Objectives: To compare the trajectories of motor and cognitive decline in older adults who progress to dementia with the trajectories of those who do not. To evaluate the added value of measuring motor and cognitive decline longitudinally versus cross-sectionally for predicting dementia.
Design: Prospective cohort study with 5 years of follow-up.
Setting: Clinic based at a university hospital in London, Ontario, Canada.
Participants: Community-dwelling participants aged 65 and older free of dementia at baseline (N=154).
Measurements: We evaluated trajectories in participants' motor performance using gait velocity and cognitive performance using the MoCA test twice a year for 5 years. We ascertained incident dementia risk using Cox regression models and attributable risk analyses. Analyses were adjusted using a time-dependent covariate.
Results: Overall, 14.3% progressed to dementia. The risk of dementia was almost 7 times as great for those whose gait velocity declined (hazard ratio (HR)=6.89, 95% confidence interval (CI)=2.18-21.75, p=.001), more than 3 times as great for those with cognitive decline (HR=3.61, 95% CI=1.28-10.13, p=.01), and almost 8 times as great in those with combined gait velocity and cognitive decline (HR=7.83, 95% CI=2.10-29.24, p=.002), with an attributable risk of 105 per 1,000 person years. Slow gait at baseline alone failed to predict dementia (HR=1.16, 95% CI=0.39-3.46, p=.79).
Conclusion: Motor decline, assessed according to serial measures of gait velocity, had a higher attributable risk for incident dementia than did cognitive decline. A decline over time of both gait velocity and cognition had the highest attributable risk. A single time-point assessment was not sufficient to detect individuals at high risk of dementia.
Keywords: cognition; cohort study; dementia; gait velocity; trajectories.
© 2018 The Authors. The Journal of the American Geriatrics Society published by Wiley Periodicals, Inc. on behalf of The American Geriatrics Society.