Background Diminished ability to perceive one's own impairments, whether cognitive or social, is common in dementia, in particular frontotemporal dementia (FTD), where ‘lack of insight’ is listed as a core diagnostic feature. Yet, there is no currently accepted method for measuring insight in dementia. The most commonly used methods, which involve comparing patients' opinions of their level of impairment with the opinions of care givers or close family members, are subjective and require the participation of a knowledgeable informant. Here, the authors introduce a new method that allows objective quantification of an individual's awareness of their cognitive abilities and relies upon objective bedside testing.

Methods The authors administered several tests of everyday, real-world functions to patients with FTD (n=10) and Alzheimer's disease (AD, n=10) and to control subjects (n=10). Prior to the tasks, participants were asked to predict their performance using a percentile-based rating system. They were also asked to estimate their performance after task completion. Differences between their self-rated and actual performances were calculated.

Results Whereas the control group showed very little discrepancy between pretest predictions, post-task estimates and actual performance (mean difference of 3.9 percentile points for prediction/3.0 percentile points for post-task estimate), both patient groups overpredicted and overestimated their performance, with a significantly greater discrepancy for FTD (49.0/54.3 percentile points) than AD (27.2/28.3 percentile points).

Discussion Failures of insight and self-awareness of cognitive dysfunction can be objectively measured in dementia without the assistance of an informant, which will facilitate further study of this key component of higher cognitive functioning.