How many steps a day actually matter? Fewer than 10,000, and it depends on your age

The clearest answer comes from pooling 15 cohorts that measured steps with devices rather than questionnaires (Paluch 2022). Against the least active quarter, hazard ratios for all-cause mortality were 0.60 at a median of 5,801 steps a day, 0.55 at 7,842 and 0.47 at 10,901. The curve flattened at about 6,000–8,000 steps for people 60 and over, and at about 8,000–10,000 for those under 60.

Two things follow. The biggest gains are at the bottom: moving from about 3,500 steps to about 6,000 was associated with 40% lower risk. And the ceiling is lower with age, which fits a simple reading: for an older adult, 7,000 steps is a lot of activity relative to capacity.

Steps are a proxy for total movement. When activity is measured by accelerometer and split into quarters, the pattern is the same and larger: hazard ratios of 1.00, 0.48, 0.34 and 0.27 from the least to the most active quarter, so the most active quarter had about a quarter of the risk of the least active (Ekelund 2019). Light activity counted, not only brisk walking.

The questionnaire literature agrees on the shape. Against no leisure-time activity, hazard ratios were 0.80 below the recommended minimum, 0.69 at one to two times the minimum, 0.63 at two to three times, 0.61 at three to five times, and 0.69 at ten times or more (Arem 2015). Most of the benefit comes with the first 150 minutes a week, with no sign of harm well beyond it.

The WHO’s 2020 guideline is in minutes, not steps: 150–300 minutes of moderate or 75–150 minutes of vigorous aerobic activity a week, or a combination, plus muscle strengthening (WHO 2020). A brisk 30-minute walk is roughly 3,000–4,000 steps for most adults, so the lower end of the guideline, spread over a week, lands near the bottom of the step curve’s useful range. The step studies suggest the guideline is a floor, not a ceiling.

Why the two measures differ: steps count everything, including slow walking around the house, while the minutes guideline counts moderate effort. Both are linked to lower mortality. Neither tells you which caused what.

All of these are observational. People who walk more may be healthier for reasons the adjustments do not capture, and people who are already ill walk less, which inflates the apparent benefit at the low end. The pooled step analysis adjusted for the usual covariates and excluded early deaths in sensitivity analyses, which reduces but does not remove that problem.

The step counts also depend on the device and where it was worn. A wrist count and a hip count differ, and so do a watch and a research accelerometer. Treat the thresholds as regions, not lines: “around 7,000” after 60, “around 9,000” before, with the steepest part of the curve below 6,000.

Nothing directly. The app does not count steps, because a watch already does that better, and because a daily count is not a quarterly benchmark. Where everyday activity enters the app is zone 2 minutes: with Apple Health or Health Connect linked, workouts whose average heart rate sits in your moderate band are counted, and the result is shown as a share of the 150-minute guideline rather than a percentile.

The step evidence is still a good reason to walk. If your VO₂max, stair time or zone 2 minutes are behind their lines, the first and cheapest change is more daily walking, and the curve above says the first few thousand steps are worth the most.