Investigating Reward Processing in the Longitudinal European Autism Project (LEAP) Cohort with Fmri

Background:

Studying reward processing in autism spectrum disorder (ASD) has so far yielded inconsistent results. Some previous studies have proposed impaired reward processing in ASD specifically regarding social reward, while others proposed a general reward processing deficit across social and monetary reward types, possibly differentiating ASD from other disorders. However, current evidence does not clearly support either of these hypotheses. This likely reflects the heterogeneity of the disorder, which has been insufficiently addressed due to small sample size, along with the unknown reliability of the measures.

Objectives:

To address these issues, the Longitudinal European Autism Project (LEAP) is using functional magnetic resonance imaging (fMRI) during reward processing (for general aspects of multi-site fMRI assessment and effects of all applied tasks see abstract by Moessnang et al.) alongside an extensive cognitive and neuropsychological task battery in a large cohort of patients with ASD as well as typically developing (TD) subjects.

Methods:

A social and a monetary reward processing task were designed, as part of a comprehensive fMRI task battery which allows for the inclusion of participants across a broad age and ability range (including those with mild intellectual disabilities). FMRI was performed at six European centers following standard operations procedures. A total of 635 subjects completed the social and 565 subjects completed the monetary reward task, with an overlap of 562 subjects completing both reward processing tasks. Subjects were aged 6-30 years, with 53% ASD and 69% male participants. Baseline assessment was performed between April 2014 and September 2016. First pass analyses have been performed using standard processing routines implemented in SPM12 (http://www.fil.ion.ucl.ac.uk/spm/).

Results:

Preliminary statistical analyses based on the General Linear Model revealed robust activation of the network of interest, while simple case-control differences did not pass the significance threshold (p<0.05, family-wise error corrected across the whole brain). However, region of interest (ROI) analysis of the ventral striatum yielded hypoactivation of the right ventral striatum in ASD patients during monetary reward anticipation (F(1,430)=5.447, p=.020) but not during social reward anticipation (F(1,501)=2.249, p=.134).

Conclusions:  The LEAP cohort represents the largest European task-based fMRI data set on autism. The employed tasks proved reliable and successfully engaged functional activation of the reward processing circuit both during monetary and social reward anticipation. In line with some previous smaller studies, ROI analysis of ventral striatal activation yielded a significant difference between ASD patients and TD subjects during monetary reward, while there were no effects on the whole brain level. However, further thorough investigation of different sources of variance, including direct social/monetary task comparisons, between-site differences, symptom profiles and developmental effects, is required. Further, the analysis of functional activation during reward receipt, as well as connectivity-based measures may aid in deeper understanding of the autism phenotype.