Underconnectivity of the superior temporal sulcus predicts emotion recognition deficits in autism

Abstract

Neurodevelopmental disconnections have been assumed to cause behavioral alterations in autism spectrum disorders (ASDs). Here, we combined measurements of intrinsic functional connectivity (iFC) from resting-state functional magnetic resonance imaging (fMRI) with task-based fMRI to explore whether altered activity and/or iFC of the right posterior superior temporal sulcus (pSTS) mediates deficits in emotion recognition in ASD. Fifteen adults with ASD and 15 matched-controls underwent resting-state and task-based fMRI, during which participants discriminated emotional states from point light displays (PLDs). Intrinsic FC of the right pSTS was further examined using 584 (278 ASD/306 controls) resting-state data of the Autism Brain Imaging Data Exchange (ABIDE). Participants with ASD were less accurate than controls in recognizing emotional states from PLDs. Analyses revealed pronounced ASD-related reductions both in task-based activity and resting-state iFC of the right pSTS with fronto-parietal areas typically encompassing the action observation network (AON). Notably, pSTS-hypo-activity was related to pSTS-hypo-connectivity, and both measures were predictive of emotion recognition performance with each measure explaining a unique part of the variance. Analyses with the large independent ABIDE dataset replicated reductions in pSTS-iFC to fronto-parietal regions. These findings provide novel evidence that pSTS hypo-activity and hypo-connectivity with the fronto-parietal AON are linked to the social deficits characteristic of ASD.

Keywords: autism spectrum disorders; emotion recognition; functional connectivity; functional magnetic resonance imaging; superior temporal sulcus.

Fig. 1

Participants determined the emotional state of PLDs in which moving white dots reflected the main joints of the human body. (A) The emotional state of the blue-bordered PLD had to be indicated relative to the baseline yellow-bordered PLD (always showing a neutral emotional state). The same PLD stimuli were presented in a four-choice control task matched for cognitive and motor demands. Here, one of the dots in the yellow-bordered PLD briefly changed color to either red or green. Subsequently, participants had to indicate the number of dots that changed into the same color in the blue-bordered PLD (2 in this example). (B) The TD group was more accurate than the ASD group on the emotion recognition task, but not on the control task.

Fig. 2

Task-based brain activity. (A) Brain activity during performance of the emotion recognition test (>fixation) in the ASD and TD groups. (B) Brain activity during performance of the control test (>fixation) in the ASD and TD groups. (C) Brain activations during emotion recognition (>control task) in ASD and TD groups (one-sample t-tests, P < 0.05, corrected).

Fig. 3

Task-based brain activity: group differences and brain–behavior relationships. (A) Between-group comparisons revealed stronger activity in bilateral pSTS in TD controls (small volume, two-sample t-test, P < 0.05, corrected). Only two additional brain regions (left IPL and left MOG) showed a similar group effect when tested at a whole-brain level (whole-brain, two-sample t-test, P < 0.0005, uncorrected), highlighting the spatial specificity of this finding. No clusters were activated more in the ASD compared with the TD group. (B) Cluster-based analysis revealed that group differences were specific to emotion recognition within bilateral pSTS and left IPL (not within left MOG). This was indicated by a significant group × task (Emo, Control) interaction. (C) Emotion recognition abilities were positively correlated with activity in right pSTS (small volume, P < 0.05, corrected). No other brain regions showed this relationship when tested at the whole-brain level (whole-brain, P < 0.0005, uncorrected).

Fig. 4

Resting-state iFC. (A) Resting-state iFC pattern of right pSTS in ASD and TD groups (one-sample t-test, P < 0.05, cluster-wise corrected). (B) Between-group comparisons showed that pSTS-iFC with bilateral IPL, left premotor and IFG was stronger in the TD group than in the ASD group (two-sample t-test, P < 0.05, cluster-wise corrected) (red–yellow). Stronger pSTS-iFC in the ASD group was observed with lingual/calcarine gyrus and superior occipital gyrus (blue–green). (C) Emotion recognition ability was positively correlated with the strength of pSTS-iFC with left/right IPL, left IFG, premotor area and left SMA (not shown) (P < 0.05, cluster-wise corrected).

Fig. 5

Relationship between task-based brain activity and resting-state iFC. Participants with high pSTS activity during emotion recognition also exhibited strong iFC of pSTS with left IPL and left premotor.

Fig. 6

ABIDE: resting-state iFC. Brain regions showing differences in resting-state iFC with right pSTS between the TD group (n = 306) and ASD group (n = 278) of the ABIDE sample (ABIDE) (two-sample t-test, P < 0.05, cluster-wise corrected). Stronger pSTS-iFC was observed in the TD group with right IPL, left premotor, bilateral fusiform gyrus and bilateral SOG (red–yellow) (regions indicated in bold font were previously identified in the original sample of 30 participants). Stronger pSTS-iFC in the ASD group was observed with right IFG (part triangularis) (blue–green) and left thalamus (not shown).