The predictive value of social cognition assessment for 1-year follow-up functional outcomes in behavioral variant frontotemporal dementia
摘要
Behavioral variant frontotemporal dementia (bvFTD) is characterized by prominent socioemotional deficits that negatively impact daily living. This study aims to identify the strongest socioemotional predictors of functional status at 1-year follow-up and examine whether socioemotional assessment provides additional predictive value for functional outcomes beyond neuroimaging data in bvFTD.
MethodsA total of 78 bvFTD patients were assessed at baseline using a comprehensive social cognition battery and the Functional Activities Questionnaire (FAQ), and were reassessed 1 year later to examine longitudinal changes in their functional abilities. Regression modeling was performed to identify the strongest socioemotional predictors of functional outcomes at 1-year follow-up. In a subset of 57 patients with available baseline MRI data, the volumes of functional brain networks particularly vulnerable in bvFTD were obtained using voxel-based morphometry. Hierarchical regression analysis was employed to compare the predictive value of social cognition measures above neuroimaging data.
ResultsNeuropsychological tests assessing socioemotional semantic knowledge, sarcasm perception and cognitive perspective taking appeared to be the strongest socioemotional predictors of FAQ at the 1-year-follow-up assessment. The structural MRI data alone accounted for 42.01% of the variance in FAQ outcome, but increased to 55.85% when social cognition measures were included in the model (p < 0.001).
ConclusionsThis study highlights the significant contribution of quantifying social cognition in predicting which bvFTD patients are more likely to quickly develop adverse functional outcomes. Integrating socioemotional assessments in clinical practice with bvFTD can significantly enhance clinicians’ prognostic accuracy for functional outcomes, thus informing refined treatment strategies for these individuals.