Do you see what I am saying? Studies on multisensory in autism

Abstract

The central research aim in the present thesis was to increase our understanding of multisensory processing abilities in individuals with autism, and investigate how possible impairments might be related to clinical features of the disorder. Impairments in sensory perception have repeatedly been associated with autism, and might ultimately result in impairments of social communication. An important aspect of everyday social processing is that signals arrive from different modalities at the same time. Integration across sensory systems plays an essential role in communication, especially during face-to-face interactions. We can, for instance, hear a laughing face and see what someone is saying. Given the impairments in those domains of communication in which crossmodal associations are particularly important (e.g., emotions and speech), there is good reason to hypothesize that multisensory processing will be especially impaired in individuals with autism. Using several research techniques (EEG and EMG) and paradigms, this thesis provides clear improvements in our knowledge of multisensory integration (MSI) in individuals with autism. On a point by point basis, the evidence given in this thesis comprises the following findings: Automatic facial muscle activity in response to seeing happy and fearful faces is larger in individuals with autism than matched controls. These results can be interpreted as an increased physiological responsiveness in individuals with autism to the presentation of facial stimuli. Possibly, individuals with autism develop a kind of homeostasis-driven behavior, such as the typical gaze avoidance, to control for hyperactive physiological responses to faces. Low-level MSI, as measured by a crossmodal EEG gating paradigm, is intact in individuals with autism. Intact rapid crossmodal associations indicate normal functional connectivity between primary visual and auditory processing areas in autism. Atypical crossmodal gating was found in a group of schizophrenia individuals, which makes the crossmodal sensory gating paradigm a potential psychophysiological measure from which autism and schizophrenia can be distinguished. Using naturalistic audiovisual (AV) speech stimuli, we found normal low-level, but atypical higher-order MSI in individuals with autism. Although these findings could indicate abnormal MSI of phonological information, one should not overlook the possibility that the observed results are due to others factors, such as attention impairments, as these are known to influence the way in which crossmodal associations are made. Individuals with autism are able to integrate emotional signals from face and voice, on low-level as well as higher-order levels of processing, although there are several factors that have an atypical influence on MSI in autism. Together with anomalies in the perceptual processing of emotional signals-in-noise, inappropriate attention mechanisms may cause aberrant multisensory EEG responses. Importantly, in combination with the observed hyper-responsiveness to emotional faces at the level of reflex-like motor activity, this pattern of information processing may cause overloading of higher-cognitive processes under naturalistic multisensory situations, and consequently lead to the specific deficits in social cognition associated with autism.