) low-level pitch direction ability strongly predicts performance of higher-level global-local pitch perception in general, but most prominently for local pitch judgments in ASD compared to TD, 2) pitch direction ability improves with age similarly in TD and ASD, but global-local perception is constant, and 3) superior abilities in ASD in low and high-level pitch tasks may be specific to particular task paradigms, or subgroups within the autism spectrum. The study of auditory perception in ASD serves as a complementary lens to more symptom-based studies, to better understand individual differences and to refine ASD endophenotypes.