What the research says about steps

A plain-language page for walkers. Every number below comes from a published study, cited with its DOI. Findings are associations from cohort studies, not promises: they describe what was true of large groups of people, not what will happen to you. Nothing here is medical advice.

Where 10,000 came from

The 10,000-step target began as a product name. In 1965 Yamasa Tokei Keiki sold a pedometer called the manpo-kei, 万歩計, literally “10,000-step meter”, in the enthusiasm around the 1964 Tokyo Olympics. Dr Yoshiro Hatano’s work on energy expenditure later gave the number a rationale, but no study picked 10,000 by testing it against other targets. The researchers who have tested step counts since, among them Dr I-Min Lee, describe it as a marketing figure that happened to be memorable.

How many steps

  • Older women, 2019. In 16,741 women averaging 72 years old, mortality was markedly lower at about 4,400 steps a day than at about 2,700, kept falling with more steps, and levelled off around 7,500. (Lee et al., JAMA Internal Medicine 2019, doi:10.1001/jamainternmed.2019.0899)
  • Fifteen cohorts, 2022. In 47,471 adults, the risk of death fell progressively with more daily steps and levelled off at about 6,000 to 8,000 a day for adults 60 and older, and about 8,000 to 10,000 for adults under 60. (Paluch et al., Lancet Public Health 2022, doi:10.1016/S2468-2667(21)00302-9)
  • Seventeen cohorts, 2023. In 226,889 people, the association with lower all-cause mortality became significant above about 3,967 steps a day, and with cardiovascular mortality above about 2,337. Each additional 1,000 steps a day was associated with roughly 15 percent lower all-cause mortality. (Banach et al., European Journal of Preventive Cardiology 2023, doi:10.1093/eurjpc/zwad229)
  • Fifty-seven studies, 2025. Across about 160,000 adults and outcomes from mortality to dementia and depressive symptoms, 7,000 steps a day was associated with 47 percent lower all-cause mortality than 2,000, and most associations weakened or plateaued around 7,000. (Ding et al., Lancet Public Health 2025, doi:10.1016/S2468-2667(25)00164-1)

Read together: benefit starts low, rises steeply through the middle thousands, and levels off somewhere between 7,000 and 10,000 depending on age and outcome. More is not harmful in these studies; it is simply less additional.

Does speed matter?

Less than volume. In both the 2019 and 2022 studies, once total steps were accounted for, step intensity was not associated with mortality. Taking the steps is what the evidence rewards; taking them fast is optional.

Why rest days work

Goals with a little slack are kept longer. In a series of experiments, people preferred a goal such as “seven days, with two emergency skips” over an otherwise equivalent easier goal, and persisted at it longer than at a rigid one, especially after a setback. The reserve is meant to feel costly to use, which is why it motivates rather than excuses. (Sharif and Shu, Journal of Marketing Research 2017, doi:10.1509/jmr.15.0231)

A walk after eating

Breaking up sitting with light walking lowered the rise in blood glucose after meals by about 1.0 mmol/L and insulin by about 5 mU/L, compared with staying seated, across five randomised crossover trials. The studies did not establish a best time to walk; light walking after meals is what they tested. (Buffey et al., Sports Medicine 2022, doi:10.1007/s40279-022-01649-4)

Why your number differs between devices

Consumer devices count steps differently, and Apple Health merges the sources it is given, removing overlap where two devices recorded the same minutes. That is why a phone, a watch and a third-party app rarely agree, and why a total can change after a watch syncs. It is a property of the sources, not a fault in any one of them.

What Manpo Steps does with this

  • The default goal is 8,000, in the band where the evidence levels off, and you can set any number.
  • Rest days are allowed and extend a streak, because slack with a cost keeps people walking.
  • The app shows the time of its reading and whether it matches Apple Health, and explains the arithmetic behind the number on request.
  • It never grades your pace. Steps are the unit.