Does physical activity improve sleep in impaired nursing home residents?

Abstract

Objectives: To determine if two physical activity programs of varying intensity would result in improved sleep among incontinent and physically restrained nursing home residents.

Design: Controlled trials of two physical activity programs.

Setting: Seven community nursing homes in the Los Angeles area.

Participants: Residents were included if they had urinary incontinence or were physically restrained. Sixty-five subjects were studied. Mean age was 84.8 years, 85% were female, mean length of residency in the nursing home was 19.9 months, and mean Mean Mini-Mental State Exam score was 13.1.

Intervention: The first physical activity program involved sit-to-stand repetitions and/or transferring and walking or wheelchair propulsion. These activities were performed every 2 hours during the daytime, 5 days per week for 9 weeks. The second, less frequent physical activity program involved rowing in a wheelchair-accessible rowing machine plus walking or wheelchair propulsion once per day three times per week for 9 weeks.

Measurements: The physical function measures reported here include mobility endurance (maximum time walking or wheeling) and physical activity as measured by motion sensors (Caltrac). Nighttime sleep was estimated by wrist activity monitors. Nighttime sleep measures included total time asleep, percent sleep, average duration of sleep, and peak duration of sleep. Daytime sleep was measured by timed behavioral observations of sleep versus wakefulness performed every 15 minutes during the day.

Results: Nighttime sleep was markedly disrupted in both groups at baseline. Across all subjects at baseline, the average total sleep time was 6.2 hours and the percent sleep was 72.0%, but the average duration of sleep episodes was only 21.2 minutes and the peak duration of sleep episode averaged only 83.8 minutes. During the daytime, subjects were observed asleep during 14.5% of observations. Although there was improvement in mobility endurance in the intervention subjects compared with controls (MANOVA F = 4.36, P = .042), there were no differences in the night and day sleep measures at follow-up testing. Even among a subgroup of intervention subjects who showed a 30% or greater improvement in mobility endurance, sleep did not improve at follow-up compared with baseline.

Conclusion: This study supports our previous findings of marked sleep disruption in impaired nursing home residents. In addition, despite documented improvements in physical function with activity, we did not find improvements in sleep in the intervention versus control groups. These results suggest that increasing daytime physical activity alone is not adequate to improve sleep in impaired NH residents. Future efforts to improve sleep in this population should take into account the multifactorial nature of sleep disruption, including individual health problems that effect sleep and the disruptive nature of the nighttime NH environment.