Self-reported exercise prevalence and determinants in the long term after stroke: The North East Melbourne Stroke Incidence Study

Background: Exercise has established benefits following stroke. We aimed to describe self-reported exercise 5 and 10 years after stroke, change in exercise over time, and to identify factors associated with long-term exercise.

Methods: Data on exercise (defined as 20 minutes' duration, causing sweating and increased heart rate) were obtained by questionnaire from a population-based stroke incidence study with 10-year follow-up. For change in exercise between 5 and 10 years (n = 276), we created 4 categories of exercise (no exercise, ceased exercising, commenced exercising, continued exercising). Multinomial regression determined associations between exercise categories and exercising before stroke, receiving exercise advice and functional ability and demographic factors.

Results: The prevalence of exercise at 5 years (n = 520) was 18.5% (n = 96) (mean age 74.7 [standard deviation {SD} 14] years, 50.6% male) and 24% (n = 78) at 10 years. In those with data at both 5 and 10 years (mean age 69 [standard deviation 14] years, 52.9% male), 15% (n = 42) continued exercising, 10% (n = 27) commenced exercising, 14% (n = 38) ceased exercising, and 61% (n = 169) reported no exercise. Continued exercise was associated with younger age (relative risk [RR] .47 95% confidence interval [CI] .25-0.89), greater Barthel score (RR 2.97 95% CI 1.00-8.86), independent walking (RR 2.32 95% CI 1.16-4.68), better quality of life (RR 10.9 95% CI 2.26-52.8), exercising before stroke (RR 16.0 95%CI 4.98-51.5), and receiving advice to exercise (RR 2.99 95% CI 1.73-5.16).

Conclusions: Few people exercise after stroke and fewer commence exercise long term. Innovative interventions to promote and maintain exercise are required after stroke.