Impact on dementia on behavioral indicators of pain and emotions

• Applicant: Professorin Dr. Miriam Kunz
• Subject Area: Biological Psychology and Cognitive Neuroscience
• Term: since 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 491518257

Across the course of dementia, individuals lose the ability to communicate verbally. As a consequence inner states, such as pain and negative emotions are not verbally expressed and thus, remain unnoticed and untreated. When the verbal communication of pain and affective states gets lost, one needs to rely on alternative communication channels, such as facial expression, vocalization and body movements. To this date, there are only a few studies on dementia-related changes in facial expressions. These studies point to an increased facial expression of pain and (negative) emotions in individuals with dementia; which seems to be linked to impairments in cognitive inhibitory functioning. With regard to the nonverbal behaviors “vocalization” (e.g. phonetic analyses) and “body movements” (e.g. actigraphy), research in individuals with dementia is mostly missing. Thus, the aim of this research proposal is to investigate the impact of dementia on nonverbal expressions of pain and emotions, which will be assessed in a multimodal way by relying on facial expression, vocalization and body movements. Moreover, it will be investigated whether dementia-related changes in pain and affective responses can be linked to specific neuropsychological deficits (especially in inhibitory functioning). To ensure ecological validity, pain and emotions (anger, happiness) are elicited during standardized activities of daily living (ADLs) in a controlled, real-life environment (living lab). Pain and emotional responses are assessed in 40 individuals with Alzheimer disease and in 60 cognitively healthy controls in a multimodal way (Facial Action Coding System for facial expression analyses; phonetic analyses of the voice; actigraphy and functional classification of movement patterns). A multimodal assessment is crucial when studying verbally compromised individuals, given that no other modality - besides self-report - can guarantee sufficient diagnostic certainty on its own. The data analyses will be performed using traditional statistical approach and will be complemented by machine learning techniques (interpretable machine learning).

• DFG Programme: Research Grants
• Co-Investigator: Professor Dr. Stefan Lautenbacher