The Relationship Between Pitch Discrimination and Enhanced Local Processing of Melodies In Autism

Background:  Enhanced pitch discrimination is one of the most robust findings within the perceptual domain in autism (Bonnel et al. 2003; O’Riordan & Passeti 2006; Heaton et al. 2008; Jones et al. 2009; Bonnel et al. 2010). Autistics have also displayed a local processing bias which may allow for more efficient detection of subtle changes in melodies under conditions where local analysis is advantageous (Mottron et al. 2000 vs. Heaton et al. 2005). These two findings suggest that enhanced auditory perception is manifested at both lower- (pitch) and higher-levels (melody) of processing in autism. However, since they were found independently, the relationship between enhanced pitch and melody discrimination remains unknown.

Objectives:  The aim of this study is to investigate the relationship between enhanced pitch discrimination and improved local processing of simple melodies in autism. Furthermore, it will serve as a replication study for the enhanced local processing bias found by Mottron et al. (2000).

Methods:  Pitch and melody discrimination abilities in 21 typical and 21 autistic individuals were measured. Autistic participants were diagnosed using DSM, ADI and ADOS standards and were slightly older than their typical counterparts (mean age: AUT=23.19yrs, TYP=20.0yrs; p=0.04); participants were matched on Wechsler IQ (FSIQ range: 81-126). There were two tasks. In a pitch discrimination task, participants were asked to indicate whether two pure tones were the same or different. Discrimination thresholds for each of three frequencies (500Hz, 1000Hz, 1500Hz) were found using a PEST adaptive staircase procedure. In a melody discrimination task investigating local-global processing abilities (based on Mottron et al. 2000), participants were asked to indicate whether two melodies were the same or different. This task included a “different contour” (DC) condition, in which change of a single tone affected the qualities of contour and an “identical contour” (IC) condition, in which this change did not affect the qualities of contour. Successful performance on the IC condition required better local processing abilities.

Results:  Autistics had significantly lower pitch discrimination thresholds than typical controls when performance was averaged across frequency conditions (p=0.042).  The results found in Mottron et al. (2000) were not replicated. When compared to the typical group, autistics did not display enhanced local processing through an enhanced performance on the IC condition (p=0.305). Nevertheless, group means, where higher scores reflect better performances, were in the expected direction (AUT=33.33>TYP=21.42). Preliminary correlation analyses showed a significant relationship between enhanced pitch perception and improved performance in the IC condition (AUT:r=-0.492, p=0.038; TYP:r=-0.506, p=0.019). Enhanced pitch perception was related to improved performance on the easier DC condition for typical controls only (r=-0.508, p=0.019).

Conclusions:  Results support previous findings of enhanced pitch perception in autistics. However, an enhanced local processing bias during melody processing was not supported. Although statistical power remains limited, it appears that for typical individuals, enhanced discrimination of pitch is associated with a better ability for discriminating between melodies, regardless of the level of difficulty. On the contrary, in autistics, enhanced pitch discrimination abilities were only related to the more difficult, IC melody discrimination condition.