Data Donation as a Method to Measure Physical Activity in Older Adults: Cross-Sectional Web Survey Assessing Consent Rates, Donation Success, and Bias.

Abstract

Background: Accurate measurement of physical activity (PA) is key to identifying determinants of health and developing appropriate interventions. Self-reports of PA (eg, in surveys or diary studies) often suffer from measurement error. Providing study participants with wearable devices that passively track PA reduces reactivity and recall error but participants' noncompliance and high device costs are problematic. Many older adults now have smartphones that track PA. Based on legal requirements, data controllers (eg, health apps) must provide users with access to their data, and individuals can request and donate these data for research. This user-centric approach provides researchers with access to individual-level data, and it gives users control over what data are shared.

Objective: We conduct a first test of the data donation approach for PA data among older adults. We study (1) how willing and successful older adults are to donate their PA data from different smartphone apps, (2) what drives donation of PA data at the different stages of participation, and (3) what biases arise from selective data donation.

Methods: To answer our research questions, we use cross-sectional observational data from a probability-based online panel of the Dutch general population. A total of 2086 members of the Longitudinal Internet Studies for the Social Sciences panel aged 50 years and older completed a web survey in 2024. All iPhone and Android smartphone owners were asked to download passively collected PA data from their devices (Apple Health, Google Location History, or Samsung Health) and donate them via the Port platform.

Results: Out of the 2086 survey participants, 1889 (91%) reported owning an iPhone or Android phone compatible for data donation, 606 (29%) reported willingness to donate PA data, 354 (17%) started the data donation, and 256 (12%) successfully provided a data package. Gender, age, educational attainment, monthly personal net income, smartphone usage behavior, privacy- and trust-related attitudes, and type of health app from which the data were requested correlated with behavior at the different stages of study participation. Self-reported reasons for nonwillingness to donate related mainly to expected technical issues, privacy concerns, and perceived usefulness. Compared with the entire sample, data donors reported better health, fewer health-related limitations, fewer difficulties performing tasks, and more PA.

Conclusions: Our study shows that data donation from smartphones as part of a probability-based web survey of older adults is a feasible alternative for the measurement of PA, especially for iPhone owners younger than 70 years. Limitations relate to nonparticipation which correlates strongly with characteristics of smartphone ownership and comfort with device use. Substantive bias in health and PA outcomes persists for those who donated in comparison with all survey respondents.