The largest step-count review in history tracked 161,176 adults across nine health outcomes. The number where benefits actually plateau will change what you do tonight.
Seventy percent of the daily step goal delivered roughly 98% of the mortality benefit. The last 30% added almost nothing.
The number where walking benefits actually start leveling off is 5,391 steps per day.
Not 10,000. Not 8,000. 5,391. That’s the inflection point — the bend in the curve where adding more steps stops making much difference — from the largest step-count review ever published.
Ding and colleagues at the University of Sydney pulled together 57 studies, covering 35 research cohorts and 161,176 adults tracked across nine different health outcomes. Published in The Lancet Public Health in 2025, this review didn’t ask whether walking helps. It asked exactly how much.
If you checked your step counter today and saw a number starting with 6, maybe 7 — and felt that familiar pang of falling short — the answer might rewrite what “short” means.
The largest step-count review ever published found that most of the health benefit from walking arrives thousands of steps before your tracker turns green.
- The 10,000-step target traces to a 1965 Japanese pedometer ad, not a medical recommendation.
- Walking 7,000 steps per day was linked to a 47% lower risk of dying compared to 2,000 steps — 10,000 steps added just one more percentage point.
- A single daily walk to 4,000 steps delivered 36 times more benefit per step than the push from 7,000 to 10,000.
- For adults over 65, the curve never flattened — every additional step kept reducing risk with no ceiling the researchers could find.
- For depression and type 2 diabetes, the relationship was linear all the way to 10,000 steps and beyond, unlike the plateau seen for overall mortality.
A 1965 Pedometer Ad and the Goal That Stuck
Before this review existed, the number 10,000 was already everywhere. Every fitness tracker defaults to it. Every wellness app reinforces it. Every “close your rings” notification assumes it’s the finish line.
The origin of that number is not a clinical study.
In 1965, the Yamasa Clock and Instrument Company in Japan released a pedometer called the Manpo-kei — which translates, literally, to “10,000 steps meter.” [1] It was a consumer product marketed during the fitness wave after the 1964 Tokyo Olympics. The name was catchy. The number was round. And it stuck.
A 2019 paper in JAMA Internal Medicine confirmed what researchers had long suspected: the 10,000-step target had “limited scientific basis” and was “perpetuated by the lay press” and wearable software defaults. [1] The goal that governs the daily routine of hundreds of millions of tracker users traces back to a product name — not a medical recommendation.
That’s the backdrop for what the dose-response data actually showed. Because the real question was never whether 10,000 steps is helpful. It was whether the number that makes you feel guilty at 6,400 — or 7,000, or 8,200 — has any basis in what the research actually found.
One Million Extra Steps for One Percent
Here’s what the Ding 2025 review found when it modeled the relationship between daily steps and the risk of dying from any cause. Fourteen studies. 161,176 adults from more than a dozen countries, roughly split between adults under 65 and over 65.
At 7,000 steps per day, the risk of dying from any cause dropped by 47% compared to people taking around 2,000 steps. That’s roughly half the risk — from walking an amount most people hit without trying particularly hard.
At 10,000 steps? 48%.
Read that again. The jump from 2,000 to 7,000 steps — the part where you’re just... walking through your day — bought 47 percentage points of risk reduction. The jump from 7,000 to 10,000 — the part where you’re pacing your apartment at 11pm, walking circles around the parking lot, bargaining with your tracker — bought one.
Three thousand extra steps per day. Three hundred and sixty-five days. That’s 1,095,000 additional steps per year for a single percentage point of mortality reduction.
A million steps. For one percent.
The 10,000-step target had limited scientific basis — it traces to a 1965 Japanese pedometer marketed as the “ten-thousand steps meter.”
The Walk That Outperforms Three Thousand Steps
If the ceiling of step benefits is lower than everyone assumed, the floor is even more surprising.
At just 4,000 steps per day — for most people, that’s their baseline movement plus a single 20-minute walk — the review found a 36% lower risk of dying from any cause compared to the 2,000-step reference point. That single walk bought a 36-percentage-point drop in risk.
The push from 7,000 to 10,000 bought one.
Put differently: one daily walk is worth 36 times more than the entire grind from 7,000 to 10,000. The lowest-effort threshold the study measured delivered the biggest per-step return.
This wasn’t a small finding in a small sample. The mortality analysis alone drew from 14 studies and 161,176 participants across four continents. The pattern held: every additional step helped, but the first few thousand delivered the vast majority of the benefit, while the last few thousand barely moved the needle.
