How Fast You Walk When Nobody's Timing You

There is a test that predicts survival in older adults about as well as blood pressure, body-mass index, smoking history and a full list of chronic conditions do together. It takes about four seconds. You have already taken it several thousand times this week without noticing.

It is the speed you walk down a corridor when nobody has asked you to hurry.

This sounds like something a wellness newsletter invented on a slow week. It isn't. In January 2011, a team led by Stephanie Studenski pooled the individual records of 34,485 adults aged 65 and over from nine separate cohort studies, followed them for six to twenty-one years, counted 17,528 deaths, and found that ordinary walking speed tracked survival across the entire range of speeds — not only at the frail end, where anyone would expect it, but all the way up.

Four metres, a stopwatch, and 34,485 people

The protocol is almost insultingly plain. A four-metre stretch of floor. A standing start. One instruction: walk at your usual pace, as if you were walking down the street. No encouragement, no second attempt, no talk of records. Distance divided by time.

Every extra 0.1 metres per second — roughly the difference between a slightly weary walk and an ordinary one — came with a pooled hazard ratio for death of 0.88. About twelve per cent lower risk, per notch, with no ceiling where the benefit flattened out. The average across the pooled group was 0.92 m/s at a mean age of 73.5.

Then the authors ran the comparison that should make you sit up. Predicting five-year survival from age, sex and gait speed alone was about as accurate as predicting it from age, sex, chronic conditions, smoking history, blood pressure, body-mass index and hospitalisation history. One number, measured in a hallway, did roughly the work of an entire intake form.

Why a hallway can speak for the whole body

Walking feels free, which is why the result seems implausible. It isn't free. Moving your own body forward at a comfortable pace draws on the heart and lungs to supply it, the circulation to distribute it, the peripheral nerves to time it, the hips and ankles to transmit it, the eyes and the vestibular system to keep you upright, and a quiet slice of attention to run the arrangement without your noticing.

You cannot be failing at any one of those and still cross a room at 1.3 m/s. That is the whole trick. Studenski's group described gait speed as a summary indicator of vitality — one output fed by a dozen inputs, several of which nobody has measured in you.

Most health metrics are the opposite: exquisitely precise about one system and blind to the rest. A resting heart rate reports on one organ's negotiation with your nervous system. A walking pace reports on whether the assembled machine still works. It is the logic that makes a handshake surprisingly informative and ten seconds on one leg worth attention, except that walking pulls in more systems than either.

There is a second, less flattering mechanism. Slowing is not only a symptom, it is also a cause: walking less deconditions you, deconditioning slows you further, and the loop tightens quietly over years. Nobody notices the year they stopped taking the long way round.

What the numbers mean, and where the thresholds go soft

In the pooled data, a usual pace of about 0.8 m/s sat at median life expectancy for a given age and sex — the middle of the distribution, not a diagnosis. Speeds of 1.0 m/s and above consistently ran better than age and sex alone would predict. Below roughly 0.6 m/s has been proposed as the point at which poor health and function become likely.

Above that, the evidence thins out fast, and the authors say so themselves: 1.2 m/s may mark exceptional life expectancy, but that line is a hypothesis awaiting research, not a finding. Anyone quoting it at you as a target has read the headline and not the discussion.

The spread is what makes this worth knowing. At age 75, predicted ten-year survival across the full range of walking speeds ran from 19% to 87% in men and from 35% to 91% in women. Same age, same sex, wildly different outlooks — separated by a measurement most people have never taken.

The crossing signal already formed an opinion about this

Here is the part that turns an abstract number into something you can feel. Pedestrian crossings in the United States are timed on an assumption about how fast you walk. Since 2009 the federal standard has required the clearance interval to be calculated at 3.5 feet per second — about 1.07 m/s. It used to be 4.0 feet per second, roughly 1.22 m/s, and it was lowered because too many people were still in the road when the light changed.

Set that against the pooled cohort average of 0.92 m/s and you get an uncomfortable piece of arithmetic. The average walking pace of adults over 65 is below the speed the road system assumes of them. Every signalised junction in the country is running an unblinded field test, several million times a day, and a substantial share of the participants are failing it.

