Assessing the dysexecutive syndrome in dementia.

OBJECTIVE: We compared performance on tests of dysexecutive behaviour (DB) and executive function (EF) in patients with behavioural variant frontotemporal dementia (bvFTD), primary progressive aphasia (PPA) and corticobasal syndrome (CBS).

METHODS: Patients diagnosed with bvFTD (n=124), PPA (n=34) and CBS (n=85) were recruited. EF was measured with the Delis-Kaplan Executive Function System (DKEFS: performance based), and DB was measured with the Frontal Systems Behavior Scale (FrSBe: caregiver-report based). Confirmatory factor analysis characterised the relationship between EF and DB, binary logistic regression evaluated the incremental diagnostic utility of the measures and neuroimaging data from 110 patients identified neural correlates.

RESULTS: EF was lowest and DB was highest in bvFTD participants. EF and DB were distinct but related (r=-0.48). Measures correctly classified 89% of bvFTD from CBS patients and 93% of bvFTD from PPA patients-30% and 13% above base rates (59%, 80%), respectively. All modalities were useful in identifying CBS and PPA, whereas DB alone was useful for identifying bvFTD. EF was uniquely associated with caudal left dorsolateral prefrontal and lateral temporo-parietal cortices. DB was uniquely associated with the cingulate (R>L), right subcallosal and right anterior frontal cortex. EF and DB were associated with the rostral dorsolateral prefrontal cortex bilaterally.

CONCLUSIONS: EF and DB measures displayed criterion and construct validity, had incremental utility at low DB levels (CBS and PPA) and were associated with overlapping and distinct neural correlates. EF and DB procedures can conjointly provide useful diagnostic and descriptive information in identifying and ruling out the dysexecutive syndrome.