Centre de gérontologie clinique Rainier-III, centre hospitalier Princesse-Grace, Monaco
Département de santé publique, CHU de Nice, France
Association de recherche bibliographique pour les neurosciences (AREBISN), Nice, France
Laboratoire d’anthropologie et de psychologie cognitives et sociales (LAPCOS, EA 7278), Université Nice-Côte d’Azur, Nice, France
Objective. Theory of mind (ToM) and empathy are severely impaired in the behavioral-variant of frontotemporal dementia (bvFTD) and more mildly in Alzheimer's (AD) and Parkinson's diseases (PD). Such impairments are associated with behavioral disorders (BD). Modification of visual scanning strategies of complex visual scenes is also found in these pathologies. We hypothesized that these patients applied atypical gaze strategies when observing social events, which would not allow to properly process social cues and would result in the production of erroneous inferences and lack of empathy towards others. Methods. Fifty-five participants were divided into four groups: five bvFTD, 19 AD, 17 PD and 14 matched controls subjects. ToM and empathy were assessed by eye movements recording (eye-tracking) and by a questionnaire during a painting observation. Scores obtained were compared between each group and to social cognition reference tests, and correlated to the NeuroPsychiatric Inventory. Results. Our paradigm was suitable for assessing cognitive ToM while it lacked sensitivity for empathy assessment. Severe ToM impairment was highlighted in bvFTD while milder difficulties were observed in AD and for PD. bvFTD and AD groups produced erroneous inferences from cognitive mental states. ToM performances were linked to visual exploration strategies of the painting. Atypical visual observation was highlighted in bvFTD and AD groups causing a time shift in perspective taking of the character. Finally, we have highlighted that social cognition performances, gaze strategies and BD were correlated. Conclusion. The observation of a painting in association with eye-tracking technology can be a good support for social cognition assessment. We highlighted a link between atypical visual scanning strategies, ToM impairment and BD in these pathologies. ToM skills could be improved by training in the search for visual social cues. Therefore, this kind of remediation could have positive effects on BD.