Differences in Temporal Profile of Brain Responses By Pleasantness of Somatosensory Stimulation in Individuals with ASD

Background: Individuals with autism spectrum disorders (ASD) show varied behavioral responses to somatosensory stimuli, including hyper-reactivity to light touch and hypo-reactivity to pain. Investigation of the neural processing underlying these differences can be informed by previous fMRI studies of sustained painful stimuli. These studies have isolated three distinct phases of pain processing, including a late phase encompassing stimulus offset and reflecting cognitive-affective evaluation of pain (Moulton et al., 2012). Individuals with ASD show diminished response in the two later phases of this triphasic pattern (Failla et al., 2017), suggesting differences in pain evaluation but not sensation or perception. It is not known whether this is unique to painful stimulation or if this pattern extends to non-painful stimuli, either pleasant or aversive.

Objectives: Our objective was to investigate the temporal profile of brain responses to non-painful somatosensory stimulation in individuals with ASD.

Methods: 28 Adults (ASD=14, typically developing (TD)=14) underwent functional magnetic resonance imaging (fMRI) in a block design using 3 textures with previously normed pleasantness ratings in the general population (Essick et al., 2010): soft brush (pleasant), burlap (neutral), and plastic mesh (unpleasant). On each trial, a texture was stroked along the right forearm with moderate speed and pressure for 30 seconds. Changes in blood oxygenation level-dependent (BOLD) signal were analyzed during early, intermediate, and late phases, consistent with prior studies. Results were cluster-corrected, thresholding at Z>2.3 and p< .05.

Results: Between groups, we found significant differences (TD>ASD) in BOLD signal change for pleasant (brush) and neutral (burlap) textures. For the brush, the TD group had greater response in left postcentral gyrus and right superior parietal lobule during the early and intermediate phases (all peak Zs> 4.0, all p_corr < .0001). For the burlap, group differences were only significant during the early phase and occurred in the paracingulate gyrus (Z_peak=4.16, p_corr<.005) and right angular gyrus (Z_peak=4.01, p_corr<.05). There were no group differences for the plastic mesh texture.

Within groups, the TD group showed significant early and intermediate phase responses to all stimuli. The ASD group had a more limited pattern of response, with increased BOLD signal during the late phase for the burlap texture (Left Central Operculum/Insular Cortex, Z_peak=3.62, p_corr<.005) and during the early phase for the mesh texture (Right Parietal Operculum, Z_peak=3.64, p_corr<.0001 and Left Insula, Z_peak=3.87, p_corr<.0001).

Conclusions: Group differences for the more pleasant textures seemed driven by limited responding in the ASD group, who showed more reliable responses to more aversive stimuli. As cognitive-affective evaluation may be less critical for non-painful stimuli, the TD group did not show late phase responses; however, the late phase response to the neutral burlap in the ASD group may reflect difficulty evaluating ambiguous, neutral sensory stimuli. The finding of an intact early but not intermediate and late phase response by the ASD group to the mildly aversive mesh texture extends the finding by Failla et al. (2017) in pain processing, suggesting a continuum of aversive sensation for which cognitive-affective evaluative processes may be affected in ASD.