Advancing the understanding and translational potential of physical activity tools and prescription
Why the research is needed
Regular physical activity is beneficial to health, but many people are not active enough. Wearable accelerometers are part of a rapidly growing trend in medicine and are increasingly being used in large scale studies to better understand relationships of physical activity with health outcomes, and in clinical practice to facilitate behaviour change. However, the complexity of data from wearable accelerometers can be challenging to understand. The use of simpler stepping-based metrics and messages may therefore be better received and actionable by both health practitioners and the general public, and facilitate better discussions and engagement within clinical services.
What is already known about the subject
Data from accelerometers are generally more accurate than self-reported measures of physical activity. This advancement in measurement capacity has allowed greater precision in linking physical activity with health outcomes and helping to further target physical activity messaging. In addition, simple step counts can be estimated from most wearable activity trackers, to varying degrees. and are more easily understood by clinicians and patients. However, in order to better estimate and translate research accelerometer metrics into stepping-based metrics, we first need to develop methods for extracting steps from research-grade accelerometers and develop a framework to assess their validity.
Who we will be working with
Public and patient involvement and engagement (PPI/E) events facilitated through The DRC have been fundamental in formulating and refining this proposal and will also be built upon in the later stages of this project to inform implementation. The overriding message and ‘need’ resonating from both healthcare practitioners and patient groups so far was that simpler physical activity metrics were helpful and that stepping or walking-based messages were the easiest to understand and take action upon in both clinical and real-life settings. We are also working with algorithm developers to refine our methods.
How patients and the public are involved
PPI/E members and clinical/patient groups already have and will continue to contribute to this proposal and the portfolio of lifestyle research that is undertaken at UoL. Representatives will inform the dissemination of findings through co-production of contextually relevant and culturally sensitive public- and clinician-facing materials and summary reports, infographics and social media channels (see below for more details). Through our links with patients and clinicians and the Centre for Ethnic Health Research, we will also gather further feedback and qualitative data on how clinicians and the general public understand different forms of physical activity messaging (including our derived stepping metrics), and ways this understanding can be improved to promote action and linkage with wearable devices for more tailored feedback, engagement, and effectiveness of behaviour change interventions.
What we will do
This research aims to understand whether wearable data on physical activity behaviours can be harnessed to provide more individualised information on disease risk and quantify how behaviour change may act to change this risk. The overall aim is to assess the validity of a step detection algorithm for wrist-worn accelerometers to inform more translatable PA outcomes and tools for both research and clinical contexts. We will achieve this through 3 x key objectives:
- Assess the accuracy and validity of an algorithm to derive stepping-based metrics from wrist-worn accelerometer data, using a thigh-based accelerometer as a benchmark.
- Develop a conversion Table that allows acceleration metrics (e.g. ENMO) to be easily/quickly converted into stepping-based metrics (i.e. total amount and cadence/intensity), allowing for easier interpretation in health research and more tailored/effective activity prescriptions.
- Investigate the associations of these step-based metrics with all-cause mortality, incident CVD and type 2 diabetes, and multimorbidity outcomes in UK Biobank.
What the benefits will be
Deriving step-based metrics from research-grade accelerometers will allow us to better translate the relationships with health outcomes (e.g. “every 500 steps/day difference in activity is associated with a specific difference in the risk of a type 2 diabetes incidence”) in turn allowing clinicians and public health professionals to provide more understandable messages to the public. This will enable more informed conversations between patients and HCP’s, allowing for better shared decision-making and care planning approaches. This will also inform how wearable devices are used in future interventions to help with increasing activity.
How we are planning for implementation
Findings will be reported to key stakeholders as evidence of the effects of stepping volume and intensity with health. Findings will also be shown to patient-public involvement (PPI) groups to refine messages for patient and public benefit, including messaging within the App to make it culturally sensitive and relevant. This evidence will also be embedded into future behavioural change research conducted at the Leicester Diabetes Centre as educational material. Study results will be submitted for publication in an open-access peer-reviewed journals and presented at international conferences.
When the findings will be available
By April 2023
Paddy Dempsey, firstname.lastname@example.org.
Ben Maylor, email@example.com.