The Curve That Doesn’t Flatten After Sixty
The pattern described above — big early gains, diminishing returns — held for adults under 65. But for older adults, the researchers found something different.
When the researchers looked at mortality separately by age, the relationship between steps and survival risk was steeper at lower step counts for younger adults, with gains flattening around 5,410 steps per day. Past that number, additional steps still helped — but less and less.
For adults 65 and older, the curve didn’t flatten at all. The relationship stayed straight — every additional step continued to reduce the risk of dying at roughly the same rate, with no ceiling the researchers could find.
A separate large-scale review confirmed the pattern. Paluch and colleagues tracked 47,471 adults from four continents and found the same age split — older adults benefited at lower step thresholds than younger ones. [2]
The practical meaning: if you’re in your twenties or thirties, the biggest health gains from walking arrive early. If your parents are in their sixties or seventies, every additional step they take keeps paying off. There is no “enough” on their curve.
At 4,000 steps per day — one additional 20-minute walk — the risk of dying from any cause dropped 36%. That single walk was worth 36 times more than the push from 7,000 to 10,000.
What 161,176 People Can’t Tell You
The mortality finding — the one most people will remember — drew from 14 studies and earned a Moderate certainty rating from the researchers using a standard grading system. That’s the second-highest grade, and it means the evidence is strong enough that future research is unlikely to reverse the direction of the finding.
But not every outcome in this review earned the same confidence.
The dementia data came from two studies. Cancer incidence: also two. Falls earned the lowest certainty rating the system offers, partly because the available studies pointed in different directions. These are early signals, not settled verdicts.
And not all outcomes followed the same curve. For depression and type 2 diabetes, the relationship was linear — more steps kept helping, with no plateau at any threshold. If your step count dropped from 10,000 to 7,000 and you care specifically about depression risk, this review suggests the reduction may matter for that outcome even if it barely registers for overall mortality.
There’s a bigger caveat the study’s authors were careful to name. All of this data is observational.
The researchers didn’t randomly assign people to step counts and measure what happened. They tracked people’s existing walking behavior and compared outcomes — which means they found an association, not proof of causation. It’s possible that healthier people simply walk more, rather than walking making people healthier.
Most of the primary studies scored well on quality assessments, and removing lower-quality studies from the analysis produced similar results. But no causal framework was applied. The researchers said so explicitly.
Here’s what the review added up to.
If you’re walking 6,000 or 7,000 steps most days, you have already captured the vast majority of the mortality benefit this research could measure. The guilt your tracker generates past that point is chasing a number that was never based on the science.
A 1965 pedometer ad set the target. Sixty years of software kept it alive. And the largest review of the evidence to date found the real inflection point thousands of steps below it.
The part that might change your evening: 70% of the daily step goal delivered roughly 98% of the mortality benefit. The last 30% — the part that had you pacing at 11pm — added almost nothing.
That’s not a reason to stop walking. For some outcomes, the data showed more steps kept helping all the way to 12,000 and beyond. And for older adults, there was no ceiling at all.
But it is a reason to stop feeling guilty about 6,400.
One question this review doesn’t answer: if extra steps above 7,000 don’t meaningfully change health outcomes, do they still help with body composition? When you’re walking for your cut — not for longevity — the math might work differently. That’s a question about energy expenditure, not disease risk. And the answer involves a different body of research entirely.
What other research found
What this means for you
The mortality curve for younger adults bent at 5,410 steps per day. Past that, additional steps still helped — but the returns shrank fast.
Heart disease tells a slightly different story. For cardiovascular risk specifically, the inflection point for younger adults was 7,802 steps — meaning your heart kept benefiting well past the mortality plateau. If you're already hitting 6,000 or 7,000 most days, the biggest gains are behind you for mortality. But your cardiovascular system was still picking up benefit all the way to around 8,000.
The dose-response curve for older adults didn't flatten. At all. Every additional step kept reducing the risk of dying at roughly the same rate, with no ceiling the researchers could find.
For heart disease specifically, the inflection point for older adults was just 5,386 steps — lower than for younger adults. That means cardiovascular benefit kicked in earlier, and overall mortality benefit kept climbing without a ceiling. If your parents walk 4,000 or 5,000 steps a day and feel like that's not enough — the data says it's already delivering substantial benefit, and every step past that keeps paying off.
