Daily Steps and Lifespan: Where the 10,000 Target Came From

The 10,000-steps-a-day target is the default on many fitness trackers and is widely assumed to have come out of a study. It came out of a slogan. The cohort studies that have since measured steps with accelerometers and followed people for years tell a more useful story: mortality risk falls steeply as daily steps rise from very low levels, then flattens somewhere between about 6,000 and 10,000 steps depending on age.

A slogan, not a study

Catrine Tudor-Locke and David Bassett traced the origin in a 2004 Sports Medicine review: the 10,000 figure, they wrote, “can be traced to Japanese walking clubs and a business slogan” from more than three decades earlier. The Japanese word attached to it, manpo-kei (roughly “ten-thousand-step meter”), is where a 2023 British Journal of Sports Medicine editorial by Stamatakis and colleagues starts its history of step targets.

They did not dismiss the number: their review called 10,000 steps a sensible ballpark for apparently healthy adults, possibly unsustainable for older adults or people with chronic disease, and likely too low for children. But the evidence behind it was cross-sectional associations and a few short interventions measuring body composition and blood pressure; none of it followed people for years to see who died.

What the cohorts show

Older women. Lee and colleagues (JAMA Internal Medicine, 2019) gave accelerometers to 16,741 women, mean age 72, for seven days and followed them for a mean of 4.3 years, during which 504 died. Quartile median steps were 2,718, 4,363, 5,905 and 8,442 a day, with adjusted mortality hazard ratios of 1.00, 0.59, 0.54 and 0.42. About 4,400 steps a day was already associated with significantly lower mortality than about 2,700, and the hazard kept declining until roughly 7,500 steps, then levelled off.

US adults. Saint-Maurice and colleagues (JAMA, 2020) analysed 4,840 NHANES adults aged 40 and over, mean age 57, followed for about ten years. Unadjusted death rates ran from 76.7 per 1,000 person-years below 4,000 steps a day to 4.8 at 12,000 or more. Compared with 4,000 steps, 8,000 a day was associated with an adjusted hazard ratio of 0.49 and 12,000 with 0.35.

Fifteen cohorts pooled. Paluch and colleagues (Lancet Public Health, 2022) combined 15 cohorts covering 47,471 adults and 3,013 deaths over a median 7.1 years. Quartile medians were 3,553, 5,801, 7,842 and 10,901 steps a day; relative to the lowest quartile, the hazard ratios were 0.60, 0.55 and 0.47. The shape matters: risk fell progressively up to about 6,000 to 8,000 steps a day in adults aged 60 and over, and up to 8,000 to 10,000 in younger adults, with no further decline beyond that.

UK Biobank. Del Pozo Cruz and colleagues (JAMA Internal Medicine, 2022) analysed 78,500 UK adults aged 40 to 79, mean age 61, over a median seven years. More steps were associated with lower all-cause, cardiovascular and cancer mortality for up to roughly 10,000 steps a day, and with lower cardiovascular and cancer incidence.

Where the curve flattens

Several meta-analyses have tried to pin down the plateau and land in the same neighbourhood.

• Jayedi and colleagues (Sports Medicine, 2022) pooled seven cohorts: hazard ratio 0.88 per extra 1,000 steps a day, risk declining linearly from 2,700 to 17,000 steps, and a hazard ratio of 0.44 at 10,000 steps.
• Banach and colleagues (European Journal of Preventive Cardiology, 2023) pooled 17 cohorts and 226,889 people: 15 percent lower all-cause mortality per 1,000-step increment, with benefit appearing above roughly 3,900 steps a day for all-cause and 2,300 for cardiovascular mortality.
• Stens and colleagues (JACC, 2023) pooled 12 studies and 111,309 people and put the optimal dose at 8,763 steps a day for all-cause mortality (hazard ratio 0.40 versus 2,000 steps) and 7,126 for incident cardiovascular disease (0.49).
• Ding and colleagues (Lancet Public Health, 2025) is the largest, 57 studies from 35 cohorts. Compared with 2,000 steps a day, 7,000 was associated with 47 percent lower all-cause mortality (hazard ratio 0.53), 25 percent lower cardiovascular disease incidence, 38 percent lower dementia incidence and 28 percent fewer falls (very low certainty), with smaller associations for type 2 diabetes and depressive symptoms. Inflection points sat around 5,000 to 7,000 steps for mortality, cardiovascular disease, dementia and falls; the remaining outcomes looked linear.

