Time outdoors prevents myopia in hyperopic children, but protection is weaker in premyopic children - a post-hoc analysis of a cluster-randomised trial

3,194 non-myopic participants (1,369 premyopic), mean age 8.2

Children aged 6-9 without myopia nested in the Shanghai STORM cluster-randomised trial, wearable-monitored 2017-2018

Post-hoc dose-response analysis nested in an RCT, but the exposure analysed is objectively measured daily outdoor minutes rather than randomized arm - so this is an observational analysis inside a trial and graded C accordingly. Cycloplegic refraction; premyopia defined as SE -0.50 to +0.75 D.

Hyperopic children showed reduced myopic shift with increasing outdoor time, plateauing around 120 min/day. Premyopic children showed a J-shaped relationship with no statistically significant benefit at any dose: versus <60 min/day, differences in SE change were -0.03 D (95% CI -0.10 to 0.05) at 61-90 min, -0.03 D (95% CI -0.11 to 0.05) at 91-120 min, and 0.04 D (95% CI -0.05 to 0.14) above 120 min. The protective effect of outdoor time is concentrated in children who have refractive reserve left; by the time a child is premyopic, the window is largely closed.

Medium-to-high: the contrast is between children who happened to spend more or less time outdoors, not between randomized arms, so heritable propensity to outdoor activity and heritable refractive trajectory both load on the estimate.