For most outcomes in this review, the benefit curve flattened somewhere between 5,000 and 9,000 steps. Depression and type 2 diabetes were the exceptions.
Both showed a straight-line relationship — more steps, more benefit, no plateau. At 7,000 steps, depression risk dropped 22%. At 10,000, it dropped 33%. That's 11 additional percentage points of risk reduction from the same 3,000 steps that bought just one percentage point for overall mortality. If either of these outcomes matters to you personally, this review suggests the push past 7,000 genuinely counts.
Before you change anything
This review pulled from 57 studies across 11 countries, but the participant pool skewed heavily toward high-income, predominantly White populations. The USA contributed 37% of the studies, the UK 21%, and Japan 14%.
Children and adolescents were excluded entirely. So were activities that don't register as steps — cycling, swimming, rowing. If your primary exercise doesn't involve feet hitting ground, these thresholds don't directly apply.
Most participants were volunteers, which means they were likely healthier than the general population before the study even started. The specific step thresholds — 5,391, 7,000, 10,000 — come from research-grade devices worn at the hip or waist. Your Apple Watch or Fitbit on your wrist may count steps differently, so the exact numbers might not translate one-to-one to your tracker.
Every study in this review was observational. Nobody was randomly assigned a step count. The researchers tracked how much people already walked and compared their health outcomes — which means it's possible that healthier people simply walk more, rather than walking making people healthier.
Step counts were typically measured over a few days at a single point in time. That snapshot might not capture someone's actual long-term walking habits — a bad week or a vacation could skew the data.
One large study (UK Biobank) had outsized influence. Removing it changed the dose-response shape for falls and cardiovascular mortality. The core mortality finding held up across sensitivity analyses, but some secondary outcomes were more fragile than the headline numbers suggest.
The mortality finding sits on solid ground — 14 studies, 161,176 participants, moderate certainty under a standardized evidence-grading system, and no sign of publication bias. Future research is unlikely to reverse the direction.
But the review covered nine different health outcomes, and confidence varied dramatically across them. Heart disease incidence, cancer mortality, diabetes, dementia, and depression each earned moderate certainty. Heart disease mortality and cancer incidence earned low. Falls earned the lowest rating the system offers.
For dementia and cancer incidence, the entire evidence base came from just two studies each. Those findings are interesting signals, not settled verdicts. The 8,829-step inflection for dementia — the highest threshold in the review — could shift substantially with more data.
This review settled one question: how many steps matter for health. The inflection point, the age curves, the outcome-by-outcome breakdown — drawn from 57 studies and 161,176 adults.
But if you're walking for your cut, you were never walking for longevity. You were walking for the calorie burn. And whether your body actually burns more from more steps — or compensates — is a question this review never tried to answer. That's a different body of research entirely.
What This Study Found
All findings from this paper, in plain language.
- The risk of dying from any cause kept dropping as daily steps increased, but the biggest gains came before 5,400 steps — after that, additional steps helped less and less.
- At 7,000 steps per day, every health outcome the review measured showed a lower risk compared to 2,000 steps — ranging from 6% for cancer to 47% for overall mortality.
- Pushing past 7,000 steps added little or no measurable benefit for heart disease mortality, cancer, diabetes, or falls.
- Even a modest increase to 4,000 steps per day — roughly one extra walk — was linked to a 36% lower risk of dying compared to 2,000 steps.
- For younger adults, the mortality curve flattened around 5,410 steps. For older adults, it never flattened — every step kept reducing risk.
- Heart disease risk followed a different pattern by age: younger adults' curve bent at 7,802 steps, while older adults' bent at just 5,386.
- The risk of dying specifically from heart disease or cancer dropped sharply by 4,800 to 5,400 steps, with smaller gains beyond that.
- Dementia risk kept declining up to about 8,800 steps per day — the highest threshold of any outcome — but this was based on only two studies.
- Steps measured by accelerometers showed a curve that flattened, while pedometer-measured steps showed a straight line — the type of device may affect how the relationship looks.
- The widely used 10,000-step target has no clear scientific basis, and 7,000 steps may be a more realistic benchmark for most adults.
- Most of the included studies scored high on quality assessments, with nearly 80% earning 7 or more out of 9 possible points.
- For depression and type 2 diabetes, more steps kept helping in a straight line with no plateau — unlike the curve that flattened for mortality.
- The review found no sign that positive-result studies were published more often than negative ones for the main mortality analysis.