All models are wrong, the late British statistician George E.P. Box famously observed, but some are useful.
A version of this aphorism often crops up when researchers debate the merits of promoting 10,000 steps a day as a public-health goal. As either a minimum requirement or a guarantee of good health, it’s clearly wrong. But its usefulness is trickier to assess, as a series of recent studies illustrates.
The origins of the 10,000-step goal are far from scientific. In the 1960s, Japan’s Yamasa Clock and Instrument Company launched a simple pedometer they dubbed manpo-kei, which means “10,000 steps meter.” The number was chosen, according to one theory, because the Japanese character for 10,000 looks like a person walking.
Still, the goal proved to be popular, perhaps because it corresponds to a vigorous but not unattainable day for many people. With the rise over the past decade of high-tech activity trackers and phones that count steps, the default goal of 10,000 steps has become more deeply entrenched as the mark of a sufficiently active day—which, in turn, has spurred researchers to test its efficacy.
A study published earlier in The Lancet, by an international consortium of researchers led by Amanda Paluch of the University of Massachusetts Amherst, offers the most definitive verdict yet. Ms. Paluch and her colleagues pooled the results of 15 previous studies with nearly 50,000 participants on four continents. After a baseline measurement of daily step count, the researchers monitored national death records for an average of seven years to see who died and when.
As expected, the results confirmed that people who took more steps at the start of the study were less likely to die during the follow-up period. The most active quarter of the subjects, who took an average of 10,901 steps per day, were less than half as likely to die during the study as the least active quarter, who took 3,553 steps per day.
How fitness can be a form of self-care and self-compassion
But there were some surprising nuances. Even the third-most-active quarter, averaging 5,801 steps, were about 40 per cent less likely to die prematurely than the lowest group. The biggest gains, in other words, seem to result from getting over about 5,000 steps.
The researchers also split the data into two age groups, under and over 60 at the start of the study. The shape of the mortality curve was similar for both groups, but it levelled off at different points. For older adults, getting 6,000 to 8,000 steps a day seemed to max out the benefits; the younger group didn’t plateau until 8,000 to 10,000 steps.
This result shouldn’t be surprising, Ms. Paluch and her colleagues point out. For older adults, it probably takes fewer steps to impose a similar level of physiological strain. Your body doesn’t care how many steps you take, only how much you elevate your heart rate and how long you keep it there.
On the surface, this is a blow to the universality of the 10,000-step rule. Activity guidelines, like other health advice, need to be tailored to the individual. There are substantial benefits to getting even halfway to the goal. And even taking age into account doesn’t necessarily solve all the problems. There are plenty of 70-year-olds whose robust fitness means that they can and should aim to be more active than people much younger than them.
In a fitness rut? Here’s how to get out of it
As research evolves and fitness tracking technology improves, our devices will undoubtedly keep getting better at tailoring their advice to our unique circumstances. But as a public health initiative, I’m not sure that ‘Do whatever your watch tells you to do’ is an improvement over the status quo.
That’s where it’s worth recalling Mr. Box’s aphorism. A simple, round-number target like 10,000 steps gives us an anchor to compare our current activity levels against, making allowances for our current fitness, health, and circumstances. Sure, it’s wrong—but as a rule of thumb, it remains useful.
Alex Hutchinson is the author of Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance.