Cognitive profiles differ in autopsy-confirmed frontotemporal dementia and AD

Background: Frontotemporal dementia (FTD) is currently distinguished from AD primarily on the basis of behavioral features because studies of cognition have shown negligible or inconsistent differences. However, the poor discriminability of cognitive measures may relate to reliance on imprecise clinically diagnosed groups. Therefore, a retrospective examination of neuropsychological test performance in autopsy-confirmed patients is warranted. Objective: To compare the pattern of cognitive deficits exhibited by patients with autopsy-confirmed FTD and AD. Methods: The profiles of cognitive deficits exhibited by patients with neuropathologic diagnosis of FTD (n=14) or AD (n=28) were compared. The Mattis Dementia Rating Scale (MDRS), letter and category fluency tests, Wechsler Intelligence Scale for Children-Revised block design test, Boston naming test, and clock drawing test were administered. Results: Multivariate analysis of covariance controlling for age, education, and level of dementia revealed that patients with FTD performed significantly worse than patients with AD on letter and category fluency tests but significantly better on the MDRS memory subscale, block design test, and clock drawing test. A logistic regression model, validated in an independent clinical sample, used letter fluency, MDRS memory, and block design scores to correctly classify 91% of AD patients and 77% of FTD patients. Conclusions: A double dissociation in the pattern of cognitive deficits exhibited by FTD and AD patients was demonstrated. The FTD patients were more impaired than AD patients on word generation tasks (i.e., verbal fluency) that are sensitive to frontal lobe dysfunction but less impaired on tests of memory and visuospatial abilities sensitive to dysfunction of medial temporal and parietal association cortices.