Umeå University's logo

Background: A team-based, individualised rehabilitation approach may be required to meet the complex needs of people with dementia. This randomised controlled pilot trial evaluated the feasibility of a person-centred multidimensional interdisciplinary rehabilitation programme for community-dwelling older people with dementia and their informal primary caregivers.

Methods: Participants with dementia were randomised to an intervention group (n = 31, mean age (SD) 78.4 (6.0) years) or usual care (n = 30, mean age 79.0 (7.1)). The rehabilitation programme consisted of a 20-week rehabilitation period containing assessments and interventions based on each individual’s goals, and group-based physical exercise plus social interaction twice a week for 16 weeks at a rehabilitation unit. After 5 and 14 months, the interdisciplinary team followed up participants over two four-week periods. For both groups, dates of deaths and decision to move to nursing home over three years, as well as interventions for the relevant periods, were collected. Blinded assessors measured physical functions, physical activity, activities of daily living, cognitive functions, nutritional status, and neuropsychiatric symptoms at baseline and at 5, 12, 24, and 36 months.

Results: Participants in the intervention group received a mean of 70.7 (20.1) interventions during the 20-week rehabilitation period, delivered by all ten team professions. The corresponding figures for the control group were 5.8 (5.9). In the intervention group, all but one participated in rehabilitation planning, including goal setting, and attendance in the exercise and social interaction groups was 74.8%. None of the adverse events (n = 19) led to any manifest injury or disease. Cox proportional hazard regression showed a non-significant lower relative risk (HR = 0.620, 95% CI 0.27–1.44) in favour of the intervention for moving to nursing home or mortality during the 36-month follow-up period. Linear mixed-effect models showed non-significant but potentially clinically meaningful between-group differences in gait, physical activity, and neuropsychological symptoms in favour of the intervention.

Conclusions: The rehabilitation programme seems feasible among community-dwelling older people with dementia. The overall results merit proceeding to a future definitive randomised controlled trial, exploring effects and cost-effectiveness. One could consider to conduct the programme earlier in the course of dementia, adding cognitive training and a control attention activity.

Trial registration: The study protocol, ISRCTN59155421, was registered online 4/11/2015.