Principal components analysis of agitation outcomes in Alzheimer's disease

Abstract

BackgroundWe developed a composite measure of agitation as a secondary outcome of change over time in the Citalopram for Agitation in Alzheimer's disease study (CitAD). CitAD demonstrated a positive effect of citalopram on agitation on the Neurobehavioral Rating Scale agitation subscale (NBRS-A). CitAD included additional agitation measures such as the Cohen-Mansfield Agitation Inventory and the Neuropsychiatric Inventory. MethodsWe performed principal components analyses on change in individual item of these scales for the same, original CitAD subjects. ResultsThe first principal component accounted for 12.6% of the observed variance and was composed of items that appear to reflect agitation. The effect size for citalopram calculated using this component was 0.53 (95% CI 0.22–0.83) versus 0.32 for the NBRS-A (95% CI 0.01–0.62). ConclusionsResults suggest that a composite measure of change in agitation might be more sensitive than change in a single primary agitation measure.