Decreased Intrinsic Connectivity Between Motion Processing Areas in ASD

Background: A dysfunction in biological motion processing has been suggested to contribute to social deficits in autism spectrum disorder (ASD). Emerging evidence suggests a specific impairment in orienting towards biological motion. This effect has been repeatedly reported for autistic children and adolescents, however results for autistic adults are more ambiguous. Some neuro-imaging studies focused on the neural correlate of motion detection and a hypo-activation of the posterior superior temporal sulcus (pSTS) was the most consistent finding. However, so far little work has been done on spontaneous, uninstructed biological motion detection and the connectivity between brain regions involved in biological motion processing in ASD.  

Objectives: Compare networks for biological motion processing in ASD and control subjects during spontaneous, uninstructed motion detection with a special focus on network connectivity.

Methods: 15 high-functioning autistic adults and 15 controls, matched for age, handedness and IQ participated in this study. Behavioral and functional MRI data were collected during a single MRI scanning session. Three hierarchically ordered conditions were utilized: 1) displays of original point light recordings of human full body motion, 2) displays with homogeneous velocity profiles and 3) displays with homogeneous velocity profile and randomized starting position for each trajectory. Importantly, participants were not instructed to attend to the motion displays but to count the occurrences of a randomly appearing red dot. Standard fMRI pre-processing steps were applied and a group comparison in regional BOLD signal strength was executed using a 2x3 mixed-model ANOVA design. Moreover, a dynamic causal modeling (DCM) analysis was conducted using a linear forward model of the regions MT+ -> EBA -> pSTS -> PF, a driving input to MT+ (conditions: 1+2+3), and modulation effects of shape (conditions: 1+2) and biological motion (condition 1) on the intrinsic connections. 

Results: Both groups achieved a high level of performance in detecting the colored dot. The strength of BOLD signal was comparable between both groups in motion processing regions, such as MT+, EBA, pSTS and PF. As a preliminary DCM result, we found a reduced intrinsic connectivity between all the nodes involved in biological motion processing included in our model in the ASD compared to the control group (main effect of group: F(1,28)=6.7, p<0.05, mixed-design ANOVA). The input effect to the DCM model and the modulation effects of shape and biological motion were significant but did not differ between groups. 

Conclusions: In a task that does not require attention to motion, including biological motion, HFA adults recruit the same areas for motion processing as the control participants and show comparable level of BOLD activation. Thus, a strong, automatic processing of motion cues by HFA adults can be concluded. The reduced intrinsic connectivity between the DCM nodes was not specific to biological motion and might indicate reduced baseline connectivity between motion processing areas. The ‘red-dot detection’ task performance was high in both groups indicating that an influence of this distractor on the connectivity pattern is unlikely. Thus, we propose that the reduced intrinsic connectivity might reflect a neural correlate of motion processing deficits in autism.