Estimator built from Kato (2026), SSM – Population Health
Daily steps and 6-month healthcare expenditure
Inputs
0
4074
Predicted 6-month healthcare expenditure
—JPY
95% CI—
Other covariates held at sample median/mode:
Fitted surface (steps × age), as in Fig. 4
Red point: your input, with its 95% CI as a vertical bar. Drag to rotate, scroll to zoom.
How to read this
An observed association in one selected cohort (long-term A-Smile users aged 40+ in Osaka, 2020–2023), not the effect of changing a person's steps.
A back-transformed log-model prediction (geometric-mean type), far below the arithmetic mean; not for budget totals. The 95% CI is point-wise for the curve, not for an individual.
Ages 40–74 and 0–20,000 steps/day (the paper interprets the curve only within the data-dense range, 845–14,906 steps/day); other covariates fixed at the sample median/mode.