Outdoor time and children's eyesight

LGT

0.84 pooled risk ratio for new nearsightedness, 7 randomised trials, 9,437 children

Outdoor time has unusually direct evidence behind it: schools changed the day, then researchers counted how many children became nearsighted.

Settled science

Stasis Home Research Every claim on this page is graded against peer-reviewed health research and linked to its primary source.Updated August 2026

Outdoor time reduces the onset of nearsightedness in children, and that finding comes from randomised trials. Across seven school-based trials involving 9,437 children, the pooled risk ratio was 0.84. All seven ran in China and Taiwan, where baseline nearsightedness was much higher than in a typical US sample, and the finding concerns preventing onset, not slowing progression after nearsightedness begins.

Outdoor time and children's eyesight: what is it, and where does it come from?

Where most of the light domain rests on how people say they feel, this rests on counting. Seven cluster randomised trials assigned schools or groups of children to more outdoor time, then followed 9,437 children to see who became nearsighted. Pooled, the risk ratio was 0.84, with a 95 percent confidence interval from 0.78 to 0.91.

One trial makes the intervention concrete: it added a single 40-minute outdoor class period to each school day, and after three years cumulative incidence was 30.4 percent in the intervention schools against 39.5 percent in the controls, an absolute difference of 9.1 points.

Every one of the seven trials ran in China and Taiwan, in populations where baseline nearsightedness was much higher than in a typical US sample. And what was measured was whether a child became nearsighted, not what happened to a child who already was. This is evidence about how often vision changes in the first place, not about changing it back.

Outdoor time and children's eyesight: why does it matter?

Nothing else in the light domain can say what the children's eyesight trials say. This is the only randomised evidence in the domain showing that outdoor light changes a clinical result at all.

What the trials do not tell you is why. They changed one thing, time outside, and they did not isolate which part of being outside did the work.

Brightness is the obvious suspect, and the size of that gap is not in dispute. Eye-level measurements put open outdoor locations around 9,300 lux, shade under a large tree around 1,580, rooms with windows at a few hundred, and a room with one fixture around 14, with outdoors running about eightfold brighter than indoors overall. Even an overcast morning outside beats a well-lit room, and it is not close.

But nothing in these trials tested light on its own, and a lux figure is not a treatment target. What survives is narrower than a headline and still worth acting on: children given more time outdoors became nearsighted less often.

Seven cluster randomised trials covering 9,437 children found that outdoor-time interventions reduced the onset of nearsightedness, with a pooled risk ratio of 0.84 and a 95 percent confidence interval from 0.78 to 0.91. All seven trials ran in China and Taiwan, where baseline nearsightedness was much higher than in a typical US sample, and the finding concerns preventing onset, not slowing progression once nearsightedness begins.Mei et al., 2024 (Frontiers in Public Health; outdoor interventions for myopia in children, meta-analysis of 7 cluster randomised trials and 9,437 children)
One cluster randomised trial added a 40-minute outdoor class period each school day and found three-year cumulative incidence of 30.4 percent, compared with 39.5 percent in control schools, a 9.1-point absolute difference. The trial ran in China in a population with much higher baseline nearsightedness than a typical US sample, and tested prevention of onset rather than progression of existing nearsightedness.He et al., 2015 (JAMA; effect of an added outdoor class period at school on incident myopia, cluster randomised trial)
The children's outdoor-time trials are the only randomised evidence in the light domain showing that outdoor light changes a hard clinical outcome. This describes the evidence base, not a promised result for any child.Mei et al., 2024 (Frontiers in Public Health; outdoor interventions for myopia in children, meta-analysis of 7 cluster randomised trials and 9,437 children)
Eye-level measurements found open outdoor locations around 9,300 lux, shade under a large tree around 1,580 lux, rooms with windows at a few hundred lux, and a room with one fixture around 14 lux. Pooled, outdoors was about eightfold brighter than indoors, and even an overcast morning outside was substantially brighter than a well-lit room.Bhandary et al., 2021 (PLOS ONE; measured eye-level light levels across indoor and outdoor locations)

Outdoor time and children's eyesight: What did the trials find?

What the randomised trials of outdoor time and childhood nearsightedness report.
What was measured	What the trials found	Source
New nearsightedness, pooled risk ratio	0.84 95 percent confidence interval 0.78 to 0.91; all seven trials ran in China and Taiwan, where baseline nearsightedness was much higher than in a typical US sample, and the result concerns preventing onset, not slowing progression once nearsightedness begins	Mei et al., 2024
Three-year cumulative incidence	30.4% vs 39.5% 9.1-point absolute difference; same limits as the row above, in a much higher-baseline population and for onset only	He et al., 2015
Added outdoor class period per school day	40 minutes what that trial tested, not a dose anyone has established; onset only, in a much higher-baseline population	He et al., 2015

These are trial results, not limits or targets. There is no legal or health threshold for time outdoors, and the 40 minutes is what one trial tested rather than a dose anyone has established. Who those trials studied, and what they do not cover, sits in each row.

What helps against Outdoor time and children's eyesight?

