2025 Volume 72 Issue 12 Pages 942-950
Objectives Recently, step-count, as an objective measure of physical activity, has become available on smartphones and is recorded by default on an iPhone (Apple Inc.). Using screenshot images of the step-count graph of the iPhone healthcare application, an image analysis tool was developed to enable a large-scale and retrospective physical activity assessment. However, the characteristics of the respondents involved in the collection of such step-count images remain unclear. We aimed to investigate the characteristics of iPhone users and respondents who provided step-count images via an Internet survey.
Methods We conducted an Internet survey in December 2022 among 5,346 Kobe residents aged ≥ 20 years from a survey company’s panel population. They previously responded to a web-based survey on physical activity, health, and community environment administered in 2021. The participants were asked to provide their demographics and walking time. iPhone users were additionally asked to take and upload two screenshots of step-count graphs on their iPhones; one-month and one-year graphs. We conducted multivariable logistic regression analysis to examine the factors associated with iPhone use and image provision.
Results Among 3,308 respondents, 349 (11%) provided images, 1,138 (34%) were iPhone users who did not provide images, and 1,821 (55%) did not use an iPhone. Those aged ≤ 64 years exhibited a higher likelihood of iPhone usage compared to older adults (odds ratio; 2.45, [95% confidence interval; 2.04–2.94]), as did women compared to men (1.47 [1.26–1.71]), those with a household income of ≥ 6 million JPY compared to those with lower income (1.39 [1.20–1.62]), and those with a college degree or higher compared to the others (1.17 [1.01–1.36]). Furthermore, those aged ≤ 64 years were more likely to provide step-count images than older adults (3.58 [2.17–5.90]), as were those with a college degree or higher compared to the others (1.71 [1.32–2.23]) and those with longer walking time compared to those with shorter walking time (1.66 [1.21–2.28]).
Conclusion Age, sex, household income, and education were associated with iPhone use, whereas age, education, and physical activity level were associated with step-count screenshots. Possible biases must be considered among iPhone users and image providers in collecting and using step-count images in Internet surveys.
