Differential diagnosis of depression and Alzheimer’s disease with the Addenbrooke's Cognitive Examination-Revised (ACE-R).

Background: One of the usual problems psychologists and clinicians face in clinical practice is differential diagnostics of Alzheimer’s disease and depression. It has been reported that the ACE and ACE-R could discriminate the cognitive dysfunctions due to depression from that due to dementia, although this is not uniform in all studies. The current study aimed to evaluate the utility of the ACE-R to differentiate late-life onset depression (with severe episode) from mild-moderate Alzheimer’s Disease (AD). Methods: This study received approval from the Lithuanian Bioethics Committee. All participants were older than 50 years (mean age = 66.52 (±8.76) years). The study sample consisted of 295 individuals: 117 with severe depression, 85 with mild-moderate Alzheimer’s disease (AD), and 94 age, gender and education matched participants of control group. Results: The ACE-R had high sensitivity (100%) and specificity (81%) at detecting cognitive impairments related to AD. Patients with late-life onset depression (ACE-R mean 76.82, SD = 7.36) performed worse than controls (ACE-R mean 85.08, SD = 7.2), but better than the AD group (ACE-R mean 54.74, SD = 12.19). Participants with late-life onset depression were differentiated by mild impairment in the ACE-R total score with mild memory (13.79, SD = 6.29) and greater deficits in letter fluency (3.65, SD = 1.21) than in semantic fluency (4.68, SD = 1.23). Participants with AD were differentiated by severely impaired performance on attention and orientation (11.80, SD = 2.93), memory (8.25, SD = 3.47) and language subtests (17.21, SD = 4.04), and moderately impaired performance on verbal fluency (6.07, SD = 2.74). Conclusions: ACE-R has diagnostic accuracy in detecting people with AD and can be used in differential diagnostics of late-life onset depression (severe episode) and AD. Diagnostic accuracy may be improved by analyzing the neuropsychological profiles and using lower cutoffs for different age groups