Prediction of psychosis onset in Alzheimer disease: The role of cognitive impairment, depressive symptoms, and further evidence for psychosis subtypes.

BACKGROUND: Psychotic symptoms in Alzheimer disease (AD+P) identify a heritable phenotype associated with more rapid cognitive decline. The authors have proposed that AD+P is itself a composite of a misidentification and a paranoid subtype with increased cognitive impairment restricted to the misidentification type. Most prior studies of the clinical correlates of AD+P have been limited, however, by the inclusion of prevalent cases. METHODS: Subjects with possible or probable AD or mild cognitive impairment (MCI) without psychosis at study entry were assessed at the time of initial presentation and then annually. Psychotic symptoms were assessed using the CERAD Behavioral Rating Scale. Survival analyses used Cox proportional hazard models with time-dependent covariates to examine the predictors of psychosis onset. RESULTS: A total of 288 subjects completed at least one follow-up examination. Mean duration of follow-up was 22.1 months. The incidence of psychosis was 0.19 per person-year. Cognitive impairment was associated with onset of psychosis, largely as a result of its association with onset of the misidentification, but not the paranoid, subtype. Including psychotropic medication use in the model revealed an association of antidepressant use with the onset of psychosis. This latter association appeared to arise from an underlying association between depression and the risk of psychosis onset rather than from antidepressant treatment. CONCLUSION: These findings are consistent with the hypothesis that the misidentification and the paranoid subtypes each define a more biologically homogeneous group than AD+P as a whole. Further exploration of the relationship between depressive symptoms and psychosis in patients with AD is warranted.