You are here

Article Text

Article menu

Download PDFPDF

38 Assessment of device-measured physical activity and associated clinical outcomes in cardiac rehabilitation using a novel statistical tool
Free
  1. Melina Del Angel1,
  2. Dylan Thompson1,
  3. Matthew Nunes2,
  4. Michael Fisher3,
  5. Ewan J Cranwell4
  1. 1Department for Health, University of Bath, Claverton Down, Bath, BA2 7AY, UK
  2. 2School of Mathematical Sciences, University of Bath, Claverton Down, Bath, BA2 7AY, UK
  3. 3Liverpool University Hospitals NHS Trust, Prescot St, Liverpool, L7 8XP and the Liverpool Centre for Cardiovascular Science, University of Liverpool, Liverpool, UK
  4. 4KiActiv
  5. ®, 1 Duchess Street, London, W1W 6AN

Abstract

Background Physical activity (PA) is a cornerstone of Cardiac Rehabilitation (CR) and is vital for improving health outcomes, including cardiorespiratory fitness (CRF). Wearable devices can be used to obtain objective PA data but, currently, statistical tools to examine and understand longitudinal PA data during CR are lacking.

Aim To apply novel statistical tools to assess device-based PA data and better infer associated clinical outcomes during CR.

Methods This secondary analysis used device-measured PA data from a recent digital CR randomised controlled trial (RCT) that increased CRF by 2-fold compared to usual CR alone. In this RCT, 130 patients were randomised to either an intervention group (N=65) or control group (N=65). All patients wore a wrist-mounted accelerometer for 8 weeks and the number of minutes spent in differing PA intensities at an hourly resolution was analysed. Patients were excluded who had >5 incomplete days, resulting in a sample of 53 patients (control N=31; intervention N=22). A day was considered complete if the accelerometer was worn for ≥80% of the waking day. PA data was assessed using the Trend Locally Stationary Wavelet (TLSW) model.

Results Sedentary time was elected as the PA outcome of interest. In figure 1, intervals where the confidence bands do not overlap indicate where sedentary time was statistically significantly different between groups. The novel TLSW tool successfully characterised the intervention response, indicating that the intervention group spent 41% of the intervention period (equivalent to 26 days) engaged in statistically significantly lower sedentary time than controls.

Conclusion The application of a novel TLSW tool to assess device-measured PA enhances understanding of PA behaviour change during CR and has highlighted the significant importance of reducing sedentary time for improving CRF. This presents new opportunities to optimise CR outcomes by harnessing everyday PA at all intensities, to deliver more personalised care.

Abstract 38 Figure 1
Abstract 38 Figure 1

Standardised trend estimations at hourly resolution for sedentary time. A trend below zero indicates an individual decreased the number of minutes spent sedentary per hour. Control (N=31) and Intervention (N=22) mean trend lines are displayed along with their 95% confidence bands (grey shaded areas). Intervals where the confidence bands do not overlap (i.e., white space between the shaded grey areas) indicate where performance (in this case, sedentary time) between groups is statistically significantly different

https://doi.org/10.1136/heartjnl-2024-BACPR.38

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.