7,000 steps per day is a worthy goal

In a previous post I described the somewhat arbitrary origin of the 10,000 steps per day recommendation:

“It basically started around the Tokyo Olympics” in 1964, said Catrine Tudor-Locke, a professor who studies walking behavior at LSU’s Pennington Biomedical Center. “A company over there created a man-po-kei, a pedometer. And man stands for ‘10,000,’ po stands for ‘step,’ and kei stands for ‘meter’ or ‘gauge.’” Ten thousand, it turns out, “is a very auspicious number” in Japanese culture, said Theodore Bestor, a Harvard researcher of Japanese society and culture, in an email. “That is, it seems likely to me that the 10,000 steps goal was subsidiary to having a good-sounding name for marketing purposes.” Whatever the reason for the adoption of this particular number, “It resonated with people at the time, and they went man-po-kei-ing all over the place,” said Tudor-Locke.

A recent systematic review reinforces this point, and instead argues for a modestly lower target: 7,000 steps/day. Do we lose much taking 3,000 fewer steps?

The interest in steps is because they are a straightforward, trackable volume metric that correlate with a broad array of outcomes – all-cause mortality; cardiovascular disease (incidence and mortality); cancer (incidence and mortality); type 2 diabetes; dementia; depressive symptoms; and even falls. Because modern wearable devices tally steps in daily life, they offer a common yardstick across studies and populations.

The above-mentioned review synthesized 57 studies drawn from 35 cohorts, with 31 studies (24 cohorts) eligible for meta-analysis (some studies used the same cohort data). Using random-effects dose–response models, it mapped how risk changes across the step-count spectrum rather than only comparing arbitrary categories. The researchers fit a dose response curve to the data from the various cohorts while taking into account the heterogeneity between the different studies.

Hazard ratios were calculated for a number of outcomes comparing those who had taken ~7,000 steps/day versus ~2,000 steps/day, which represents the lower limit of the typical range for older adults:

• For all-cause mortality, ~7,000 steps/day was associated with a hazard ratio (HR) of 0.53, implying ~47% lower risk (95% CI 0.46–0.60). 
• For cardiovascular disease (CVD) incidence, the pooled HR was 0.75 – about a 25% lower risk (0.67–0.85).
• For CVD mortality, the estimate was HR 0.53 -- ~47% lower risk (0.37–0.77).
• For cancer incidence, the pooled HR at ~7,000 steps/day was 0.94 -- ~6% lower risk (0.87–1.01).
• For cancer mortality, the HR was 0.63 -- ~37% lower risk (0.55–0.72). It is interesting that there was a bigger impact on cancer mortality than incidence. 
• For dementia, the HR was 0.62 -- ~38% lower risk (0.53–0.73), a robust decrease.

The 2,000 and 7,000 steps/day represent two points on the steps versus health outcome (HR) dose-response curves (Figure 1). The shapes of the curve were quite informative, especially where the health benefits start to flatten. For all-cause mortality, CVD incidence, dementia, and falls, risk curves are non-linear and inverse -- substantial early gains as steps rise from very low levels, followed by slower improvements. The inflection/plateau zone typically appears around 5,000–7,000 steps/day. For CVD mortality, cancer incidence and mortality, type 2 diabetes, and depressive symptoms, patterns are approximately linear across observed ranges—each additional increment of steps continues to reduce risk without an obvious inflection. Practically, moving from ~2,000 => 4,000 =>7,000 steps/day delivers large, tangible drops in risk; going beyond ~7,000 remains helpful, but each extra 1,000 steps usually buys less additional benefit than the early increases.

This flattening raises the question of what step count is optimal. The authors suggest that a "sweet spot" sits around ~7,000 steps/day, balancing sizeable relative risk reductions with broad achievability. It captures most of the non-linear gains while being realistic for many adults. For performance-minded or highly active individuals, 8,000–10,000+ steps/day is a reasonable stretch, as incremental benefits persist for several outcomes even after 7,000. The key is recognizing that the marginal returns per extra 1,000 steps tend to shrink. If your baseline is ~2,000–3,000 steps/day, adding ~1,000–2,000 steps is a powerful first move that markedly improves risk long before 10,000 becomes relevant. The advice is to build upward in manageable increments.

Some caveats of the meta-analysis include the observational nature of the studies -- not randomized trials -- so causality can’t be proven. Indeed, reverse causation is possible: people with subclinical illness or mobility limitations may walk less because they are unwell, which can inflate the apparent benefits of higher steps. In addition, residual confounding (diet, socioeconomic factors, comorbidity, medications, neighborhood walkability) may bias associations. There is also heterogeneity across cohorts (population age, outcome definitions, follow-up duration, covariate adjustment) meaning that a single “optimal” number will always be context-dependent. Finally, measurement error from different tracking devices and activities is an issue along with the question of step intensity and type (e.g. climbing stairs).

 

Nevertheless, the cumulative data argue strongly that ~7,000 steps/day delivers substantial benefits for most adults without demanding athletic lifestyles, and it aligns with where several non-linear curves begin to flatten. As a stretch, 8,000–10,000+ steps/day can produce additional -- albeit smaller -- reductions for multiple outcomes, suitable for those who can and want to do more. The governing principle is that every step counts, especially when moving up from low baselines. In other words, don’t let an aspirational 10,000 steps/day become a barrier to capturing the health benefits at more attainable levels.

Figure 1. Dose-response curves showing association between steps/day and risk of various health conditions. The x-axis denotes steps per day, and the y-axis is the hazard ratio. Data are from a meta-analysis of 31 studies. The vertical line marks the 2,000 steps/day reference group (Ding et al. The Lancet Public Health, 2025).