Two things stand out. First, the biggest gains per step are at the bottom of the range: moving from 2,000 or 3,000 steps to 5,000 to 7,000 is where the curves are steepest. Second, the plateau sits lower in older adults (6,000 to 8,000 in Paluch, about 7,500 in Lee’s older women) than in adults under 60 (8,000 to 10,000). The 10,000 target is not wrong so much as the upper end of the range for younger adults and beyond the point of measurable return for older ones; Ding’s group called it still viable for the already active, with 7,000 more realistic for some. (See also VO2 max and lifespan.)

Does walking faster add anything?

On cadence the studies disagree.

In Lee’s older women, faster stepping looked protective on its own, but after adjusting for total steps every intensity measure weakened and most lost significance. In Saint-Maurice’s NHANES adults, no intensity measure was significantly associated with mortality after adjusting for total steps. Jayedi’s meta-analysis found a weaker per-1,000-step association (hazard ratio 0.92 rather than 0.88) in the two studies that controlled for intensity, suggesting some of the volume effect travels with pace.

The larger datasets found more. In the Paluch pooled analysis, after adjusting for daily steps, peak 30- and 60-minute cadence still carried a significant association (hazard ratio 0.67, top versus bottom quartile), though minutes spent at 40 or 100 steps per minute or faster did not. In UK Biobank, peak-30 cadence (average pace over a person’s 30 fastest, not necessarily consecutive, minutes) was linked to lower risk for every outcome over and above total steps. Stens’ meta-analysis found moving from a low to a moderate or high cadence independently associated with lower mortality.

The honest summary: total steps carry most of the association, and the fastest half-hour of the day appears to add a modest amount on top in the bigger datasets, while the two single US cohorts found no clear independent effect of pace. Zone 2 Cardio and Longevity covers where brisk walking sits on the intensity scale, and Cardio and Heart Risk handles intensity as a separate lever.

Does it have to be every day?

Inoue and colleagues (JAMA Network Open, 2023) asked a different question of NHANES data from 3,101 adults: on how many days of the monitoring week did they reach 8,000 steps? Compared with people who never did, those who reached it on one or two days had an adjusted absolute ten-year mortality risk 14.9 percentage points lower; three to seven days was 16.5 points lower, and the association plateaued at about three days a week, after adjusting for average daily step count. One week of monitoring is a thin basis for a ten-year prediction, but it fits the pattern: clearing the floor matters more than daily consistency.

Beyond mortality, Master and colleagues (Nature Medicine, 2022) linked Fitbit data from 6,042 All of Us participants (median 7,731 steps a day over four years) to their health records. Higher counts were associated with lower incidence of obesity, sleep apnoea, reflux and major depressive disorder, while diabetes and hypertension risk showed no further reduction above 8,000 to 9,000 steps. That cohort was 73 percent female, 84 percent white, 71 percent college-educated and made up of existing Fitbit owners: a narrow slice of the population.

What this kind of data cannot tell you

All of the studies above are observational; no trial has randomised thousands of adults to 4,000 versus 8,000 steps a day and followed them for a decade. The usual problems apply:

• Reverse causation. People who are already ill walk less and die sooner; cohorts adjust for baseline health and diagnoses, but undiagnosed illness can drive both.
• Confounding. Saint-Maurice’s models adjusted for diet, smoking, body mass index, education, self-rated health and existing diagnoses, but adjustment is only as good as the variables measured, and the Ding review names residual confounding as a limitation.
• One week of measurement. Lee, Saint-Maurice and Inoue each measured steps for about seven days at baseline, then followed people for four to ten years.
• Device and population differences. Stens found stronger risk reductions with hip-worn accelerometers than with pedometers or wrist devices, and a threshold derived in 72-year-old women does not automatically transfer to a 35-year-old man.

Consistency across cohorts, countries and ages, plus the dose-response shape, makes a causal contribution plausible; Jayedi’s group graded the certainty strong because the effect was large and dose-dependent. But the hazard ratios are associations, not the reduction any individual would get by adding steps.

Bottom line

The 10,000-step target came from a slogan, not a trial, and the trials that would settle it do not exist. The observational evidence says the steepest mortality gains come from moving out of the 2,000 to 4,000 range, that risk keeps falling to about 6,000 to 8,000 steps a day after 60 and 8,000 to 10,000 before it, and that pace adds a modest amount once volume is counted. A plateau is not a cutoff: in Jayedi’s analysis risk kept falling in a roughly straight line to 17,000 steps. But if 10,000 has felt like a pass-fail line, a large share of the measured association is already present by about 7,000, the figure the 2025 review called a more realistic target for some.