Association of Implantable Device Measured Physical Activity With Hospitalization for Heart Failure

Abstract Objectives Evaluate the association of physical activity (PA) level and longitudinal PA trajectory with a composite heart failure hospitalization and mortality endpoint over a 5-year follow-up period following implantation. Background Low device measured PA early after implantation of an implantable cardioverter-defibrillator (ICD) or cardiac resynchronization therapy defibrillator (CRT-D) is associated with poor outcomes. Methods We linked daily PA data from the Boston Scientific ALTITUDE dataset of patients with ICD or CRT-D implantation to Medicare claims data. We used a joint model to investigate the association of the composite endpoint with 1) the time-varying point estimate of PA and 2) the time-varying trajectory/slope of PA during follow-up. Results Among 20,927 patients with median activity level 85 min/day, 14.1% and 49.6% experienced the composite endpoint at 1 and 5 years. Adjusted joint model results showed that there was a 1.13 (95% confidence interval: 1.12 to 1.13)-fold increase in the hazard of the composite endpoint for 75 min of daily PA relative to 85 min of PA; and a within-patient 10-min decrease in average daily PA over an 8-week period from 85 to 75 min was associated with a hazard ratio of 4.02 (95% confidence interval: 3.82 to 4.22) for the composite endpoint. Conclusions Patients with large decreases in PA have significantly higher risk of experiencing heart failure hospitalization or death. PA data from implantable devices may identify patients before clinical decompensation.