How Much Walking Is Enough? What the Step Studies Show

The most useful thing in the walking literature is not the target. It is the shape of the curve.

What the steps actually buy#

A 2022 analysis pooled fifteen cohorts, 47,471 adults and 3,013 deaths over a median 7.1 years. Sorted into quarters by daily steps, the medians were about 3,550, 5,800, 7,840 and 10,900.

Read the first two rows again. Most of the available benefit arrives between the first and second quarter, which is the difference between a mostly seated day and a day with a couple of short walks in it. Everything after that is real but smaller.

A 2025 pooled analysis of 57 studies put numbers on the same shape against a wider set of outcomes. Comparing 7,000 steps a day with 2,000: 47 percent lower all-cause mortality, 38 percent lower dementia, 37 percent lower cancer mortality, 28 percent fewer falls, 22 percent fewer depressive symptoms. Most of the inflection points sat between 5,000 and 7,000 steps, and the authors landed on 7,000 as the more realistic target.

Where 10,000 came from#

It was a product name. A Japanese company marketed a pedometer in 1965 called the Manpo-kei, literally the ten thousand steps meter, and the figure escaped into general advice and stayed there for sixty years.

The peer-reviewed comment is unusually direct. The 2022 pooled analysis opens by saying that although 10,000 steps is widely promoted, there is little evidence to support the recommendation.

Where the curve flattens depends on age. In adults over 60 it flattens around 6,000 to 8,000 steps. Under 60 it flattens around 8,000 to 10,000. There is one honest disagreement in the field worth naming: a 2023 meta-analysis of 226,889 people, with an older average age, found no plateau at all and reported benefit continuing up toward 20,000 steps a day. So the steepest gains are not in dispute, and whether there is a ceiling is.

The dose nobody mentions: half is most of it#

This is the finding most worth knowing and least often repeated.

Pooling 94 cohorts, 163 million person-years and 811,616 deaths, researchers mapped activity against mortality. The full guideline dose, about 150 minutes a week of moderate activity, carried 31 percent lower all-cause mortality. Half of it, about 75 minutes a week, carried 23 percent.

Three quarters of the benefit, for half the time. The guideline is a target, not a threshold, and somebody managing twenty minutes of brisk walking four times a week is not failing at exercise.

Does pace matter?#

Somewhat, and with a caveat.

In the pooled step analysis, a person's peak 30 minute cadence predicted mortality even after adjusting for their total daily steps. Simply accumulating time above a cadence threshold did not. In a separate study of 78,430 UK adults, peak 30 minute cadence was the strongest independent predictor of dementia in the dataset, with the best value around 112 steps a minute.

The caveat matters on a medical page. How fast you can comfortably walk is partly a readout of how healthy you already are, and slowing gait can precede a dementia diagnosis by years. None of this is evidence that forcing a faster pace changes your future. The reasonable version is to include a sustained brisk stretch in your walking rather than to hurry all of it.

What walking does and does not cover#

It does raise fitness in a sedentary person. Across 32 randomised trials of previously sedentary adults, walking programs raised VO2 max by about 3 ml/kg/min, took about 3.6 mmHg off systolic blood pressure, and produced about 1.4 kg of weight loss over an average of 19 weeks. It did not move cholesterol.

It does not build strength, and it is not a falls intervention. That second point is worth stating plainly because the opposite advice is so common. The Cochrane review of exercise for preventing falls found walking programs showed no benefit, and a trial of a brisk walking program in women who had already fractured an arm found more falls in the walking group, not fewer.

The reviewers' own wording is that walking should not be prescribed as a single falls prevention intervention, and that high-risk older people should not be told to walk briskly.

That is not an argument against walking. It is an argument that falls need balance training, which is a different thing.

Walking with a condition#

Three situations change the advice.

High blood pressure. Walking is one of the better things you can do for it, and the 3.6 mmHg average drop across the trials is roughly what a low dose of some medicines achieves. It does not replace treatment for anyone already on it, and the reading still needs checking.

Type 2 diabetes. A walk after eating blunts the rise in blood sugar more than the same walk taken at another time of day. Ten to fifteen minutes after the largest meal is the version most people keep up.

Knee or hip pain. Walking usually helps arthritic joints rather than harming them, but pain that worsens over the following day rather than settling within an hour is a sign the dose is too high, not that walking is wrong. Shorter, more frequent walks solve most of it.

Making it happen at all#

The behaviour research on this is thin and the practical experience is not. Three things reliably work.

Attach the walk to something that already happens. After a meal, before the school run, at the end of the working day. A walk with no anchor is a walk that competes with everything else in the day and usually loses.

Count in walks, not steps. A step target is a number you can fail at by 9pm. Two walks is a thing you either did or did not do.

Do not start with the target. Somebody averaging 2,500 steps who aims for 7,000 tomorrow is planning a week of failure. Adding a thousand a day for a fortnight, then another thousand, gets there without the collapse, and it happens to add the steepest part of the benefit curve first.

A reasonable target#

Somewhere around 7,000 steps a day, with a brisk stretch inside it, and the understanding that going from 3,000 to 5,000 is worth more than going from 8,000 to 10,000.

If you are starting from very little, the useful instruction is not a number at all. It is to add one walk to the day you already have.

One more thing about the curve. Because the steepest part of it is at the bottom, the person with the most to gain from walking is the person currently doing almost none of it, and the person with the least to gain is the one already at 9,000 steps wondering whether to push for 12,000. Most of the advice on the internet is aimed at the second person. It is the first one the evidence is about.

See how your walking speed compares with your age group, and the five other tests that read the same reserve.