In July 2025, The Lancet Public Health published a large review on walking and health. Ding and colleagues gathered 57 studies from 35 groups of people who were followed over time [1]. They pooled 31 of those studies to ask one question: how does the number of steps people take each day relate to their later health?
I think the study is good news for anyone who finds 10,000 steps hard to reach. However, the popular reading that 7,000 steps is the new 10,000 leaves out three things. The group the walkers were compared with walked very little, and for some outcomes the extra benefit gets smaller as steps go up. Also, people who are already sick tend to walk less, which can make walking look more helpful than it really is.
Where did 10,000 steps come from?
The number did not come from a study. A 2022 analysis led by Paluch said the goal “is not based on evidence, but instead originates from a marketing campaign in Japan” [2]. Ding and colleagues call it “an unofficial target for decades without a clear evidence base” [1].
That does not make 10,000 steps a bad goal. It means the number came before the evidence. The new review was done to help update Australia’s physical activity guidelines.
What did the 2025 review find?
This was a meta-analysis, a study that pools results from many earlier studies. The earlier studies were observational. That means researchers measured how much people walked and then waited to see who got sick. Nobody was told to walk more. Steps were counted with devices, such as step counters worn on the hip or wrist, not guessed by the walkers.
The headline numbers compared 7,000 steps with 2,000 steps a day [1]. At 7,000 steps, the risk of dying from any cause during the studies was 47% lower, a bit less than half. New cases of cardiovascular disease (heart and blood vessel disease) were 25% lower, about 1 in 4. Dementia was 38% lower, and depressive symptoms were 22% lower. For new cancer cases, the difference was only 6%, small enough that it could be due to chance.
Lower than whom?
A quick reading easily misses who the walkers were compared with. Every one of those percentages is a comparison with people who took about 2,000 steps a day. The authors chose 2,000 steps as “the lower bound of the normal range for older adults,” which in plain terms means a very inactive day.
So a 47% lower risk does not mean 7,000 steps halves anyone’s risk. It means people at 7,000 steps did much better than people who barely moved. If you already walk 6,000 steps, the study does not promise you anything close to that.
Imagine two neighbors. One walks about 2,000 steps a day, mostly around the house. The other walks 6,000 steps and has read that 7,000 cuts the risk of early death in half. The large drop in risk found in the study applies most closely to the first neighbor, not the second.
The review has the most to say about the first neighbor. The numbers changed most when people added steps from a very low start.
Why the first steps matter most
For four outcomes, including death from any cause, the review found diminishing returns. That means each extra 1,000 steps helped less than the ones before. Going from 2,000 to 4,000 steps was linked with a 36% lower risk of death [1]. For those four, the benefit began to level off at around 5,000 to 7,000 steps. For the other outcomes, risk kept falling at a steady rate as steps went up.
More steps were still linked with lower risk for some outcomes. In an extra analysis, 10,000 steps was linked with a further 10% lower risk of death than 7,000 steps. But for deaths from cardiovascular disease, new cancers, type 2 diabetes and falls, going past 7,000 did not show a clear extra benefit. Age also made a difference. For older adults, the drop in death risk did not level off.
The earlier Paluch analysis found a similar pattern [2]. It combined 15 studies with 47,471 adults. The risk of death leveled off at about 6,000 to 8,000 steps for adults 60 and older. For younger adults, it leveled off at about 8,000 to 10,000 steps. These two results come from different analyses. Ding’s team sorted whole studies by their participants’ average age. A study counted as older if that average was 65 or more, and only six such studies were pooled for death. Paluch’s team had data on each person and split people at age 60.
Does walking cause the benefit, and how sure are we?
These studies cannot fully answer that. One problem is called reverse causation. It means the cause may run the other way. Illness can make people walk less. Someone with early heart failure or a hidden cancer may take fewer steps before anyone knows they are ill. They are also more likely to die soon, whatever their step count.
Ding and colleagues say this plainly. People at about 2,000 steps “might still not be comparable with participants with higher step counts in many aspects of health and physical function” [1]. Most studies also counted steps only once, over a few days.
Paluch and colleagues tested this worry [2]. They removed people who died in the first 2 years. The link between steps and death got weaker, but it did not go away. The link was also stronger in studies that followed people for a shorter time. In my view, some of the headline effect probably comes from sick people walking less. But part of the link still looks real.
How sure the authors could be varied a lot from one outcome to another. The review used GRADE, a common system for rating how strong evidence is [1]. Under it, observational studies start with a low rating. The authors raised the rating to moderate for six outcomes, including death, because risk fell steadily as steps rose. It stayed low for deaths from cardiovascular disease, new cancers and physical function. For falls, it was very low, because the studies pointed in different directions.
Every outcome except death and new cardiovascular disease was based on only two to four studies. For those outcomes, the authors say the findings “should be interpreted as exploratory.” Most data came from wealthy countries.
The journal also published a correction in 2025 [3]. Saved copies of the article from before and after it show the same main text and the same headline numbers. The only visible change is that the online supplement (the extra files posted with the article) was replaced. What changed inside those files is not known.
So how many steps should you aim for?
The authors suggest 7,000 steps “might be a more realistic and achievable target for some” [1]. They add that 10,000 can still be a goal for people who are more active. I agree with both points. I would add that the message is most useful for people who walk the least.
I read this review as a strong case for moving more, not as proof that 7,000 steps is a set goal everyone must reach. In the data, the biggest gains went to people who started with very few steps. In my view, the exact number matters less than moving more than you do now.
If you walk very little today, a few thousand more steps is a realistic and worthwhile goal. If you already walk 7,000 steps, the extra gain from more is smaller, though walking more is still fine. And if you read that a step count cuts your risk in half, ask one question: compared with whom?
References
[1] D. Ding, B. Nguyen, T. Nau, M. Luo, B. del Pozo Cruz, P. C. Dempsey, et al., “Daily steps and health outcomes in adults: a systematic review and dose-response meta-analysis,” The Lancet Public Health, vol. 10, no. 8, pp. e668-e681, Aug. 2025, doi: 10.1016/S2468-2667(25)00164-1.
[2] A. E. Paluch, S. Bajpai, D. R. Bassett, M. R. Carnethon, U. Ekelund, K. R. Evenson, et al., “Daily steps and all-cause mortality: a meta-analysis of 15 international cohorts,” The Lancet Public Health, vol. 7, no. 3, pp. e219-e228, Mar. 2022, doi: 10.1016/S2468-2667(21)00302-9.
[3] “Correction to Lancet Public Health 2025; 10: e668-81,” The Lancet Public Health, vol. 10, no. 9, p. e731, Sep. 2025, doi: 10.1016/S2468-2667(25)00199-9.
Related Articles
