Derivation of a brief measure of agitation and aggression in Alzheimer's disease

Objective: Neuropsychiatric symptoms, including agitation and aggression (A/A), are highly prevalent in Alzheimer's disease (AD) and are associated with increased disability, functional impairment, caregiver distress, and institutionalization. Previous psychometric work suggests that individual items of agitation, irritability, disinhibition, and aberrant motor behavior from the Neuropsychiatric Inventory (NPI) may be a valid measure of A/A in AD. We provide additional confirmation of this subscale, as well as preliminary validation of it as a measure of A/A (the NPI-4-A/A).

Methods: The sample included 641 individuals identified from the South Carolina Alzheimer's Disease Registry and assessed to be at a nursing home level of care. Demographic and medical data were extracted from the Registry, and phone interviews were conducted with caregivers to collect additional information not included in the Registry. The primary statistical analysis was confirmatory factor analysis of the NPI-12 factor structure.

Results: The standardized root mean residual and root mean square error of approximation (90% CI) values of 0.060 and 0.043 (0.030, 0.057), respectively suggest adequate model fit of the data, whereas the Tucker-Lewis index estimate of 0.779 is below the criteria for adequate model fit. All but two normalized residuals (NR) suggested adequate model fit of the data (|NR| < 2.58). NPI-4-A/A scores were higher in patients residing in nursing homes and were correlated with caregiver burden.

Conclusions: The NPI-4-A/A is proposed as a measure of A/A in AD. The utility and validity of this measure should be explored further. Improved measurement and focus on subgroups of neuropsychiatric symptoms should be integrated into interventions for AD.