Does cardiac rehabilitation for people with stroke in the sub-acute phase of recovery lead to physical behaviour change? Results from compositional analysis of accelerometry-derived data

Abstract

ObjectivesDoes adapted cardiac rehabilitation (CR) improve the physical behaviours of people with mild-to-moderate stroke in the sub-acute recovery phase using a compositional data analysis (CoDA) approach? DesignBefore-after. SettingUniversity Hospitals of Leicester, Glenfield Hospital, UK. Participants24 individuals completed CR and provided valid physical activity (PA) data (mean (SD) 63.1 (14.6) years, 58% male (14/24)). Intervention6-week adapted CR program within 6-months of stroke. Main outcome measuresPhysical behaviours were assessed using waist-worn accelerometry. Step count, stationary time (ST), light PA (LPA), and moderate-to-vigorous PA (MVPA) were compared pre post CR using conventional analyses and CoDA. Analysed compositions were: Waking day (ST, LPA, MVPA); ST (1–9-minutes, 10–29-minutes, ≥30-minutes bouts); and MVPA (1–4-minutes, 5–9-minutes, ≥10-minutes bouts). ResultsFollowing CR, patients took significantly more steps (mean (SD) 3255 (2864) vs 3908 (3399) steps/day, P=0.004) and engaged in more bouts of MVPA lasting ≥5 and ≥10-minutes (≥5-minutes: mean (SD) 0.7 (1.4) vs 1.2 (1.8) bouts/day, P=0.008). Using CoDA, no changes in waking day or ST compositions occurred. For waking day, 42% (10/24) increased their LPA and MVPA at the expense of ST. For ST, 33% (8/24) increased their short bouts at the expense of medium and long bouts. For MVPA, 13% (3/24) increased their medium and long bouts at the expense of short bouts. ConclusionPeople with stroke in the sub-acute stage of recovery exhibited low levels of PA. CR appears to be an effective intervention to increase step count but did not alter the overall proportion of time individuals spent being sedentary, or engaging in LPA or in MVPA. RegistrationISRCTN65957980.