Comparing the value of dynamic versus static image-based tests of emotion recognition in neurodegenerative disease

More precise subtyping within dementia syndromes leads to better prediction of pathology, supporting individualized, disease-specific treatments, and measuring emotion recognition abilities is particularly important for disorders like semantic behavioural variant frontotemporal dementia (sbvFTD). To evaluate current tools' effectiveness, we compared two dynamic video-based affect labelling tests—the Dynamic Affect Recognition Test (DART), The Awareness of Social Inference Test-Emotion Evaluation Test (TASIT-EET)—and a static image-based Comprehensive Affect Testing System-Name Affect (CATS-NA) test. A total of 555 persons with dementia (PwD), diagnosed with Alzheimer's disease (n = 154), progressive supranuclear palsy syndrome (n = 88), non-fluent variant primary progressive aphasia (nfvPPA) (n = 77), semantic variant PPA (svPPA) (n = 53), bvFTD (n = 124), sbvFTD (n = 65) and 133 healthy participants underwent emotion testing and MRI. All emotion labelling tests differentiated PwD from healthy controls, and FTD with social cognition deficits (sbvFTD, bvFTD and svPPA) from other PwDs. DART outperformed TASIT-EET and CATS-NA in differentiating sbvFTD from bvFTD and svPPA. Beyond emotion reading, TASIT-EET performance was also mediated by verbal memory and semantics, and CATS-NA was predicted by visuospatial and executive functions, while DART required only semantic knowledge. DART corresponded to expected structural anatomy of emotion, including predominantly right insula, anterior temporal and orbitofrontal lobes. While both TASIT-EET and CATS-NA shared that pattern, TASIT-EET additionally correlated with left temporal structures and CATS-NA associated with more dorsal frontoparietal regions. Careful selection of tests that utilize dynamic stimuli, reflect key neural circuits related to emotion and minimize demands from other cognitive domains may yield more accurate diagnoses.