To test whether personality traits conferred risk of AD, we used proportional hazards regression models, controlling for age of personality assessment, sex, ethnicity (white vs. others), and education (years of schooling). The analyses were conducted separately for each of the five personality domains and each of the facets. Personality scores were standardized so that one unit corresponded to a 1 SD difference. In addition to the continuous scores, domain scores were recoded to provide a statistical and graphical comparison of the top and bottom quartiles of the distribution. We also tested a model that included all five factors simultaneously. The time end point was the year of onset of AD-type clinical dementia. Participants who did not develop AD were censored at the time of their last clinical evaluation. Because of differences in the pathophysiologic processes among dementia subtypes, we excluded 44 participants who developed non-AD dementia (e.g., vascular, Lewy body, Parkinson disease). The results were similar if we included these 44 participants censored at time of onset of non-AD dementia.
