Background: Despite striking neuropsychological and behavioural differences between Alzheimer's disease (AD) and frontotemporal dementia (FTD), clinical diagnostic criteria failed to discriminate FTD from AD patients. We therefore developed the Middelheim Frontality Score (MFS), a disease-long clinical and behavioural assessment tool that measures frontal lobe features, and set up this prospective study in clinically diagnosed AD and FTD patients to assess discriminatory power and intra- and inter-rater variability.
Methods: Patients with probable AD (n = 400) and FTD (n = 62) were included. The MFS was obtained by summating the scores obtained in a standardized fashion on ten items yielding a total maximal score of 10. Information was obtained through an interview of the patient and her/his caregiver, clinical files and behavioural observation.
Results: Comparing mean total MFS scores, FTD patients (6.3 +/- 1.8) had significantly higher scores than AD patients (3.1 +/- 1.8) (p < 0.001). Distribution of scores on individual MFS items was significantly different between both disease groups (chi(2) = 76.2; p < 0.001). A moderately positive and highly significant correlation was shown between the total MFS score and diagnosis FTD (r = 0.478; p < 0.0001). Applying a total MFS score of 5 as discriminatory cut-off, a specificity of 89.0% and a sensitivity of 88.7% were achieved. Intra- and inter-rater variability was calculated in a different study population by means of retest correlation, revealing moderate to strong positive correlations of high statistical significance.
Conclusions: The MFS is a clinical and behavioural assessment scale that measures frontal lobe features and that was shown to reliably discriminate FTD from AD patients.