Bella Struminskaya, Florian Keusch
Measurement of physical activity in older adults through data donation

Data Donation Symposium 2024, Amsterdam, May 30th to May 31st, 2024

Physical activity (PA) is a key predictor of many health outcomes, especially for aging populations. The accurate measurement of PA is key to identifying determinants of health and developing appropriate interventions. To measure PA, most population studies use self-report. However, selfreports are usually limited to global measures of PA (e.g., average daily hours of moderate/vigorous activity) and suffer from misclassification (e.g., walking the dog not considered PA). More finegrained day-reconstruction methods are burdensome for respondents and prone to recall error. As an alternative researchers are providing study participants with wearable devices that passively track PA, which reduces reactivity and recall error. However, participants’ non-compliance and high device costs are problematic. Many older adults now have smartphones that track physical activity and individuals can share these passively collected physical activity data with researchers. In this study, we test a data donation approach among older adults. Based on legal requirements such as the General Data Protection Regulation (GDPR), all data collecting companies and digital platforms need to provide users access to all of their data in a machine-readable format. We use a privacy-preserving data donation tool (Boeschoten et al. 2023) integrated in a probability-based online panel of the Dutch general population to collect PA data from health apps and Google Semantic Location History. We investigate determinants of consent and selection bias in PA data donation among ca. 2,000 individuals aged 50 and older in the Netherlands. Panel members who own a smartphone are asked to locate the in-built health apps on their smartphone (iPhone or Samsung) and perform the data request. Smartphone owners of other Android devices are asked to request Google Semantic Location History data from Google. Using the rich data available about panelists (from the last 15 years), we assess the quality of the donated PA data and evaluate how well multi-source physical activity data can predict health outcomes. Our study contributes to the development of future-proof methods for collecting high-quality PA data and innovations in surveys. Having self-report measures on technological skills (e.g., smartphone functions used) as well as measures on digital literacy already present in the LISS Panel and objective measures (i.e., donation outcome) will allow us to provide recommendations about using technologies such as apps and sensors for data collection in older populations.