The practical upshot: you do not necessarily need a tape measure. If the countdown regularly beats you across a two-lane road, you are somewhere under 1.07 m/s, and you have learned most of what the four-metre test would have told you.

The step you take while thinking about something else

In February 2026 a group at Ben-Gurion University published something that extends the idea in a direction worth watching. They went back to balance data collected between 2005 and 2011 from 120 community-dwelling older adults, average age 77.8, and followed their survival for ten to seventeen years.

They tested plenty of things: postural sway, standing stability, the whole battery. What predicted survival best was none of those. It was the time between a signal appearing and a foot actually leaving the floor — step initiation — measured while the participant was simultaneously doing a cognitive task. Adjusted for age and sex, every extra tenth of a second was associated with about a 28% higher mortality hazard (HR 1.279, 95% CI 1.119–1.462).

The sample is small, the design is retrospective, the authors could not account for how physically active anyone was, and they call the work exploratory themselves. Treat the size of that number with suspicion. But the direction is the interesting part: what carried the signal was not balance while standing still. It was the delay between deciding to move and moving, with your attention pointed elsewhere — which is exactly the situation in which people actually stumble.

Mirror, or lever?

This is where most articles about walking speed quietly change the subject, so let's not. Is gait speed something you can improve to your benefit, or only a readout of a state you are already in?

It has one more piece of evidence behind it than grip strength does. A 2007 study followed 439 adults aged 65 and over for eight years, tracking not their walking speed but the change in it. Those whose usual pace improved by at least 0.1 m/s over a single year had 31.6% mortality over the following eight years. Transient improvers: 41.2%. Those who never improved: 49.3%. Adjusted, the improvers carried a hazard ratio of 0.42.

Tempting — and still not causation. Someone whose pace improves over a year has often recovered from an illness, a bad hip, a rough winter. The improvement may be the visible edge of getting better rather than the cause of it, and an observational design cannot separate the two.

The closest thing to an experiment is the LIFE trial, published in 2014: 1,635 sedentary adults, average age 78.9, randomised either to a structured programme of walking, strength and balance work or to health education. Major mobility disability — losing the ability to cover 400 metres on foot — occurred in 30.1% of the activity group and 35.5% of the education group, a hazard ratio of 0.82. Real, modest, worth having.

Note precisely what that shows. Structured activity protected people's walking. No trial has yet shown that pushing your gait speed up adds years to your life. Walking speed is an excellent instrument and an unproven target — the same caution that applies to most longevity biomarkers, and the reason we keep saying numbers are for correcting fantasy, not replacing experience.

How to take the test, and what to do with it

Measure four metres of flat, uncluttered floor. Start standing still at the line, walk at your usual pace as if heading down the street, and have someone time from your first movement to the moment you cross the far mark. Divide four by the seconds. Two attempts, take the better one. Everyday shoes.

Then leave it alone for three months. The protocol is fussier than it looks — course length, whether you were told to "go", any turn involved — so today's digit is far less informative than the direction of travel across a year. Studenski's own caveat is worth keeping in mind: her participants had no idea what walking speed meant, and people who know what a measurement signifies tend to perform it differently. The number gets less honest the more you care about it.

If you want the number to move, the evidence points at the velocity of your movement rather than the load. Walking speed is a power problem — force applied quickly — and power fades faster with age than strength does, which is why training the speed of a movement, not just its weight is the part most programmes leave out. The LIFE trial's own recipe was unglamorous: walking toward 150 minutes a week, plus strength and balance work. If you would rather start with volume than structure, the dose-response curve for daily walking is a reasonable place to begin.

The bottom line

Your usual walking speed is one of the cheapest and most informative measurements available to you, and almost nobody takes it. In adults over 65 it tracks survival across the whole range, at roughly 12% lower risk per 0.1 m/s, and it does so about as well as a panel of established risk factors — because it quietly audits the heart, lungs, nerves, joints and attention all at once.

What it does not do is take instructions. No trial has shown that speeding yourself up buys you the outcomes fast walkers enjoy, and the cut points above 1.0 m/s are proposals, not findings. Measure it, note it, watch the trend over years, and treat a real change as information worth acting on rather than a score to beat.

And when you next find yourself striding across a junction with the countdown still in double digits, notice it. That is the test, administered for free, by the city, several times a day.