Health literacy and response to a mobile app–based exercise programme in heart failure: Insights from the EXERCISE-HF trial

Background The EXERCISE-HF trial demonstrated that a multidisciplinary, mobile app-based exercise-support programme improves functional capacity and skeletal muscle strength in patients with heart failure (HF). The application integrates structured education and behaviour-change support to facilitate patient-centred engagement. Baseline health literacy (HL) may therefore influence responsiveness to this intervention. In this sub-study of EXERCISE-HF, we aimed to evaluate whether baseline HL modified changes in peak oxygen uptake (VO2). Methods In total, 100 patients with HF were randomized to a wearable-integrated, mobile app–based exercise programme or usual care. HL was measured using the Health Literacy Survey European Questionnaire-47 (HLS-EU-Q47), generating a general index and three domain-specific indices (healthcare, disease prevention, health promotion), each scored 0–50 based on core HL competencies. The outcome of interest was percentage change in peak VO2 at 24 weeks. This was interpreted in light of evidence indicating that a 6% increase in peak VO2 is associated with reduced clinical risk in HF. Multivariable linear regression models included allocation group, HL level, and their interaction, adjusted for age, sex, peak VO2(baseline), left ventricular ejection fraction (LVEF), and N-terminal pro-B-type natriuretic peptide (NT-proBNP). Results Of the 100 patients enrolled in EXERCISE-HF, nine were excluded because of missing cardiopulmonary exercise testing or HLS-EU-Q47 data. The final cohort comprised 91 patients (median age 63 years, IQR 57–71; 74% male; 44 intervention, 47 control). Median LVEF was 47.8% (IQR 41.4–59.8) and median NT-proBNP was 447 pg/mL (IQR 239–1,005). Median baseline HLS-EU-Q47 scores were 28 (IQR 24–34) for general HL, 29 (IQR 23–34) for healthcare, 32 (IQR 26–38) for disease prevention, and 28 (IQR 20–34) for health promotion, with no significant between-group differences. In adjusted analyses, the healthcare HL domain significantly modified the effect of the intervention on percentage change in peak VO2 at 24 weeks (interaction estimate per 10-point increase: 7.30%; 95% CI 0.66–13.94; p=0.032). No significant interaction was observed for the other HL domains. Conclusions Healthcare-specific HL significantly modified response to a mobile app–based exercise programme in patients with HF. Greater improvements in peak VO2 were observed at higher healthcare HL levels. Baseline assessment of patients’ capacity to understand and apply healthcare information may enhance the effectiveness of mobile app-based exercise interventions in HF.