Reading the Mind in the Eyes: Examining a Large Multicentre Dataset

Thursday, May 11, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
R. Holt1, M. C. Lai2, H. L. Hayward3, E. Loth3, A. N. Ruigrok1, M. V. Lombardo4,5, B. Auyeung6, D. G. Murphy7 and S. Baron-Cohen1, (1)University of Cambridge, Cambridge, United Kingdom, (2)Psychiatry, University of Toronto, Toronto, ON, CANADA, (3)Forensic and Neurodevelopmental Sciences, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom, (4)University of Cambridge, Sacramento, CA, (5)University of Cyprus, Nicosia, Cyprus, (6)University of Edinburgh, Edinburgh, United Kingdom, (7)Department of Forensic and Neurodevelopmental Sciences, and the Sackler Institute for Translational Neurodevelopment, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom
Background:  Difficulties with aspects of social cognitive processing comprise a core symptom of Autism Spectrum Conditions (henceforth autism). Previous studies have identified differences in performance on the ‘Reading the Mind in the Eyes’ test (Eyes test) in individuals with autism (Baron-Cohen et al 1997, 2001 (JCPP), 2015 (PLoS ONE), Lai et al 2012 (PLoS ONE), Karlan et al 2008 (JADD), Lombardo et al., in press, Scientific Reports).

Objectives:  Here we present data from the Eyes test collected as part of the EU-AIMS Longitudinal European Autism Project (LEAP), pooling data from 6 study sites. This study aimed to examine: 1) Categorical group differences (autism vs. controls and males vs. females) in performance (accuracy and reaction time). 2) The interaction between diagnosis and sex. 3) The effects of cross-sectional age on task performance. These differences were first considered in the complete sample, and secondly in separate age cohorts of children (6-11 years), adolescents (12-17 years) and adults (18-30 years).

Methods:  Participants with a diagnosis of autism aged 6-30 years (N=329; 237 males, 92 females) and age matched typically developing controls (N=264; 174 males, 90 females) completed the Eyes task as part of a battery of cognitive tasks. Three versions of the task were administered using age appropriate mental state words and were translated into the native language of the different participating countries.

Results:  Significant differences were identified for task accuracy between the autism and control groups (F(1,588)=24.24, p<.001, r=0.2), with fewer correct responses seen in the autism group. In addition, significant sex differences were identified (F(1,588)=12.09, p=.001, r=.14) with more correct responses seen in females. There was no significant interaction between diagnosis and sex and no differences in reaction time between the groups. A significant correlation was identified between accuracy and age, with an improved performance seen with increasing age (r=.348, p<.001). When examining the age groups separately, significant differences were identified between the autism and control groups in both the adult (F(1,228)=6.15, p=.014, r=.16) and adolescent (F(2,203)=17.04, p<.001, r=.28) samples, however there was no significant difference in the child group. Significant correlations between task accuracy and age were identified in the child (r=.279, p<.001) and adolescent groups (r=.234, p=.001) but were not seen in the adult group.

Conclusions:  Performance differences were identified on this task, indicating impairment in complex emotion recognition associated with mental state attribution in individuals with autism. The finding of group differences is consistent with previous work, however in this case no interaction between diagnosis and sex was identified. Task performance was significantly associated with cross-sectional age, demonstrating improvements on the task with increasing age in children and adolescents but not in adults. Future work will examine the longitudinal data from this task to further examine the effect of development on task performance. Future work will also utilize an item level clustering approach to examining the data.