Using a second-person approach to identify disease-specific profiles of social behavior in frontotemporal dementia and Alzheimer's disease.

Elsevier BV
Publication Type:
Journal Article
Cortex; a journal devoted to the study of the nervous system and behavior, 2020, 133, pp. 236-246
Issue Date:
Filename Description Size
1-s2.0-S0010945220303531-main.pdfPublished version788.13 kB
Adobe PDF
Full metadata record
Changes in social behavior are recognized as potential symptoms of behavioral-variant frontotemporal dementia (bvFTD) and semantic dementia (SD), yet objective ways to assess these behaviors in natural social situations are lacking. This study takes a truly social (or second-person) approach and examines changes in real-world social behavior in different dementia syndromes, by analyzing non-scripted social interactions in bvFTD patients (n = 20) and SD patients (n = 20), compared to patients with Alzheimer's disease (AD) (n = 20). Video recordings of 10-min conversations between patients and behavioral neurologists were analyzed for the presence of socially engaging (e.g., nodding, smiling, gesturing) and disengaging behavior (e.g., avoiding eye contact, self-grooming, interrupting). Results demonstrated disease-specific profiles, with bvFTD patients showing less nodding and more looking away than AD, and SD patients showing more gesturing than AD. A principal components analysis revealed the presence of four unobserved components, showing atypical disengaging patterns of behavior. Whole-brain voxel-based morphometry analyses revealed distinct neurobiological bases for each of these components, with the brain regions identified previously associated with behavior selection, abstract mentalization and processing of multi-sensory and socially-relevant information, in mediating socially engaging and disengaging behavior. This study demonstrates the utility of systematic behavioral observation of social interactions in the differential diagnosis of dementia.
Please use this identifier to cite or link to this item: