7,000 steps, not 10,000

A 2025 meta-analysis of 57 studies found that 7,000 steps a day was associated with 47% lower all-cause mortality compared with 2,000 steps. Going from 7,000 to 10,000 added very little for most outcomes. The 10,000 figure was never a clinical threshold to begin with.

What did the 2025 step-count meta-analysis find?

The analysis, published in The Lancet Public Health and led by Professor Melody Ding at the University of Sydney, pooled 57 studies published between 2014 and 2025 across eight health outcomes.

Compared with a baseline of 2,000 steps per day, 7,000 steps was associated with 47% lower all-cause mortality (hazard ratio 0.53) and 38% lower dementia risk (HR 0.62). Cardiovascular disease incidence was 25% lower (HR 0.75) and cancer mortality 37% lower (HR 0.63). For most of these outcomes the curve flattened close to 7,000, and raising the target to 10,000 added little.

Type 2 diabetes was the exception. Its reduction at 7,000 steps was more modest (HR 0.86), and unlike the others the dose-response kept improving beyond that point rather than levelling off.

These are observational cohort findings, so they show association rather than causation, and people who walk more differ from people who walk less in ways that statistical adjustment cannot fully remove.

Where did 10,000 steps come from?

Not from a trial. The figure is widely attributed to a 1960s Japanese pedometer marketed as the manpo-kei, literally “10,000-step meter”, chosen in part because the character for 10,000 resembles a walking figure. It became a global default without ever having been tested as a threshold.

What the 2025 data shows is that the steepest part of the benefit curve sits well below it. The largest gains come early: moving from 2,000 to somewhere between 5,000 and 7,000 steps produced most of the observed improvement, and even 2,000 to 4,000 was associated with measurable change.

That matters because the people with the most to gain are the ones furthest from 10,000, and a target they read as unreachable is a target they do not start. Near the bottom of the curve, which exercise you pick matters far less than whether you are moving at all.

The bottom line

• In a 57-study meta-analysis, 7,000 daily steps was associated with 47% lower all-cause mortality and 38% lower dementia risk versus 2,000 steps.
• For most outcomes the curve flattens near 7,000. Type 2 diabetes was the exception, continuing to improve past it.
• 10,000 steps originated as 1960s pedometer branding, not as a clinical recommendation.