The stability of neuropsychiatric subsyndromes in Alzheimer's disease.

INTRODUCTION: Neuropsychiatric symptoms are common in Alzheimer's disease. Previous research has attempted to identify subsyndromes-sets of symptoms related to one another-to clarify underlying mechanisms and treatment targets. We examined the stability of these subsyndromes over time.

METHODS: We administered the Neuropsychiatric Inventory annually for 3 years to 447 patients with Alzheimer's disease recruited from memory clinics. We conducted principal component analyses at each time point and multiple-group confirmatory factor analyses across time.

RESULTS: Principal component analyses showed that no two time points shared the same factor structure. Factor solutions did not exhibit strong simple structures and substantial cross-loadings were common. Confirmatory analysis revealed significant differences in factor loadings and model fit over time.

DISCUSSION: Symptoms cannot be neatly partitioned into discrete clusters that are stable over time. The findings highlight the significant challenges that clinicians and caregivers face and may help explain the lack of success in intervention studies.