Objective: To assess the prevalence of depression in subjects with preclinical Alzheimer's disease (AD) and to investigate the possibility of differentiating subjects with preclinical AD and depression from subjects with depression-related cognitive impairment.
Design: A prospective, observational cohort study.
Setting: An outpatient memory clinic of a university-affiliated hospital.
Participants: Nondemented subjects with cognitive impairment older than 55 years (n = 111) without neurological or somatic causes for the cognitive impairment.
Measurements: At baseline, data were collected on patient characteristics, the severity of depression, and cognitive functioning. The course of the cognitive impairment and the presence of dementia were assessed after 2 and 5 years.
Results: Twenty-five subjects had preclinical dementia with Alzheimer's type dementia at follow-up. Sixty percent of these subjects (n = 15) were depressed at baseline. Subjects with depression and preclinical AD had at baseline a poorer performance on the cognitive tasks and were older than the subjects with depression-related cognitive impairment. Logistic regression with backward step selection selected age and memory performance as the best predictors for Alzheimer's type dementia in the depressed subjects. The specificity of these predictors for the diagnosis of future Alzheimer's type dementia in depressed subjects was 94%, sensitivity was 90%, positive predictive value was 90%, and negative predictive value was 94%.
Conclusions: Depression is common in preclinical AD. Depressed subjects with preclinical AD can be accurately differentiated from subjects with depression-related cognitive impairment by age and the severity of the memory impairment. Research that aims to investigate preclinical AD should not exclude a priori subjects with depression inasmuch as preclinical AD is often accompanied by depression.