3 approaches below reduce Outdoor time and children's eyesight in a home, ordered from the most direct fix to the smallest change in habit. The most direct is Outdoor time during the school day. Each entry also names what it cannot do, because no single fix covers every exposure route.

Free and behavioral

Outdoor time during the school day
Build a regular outdoor period into the part of the day a child already spends at school. A 40-minute outdoor class period is the clearest tested example, not a guaranteed dose for every child.
The randomised evidence comes from China and Taiwan, where baseline nearsightedness was much higher than in a typical US sample, and it concerns preventing onset rather than slowing existing nearsightedness.
Weekend outdoor time
Ordinary weekend plans keep outside time in the week when school is not supplying it. What the trials changed was time outdoors, not the activity, so nothing here needs a program or a product.
The trials do not establish a weekend dose, and they did not isolate which part of being outside did the work.
Treat outdoor time as prevention, not reversal
Use outdoor time as a practical step before nearsightedness begins, alongside ordinary eye care.
The randomised evidence does not show that going outside reverses nearsightedness or slows its progression after onset.

It sells nothing: no lamp, monitor, lens, or supplement.

We also ship the limits beside the number every time. All seven trials ran in China and Taiwan, where baseline nearsightedness was much higher than in a typical US sample, and the finding is about preventing onset, not slowing progression after nearsightedness begins. The number is useful only when those boundaries stay attached.

Common questions about Outdoor time and children's eyesight

What is a cluster randomised trial, and why does it matter here?

A cluster randomised trial assigns whole groups, in this case schools or classes, rather than individual children. That matters because it makes this the rare light finding that is experimental instead of observational. Almost everything else in this domain compares people who already differ. Here the researchers changed one thing on purpose and then counted what happened, which is what lets the result speak to cause rather than to pattern.

What does a pooled risk ratio of 0.84 actually mean?

Pooling combines seven separate trials, 9,437 children in total, into one estimate. A risk ratio of 0.84 means children given more outdoor time became nearsighted at about 84 percent the rate of the comparison groups. The confidence interval, 0.78 to 0.91, is the range the true effect plausibly sits in; because it stays entirely below 1, the direction is consistent rather than a coin flip. All seven trials ran in China and Taiwan, where baseline nearsightedness was much higher, and they measured preventing onset rather than slowing progression.

Why quote both a risk ratio of 0.84 and a 9.1 point difference?

Both numbers are quoted because they answer different questions. The pooled risk ratio of 0.84, interval 0.78 to 0.91, is relative: the rate fell to about 84 percent of the comparison rate. The 9.1-point difference is absolute: 30.4 percent of children against 39.5 percent. Neither predicts one child. All seven trials ran in China and Taiwan with much higher baseline nearsightedness, and measured onset, not progression.

How long were the children followed?

In the trial that added a 40-minute outdoor class period, children were followed for three years, and cumulative incidence of nearsightedness was 30.4 percent against 39.5 percent in the control schools. Three years is long enough to count real new cases rather than a short-term signal. All seven trials ran in China and Taiwan, where baseline nearsightedness was much higher, and measured onset rather than progression.

Is the 40 minute trial one of the seven in the meta-analysis?

Yes. The 40-minute trial is one of the seven pooled in the meta-analysis, so the two figures are not independent results and nothing is being counted twice. The pooled ratio of 0.84 summarises all seven; the 40-minute period is what one of them actually did. All seven ran in China and Taiwan, where baseline nearsightedness was much higher, and measured onset rather than progression.

Why does it matter that every trial ran in China and Taiwan?

Because the starting point was different. Those populations carry much higher baseline nearsightedness than a typical US sample, and an intervention generally has more room to move an outcome that is more common to begin with. Nothing about the finding is invalidated by that, but the size of the effect should not be assumed to transfer unchanged to a population where fewer children were going to become nearsighted anyway.

Why is preventing onset different from slowing progression?

They are two different endpoints, and these trials only measured the first. Onset is whether a child who could see clearly becomes nearsighted at all. Progression is whether a child who is already nearsighted gets worse. The trials counted new cases, so the evidence supports prevention and is silent on progression, in either direction. A child already in glasses is outside what this research answers.

Does any of this evidence apply to adults?

None of this evidence applies to adults. The seven trials enrolled school-age children in China and Taiwan, where baseline nearsightedness was much higher, and they counted whether a child became nearsighted rather than how an existing case progressed. It is still the only randomised evidence that outdoor light changes a hard clinical outcome, which describes the evidence base rather than promising anyone a result. Adult eyesight was simply not studied.

Do the trials prove that light is the active ingredient?

No. What the trials changed was time outdoors, not brightness in isolation, so light, distance viewing, movement and time away from close work all moved together. Brightness is the obvious suspect, and the size of that gap is not in dispute: measured at eye level, open outdoors ran around 9,300 lux against about 14 in a room under a single fixture. But no trial here tested light on its own.

How does this compare with the rest of the light evidence?

It is the strongest thing in the domain by a wide margin. This is the only randomised evidence anywhere that outdoor light changes a hard clinical outcome. Most of what else gets claimed about daylight rests on people reporting their own habits, or on trials that came back null. The two limits stay attached to the number every time it appears.