Exercise or Social Intervention for Nursing Home Residents with Dementia: A Pilot Randomized, Controlled Trial

To compare the effects of exercise with those of a structured nonphysical intervention on ability to perform activities of daily living (ADLs) and physical and cognitive function of persons with dementia (PWDs) living in nursing homes (NH).

Cluster-randomized pilot-controlled trial.

Seven French NHs.

PWDs living in NHs.

NHs were randomized to an exercise group (4 NHs, n = 47) or structured social activity group (3 NHs, n = 50) for a 24-week intervention performed twice per week for 60 minutes per session. The main endpoint was ADL performance (Alzheimer's Disease Cooperative Study Activities of Daily Living Inventory for Severe Alzheimer's Disease Scale (ADCS-ADL-sev); range 0-54, higher is better); secondary endpoints were overall cognitive function (Mini-Mental State Examination (MMSE)) and performance-based tests of physical function (Short Physical Performance Battery (SPPB), usual gait speed).

Ninety-one participants with at least one postbaseline ADL assessment were included in efficacy analysis. Groups differed at baseline in terms of sex, neuropsychiatric symptoms, and nutritional status. Multilevel analysis adjusted for baseline differences between groups found no significant difference between effects of exercise and social activity (group-by-time interaction), with adjusted mean differences at 6 months of 1.9 points for ADCS-ADL-sev and 0.55 points for MMSE favoring social activity and 0.6 points for SPPB and 0.05 m/s favoring exercise. Adverse events did not differ between groups, except that the social activity group had more falls than the exercise group.

A larger, longer trial is required to determine whether exercise has greater health benefits than nonphysical interventions for institutionalized PWDs.