The Lancet Regional Health. Western Pacific · Published 2026-06-19 · DOI 10.1016/j.lanwpc.2026.101901
Kazuaki Arai, Masamitsu Kamada, Takahiro Yoshida, Ikuho Yamada, Ding Ding, Kimihiro Hino
Summary: Background: Although regional inequalities in physical activity are recognised as a global concern, nationwide evidence from device-based measurements is limited. Given the multilevel nature of inequality, we quantified step count disparities across municipalities and examined cross-level interactions between individual-level socioeconomic status (SES) and walkability. Methods: We conducted municipality- and individual-level analyses using step count data from 1,542,536 smartphone users aged 20–64 years. We calculated municipality-level mean daily steps and Gini coefficients, and examined associations between mean daily steps, municipal walkability, and Gini coefficients. Linear mixed-effects models quantified cross-level interactions between individual SES (education, employment, and equivalised household income) and walkability (modelled as a continuous variable) on individual-level mean daily steps. Cross-level patterns were visualised across quintile-based walkability strata. Findings: Municipality-level mean daily steps varied substantially across the 951 municipalities, with a maximum difference of 3724 steps/day (ranging from 4026 to 7750 steps/day, nearly a 2-fold disparity). Municipalities with higher mean daily steps had greater walkability (r = 0.76) and more equal step-count distributions (Gini coefficient; r = −0.79, stronger than simulated mathematical correlations). Significant cross-level interactions on individual-level mean daily steps were evident, with patterns differing across SES domains: step-count differences across walkability strata were prominent for education and income (623–762 steps/day), whereas employment status showed substantial differences between employed and non-employed individuals, reaching 826–1514 steps/day in highly walkable areas. Interpretation: Substantial inequalities exist across and within municipalities, and individual-level disparities vary based on SES and walkability. Nationwide surveillance and cross-sectoral collaborations are needed to identify populations vulnerable to inactivity and implement targeted strategies. Funding: Japan Society for the Promotion of Science and Japan Science and Technology Agency.
Abstract from DOAJ. Public domain (CC0 1.0).
Read the article at the publisher →
Arai, K., Kamada, M., Yoshida, T., et al. (2026). Cross-level regional and socioeconomic inequalities in step counts: a descriptive epidemiological investigation of 1.5 million Japanese smartphone users. The Lancet Regional Health. Western Pacific. https://doi.org/10.1016/j.lanwpc.2026.101901