THE NEUROPSYCHOLOGICAL DIFFERENTIATION OF PATIENTS WITH VERY MILD ALZHEIMERS-DISEASE AND/OR MAJOR DEPRESSION

OBJECTIVE: To evaluate the usefulness of standardized neuropsychological tests in the psychometric differentiation of patients with very mild or mild Alzheimer's Disease (AD) and/or major depression presenting in a tertiary clinic with memory/attention complaints. DESIGN: Controlled prospective clinicoexperimental design. SETTING: Multidisciplinary Memory Clinic at Addenbrooke's Hospital, Cambridge, England. PARTICIPANTS: Twenty-four patients with a clinical diagnosis of Alzheimer's disease (12 with major depression and 12 without), 12 patients with major depressive illness but without AD, and 12 healthy control subjects, all matched for age, sex, education levels, and estimates of premorbid intellectual potential. MEASUREMENTS: Mini-Mental State Examination (MMSE), Wechsler's Logical Memory (WLM) and Visual Reproduction (WVR), immediate and delayed reproduction, Wechsler's paired Associate Learning (WPAL), including the Easy and Hard subsets. Warrington's Recognition Memory for Faces (WRMF), Kendrick's Object Learning (KOLT) and Digit Copying (KDCT)Tests. OUTCOME MEASURES: Minimum 2-year follow-up diagnosis. RESULTS: Statistically, patients with very mild AD were distinguished clearly from those without AD on most tests of memory functions. Psychometrically, only KOLT and an index of retention on WLM and WVR were specific enough to avoid false positives, a requirement for second-stage tools. They also proved sensitive enough to suggest their role as first-stage instruments when screening for primary dementia in high-functioning patients scoring above the cut-point on MMSE. CONCLUSIONS: As efforts intensify to develop more powerful means to identify patients with Alzheimer's disease in its earliest stages, inclusion of specialist tests posing greater cognitive challenge than standard mental status scales has been one strategy. Our study explored how some of these neuropsychological tools behave psychometrically when analyzed on a single-case basis, and the results suggest a few are sensitive enough to boost detection above base rates alone while also being specific enough to reduce false alarms. Retention on Wechsler's Logical Memory and Visual Reproduction tasks and scores on Kendrick's Object Learning Test helped decrease the degree of ambiguity when cognitive profiles were used to distinguish depressed patients with Alzheimer disease from these without.