If you've read anything about step counts in the last few years, you've met the number 4,400. It usually arrives as a headline: just 4,400 steps a day linked to a 41% lower risk of dying. That is close to what the paper says, and it is also easy to misread, so here is the paper, read slowly.
Where 4,400 comes from
The study is Lee and colleagues, published in JAMA Internal Medicine in 2019. It drew on the Women's Health Study, a long-running Harvard cohort. Between 2011 and 2015, about 18,000 surviving participants agreed to wear a hip accelerometer for a week, and 16,741 wore it for long enough to count. They were between 62 and 101, with an average age of 72. The researchers followed them for an average of 4.3 years, and in that time 504 of them died.
The women were split into four equal groups by average daily steps. The 4,400 figure is, roughly, the middle of the second group.
| Group, by daily steps | Median steps | Deaths | Risk vs lowest group |
|---|---|---|---|
| Lowest quarter | 2,718 | 275 | 1.00 (reference) |
| Second | 4,363 | 103 | 0.59 |
| Third | 5,905 | 77 | 0.54 |
| Highest quarter | 8,442 | 49 | 0.42 |
Risk figures are hazard ratios adjusted for age, smoking, alcohol, diet, self-rated health, existing heart disease and cancer, and other factors. Source: Table 3 of the paper.
The "41% lower" is simply 1 minus 0.59, the second group's adjusted risk compared with the lowest group's. So 4,400 is not a threshold the researchers discovered. Nobody found that 4,399 steps does nothing and 4,400 does something. It is the median of a group, and that group happened to be the first one that looked clearly better than the bottom.
The shape of the curve
The authors also fitted a smooth curve through the data, which says more than four boxes can. Risk fell steadily as steps went up until about 7,500 a day, then flattened. Compared with women taking 2,000 or fewer steps, the first statistically clear drop appeared in the 3,000 to 3,999 band. Across the whole range, each extra 1,000 steps went with roughly 15% lower risk. The big gains are at the bottom, which is the actual point of the paper.
The groups weren't alike
This is the part that tends to get cut from news coverage. The women in the lowest group weren't just less active. They were older, heavier and sicker before anyone counted a step.
| Lowest group | Highest group | |
|---|---|---|
| Average age | 75.2 | 69.7 |
| Average BMI | 28.2 | 24.2 |
| History of diabetes | 16.7% | 4.2% |
| History of high blood pressure | 83.8% | 54.3% |
| Cardiovascular disease | 4.2% | 1.2% |
Source: Table 2 of the paper.
That matters because poor health can cause low step counts just as easily as low step counts can cause poor health. A 75-year-old who manages 2,700 steps a day is not a random 75-year-old. The authors adjusted for age, smoking, alcohol, diet, self-rated health and existing disease. They also reran the analysis in ways meant to catch this problem: once dropping the first year of follow-up and any woman with existing heart disease, cancer or diabetes, and once keeping only women who rated their health good or excellent. For the top group, risk came out at 0.50 and 0.44, against 0.42 in the main analysis. That is reassuring. It is not the same as a randomised trial, and the authors list reverse causation as a limitation themselves, along with the fact that steps were measured for a single week.
What about walking faster?
The paper also looked at stepping intensity: the fastest single minute of the day, the fastest half hour, time spent at purposeful cadences. Each of them looked protective on its own. Once total steps were accounted for, most of the associations shrank to the point of not being statistically significant, and the authors concluded that in this group the number of steps mattered more than how fast they were taken.
There is a catch. These women barely walked fast. The median share of the day spent at 100 steps a minute or more, roughly 2.5 mph and up, was 0.2%. A study can't tell you much about a behaviour almost nobody in it did.
Who this does and doesn't describe
By the authors' own account, the participants were mostly white, better off, and more active than a national sample: they averaged 5,499 steps a day, against about 4,000 for American women over 60 in earlier pedometer data. They were also healthier than the Women's Health Study members who chose not to take part. And it is only women, only 62 and over. Nothing in it says what a 35-year-old man should aim for.
A later analysis in The Lancet Public Health pooled 15 cohorts and 47,471 adults and found the plateau sits in different places by age: around 6,000 to 8,000 steps a day for people 60 and older, and 8,000 to 10,000 for younger adults. So 7,500 for older women fits that picture. It isn't a number for everyone.
A footnote on 10,000
The paper's own introduction says the origin of the 10,000-step goal is unclear and probably traces to a pedometer sold in Japan in 1965 called Manpo-kei, which translates as "10,000 steps meter." The study was pitched partly as encouragement for people who find 10,000 out of reach.
How we'd use it
We'd treat 4,400 as the answer to a narrow question: is a modest increase from a very low starting point worth bothering with? The data say yes, and the effect looks biggest exactly there. We wouldn't treat it as a target, as a ceiling, or as a reason to stop at 7,500 if you're 40 and happy to walk more. If you're averaging 2,500 or 3,000 steps, getting to 4,500 probably matters more than anything you could do between 8,000 and 12,000.
The practical first step is to find out your actual average, without trying to change it, for a week. Any counter will do, including the one built into your phone. To see where you fall on the usual categories, our page on active vs sedentary step counts lays them out, and how many steps you should walk daily covers targets by age. For the older-adult angle, see walking for seniors. If you'd like a counter with streaks, StepMax, our own free app, has one: Android, iPhone.
Sources
- Lee I-M, Shiroma EJ, Kamada M, Bassett DR, Matthews CE, Buring JE. Association of step volume and intensity with all-cause mortality in older women. JAMA Internal Medicine. 2019;179(8):1105-1112. jamanetwork.com (doi:10.1001/jamainternmed.2019.0899)
- Paluch AE, Bajpai S, Bassett DR, et al. Daily steps and all-cause mortality: a meta-analysis of 15 international cohorts. The Lancet Public Health. 2022;7(3):e219-e228. thelancet.com
Numbers in this article were checked against the full text of the 2019 paper (abstract, Tables 2 and 3, and limitations section) on October 2, 2026. This is general information, not medical advice.