Home-based cardiac rehabilitation and physical activity in people with heart failure:a secondary analysis of the REACH-HF randomised controlled trials

OBJECTIVES: To quantify the impact of a home-based cardiac rehabilitation intervention (Rehabilitation Enablement in Chronic Heart Failure (REACH-HF)) on objectively assessed physical activity (PA) of patients with heart failure (HF) and explore the extent by which patient characteristics are associated with a change in PA. DESIGN: Secondary analysis of randomised controlled trial data. SETTING: Five centres in the UK. PARTICIPANTS: 247 patients with HF (mean age 70.9±10.3 years; 28% women). INTERVENTIONS: REACH-HF versus usual care (control). PRIMARY AND SECONDARY OUTCOME MEASURES: PA was assessed over 7 days via GENEActiv triaxial accelerometer at baseline (pre-randomisation), post-intervention (4 months) and final follow-up (6-12 months). Using HF-specific intensity thresholds, intervention effects (REACH-HF vs control) on average min/day PA (inactivity, light PA and moderate-to-vigorous PA (MVPA)) over all days, week days and weekend days were examined using linear regression analysis. Multivariable regression was used to explore associations between baseline patient characteristics and change in PA. RESULTS: Although there was no difference between REACH-HF and control groups in 7-day PA levels post-intervention or at final follow-up, there was evidence of an increase in weekday MVPA (10.9 min/day, 95% CI: -2.94 to 24.69), light PA (26.9 min/day, 95% CI: -0.05 to 53.8) and decreased inactivity (-38.31 min/day, 95% CI: -72.1 to -4.5) in favour of REACH-HF. Baseline factors associated with an increase in PA from baseline to final follow-up were reduced MVPA, increased incremental shuttle walk test distance, increased Hospital Anxiety and Depression Scale anxiety score and living with a child >18 years (p<0.05). CONCLUSIONS: While participation in the REACH-HF home-based cardiac rehabilitation intervention did not increase overall weekly activity, patient's behaviour patterns appeared to change with increased weekday PA levels and reduced inactivity. Baseline PA levels were highly predictive of PA change. Future focus should be on robust behavioural changes, improving overall levels of objectively assessed PA of people with HF. TRIAL REGISTRATION NUMBERS: ISRCTN78539530 and ISRCTN86234930.