Abstract
Dementia is a neurodegenerative syndrome affecting multiple aspects of life. The large heterogeneity in disease course causes high uncertainty about patients’ prognosis. Patients may live in fear of rapid cognitive decline, while in fact their disease progresses slowly and vice versa. This study aims to provide patients with an individualized prognosis by investigating trajectories of cognition and functioning simultaneously over time. In the Clinical Course of Cognition and Comorbidity study, 331 dementia patients were followed yearly from diagnosis onwards for a maximum of three years. Cognition was measured using the Mini Mental State Examination. Functioning was recorded according to the Disability Assessment for Dementia. We built a parallel-process growth mixture model to jointly model cognitive and functional progression and used the Bayesian Information Criterion to assess model fit. Three distinct classes of cognitive and functional trajectories were observed: 32% of the patients showed moderate progression speed, 12% declined rapid and 56% exhibited slowly declining trajectories for both cognition and functioning over time. Cognitive and functional rates of decline were highly correlated (r=0.86, p=0.000). This study shows that heterogeneity in the course of dementia could be modeled as latent classes of trajectories and that most individuals were members of a class with stable and slow progression. This presents a more optimistic picture of dementia progression as compared to presenting the mean trajectory across the entire population. Relating class membership to baseline variables can help explain the heterogeneity in disease progression.