Myopia Prevention and Outdoor Light Intensity in a School-Based Cluster Randomized Trial

693 grade-1 children in 16 schools (267 intervention, 426 control) after 8 of 24 randomized schools withdrew

Grade-1 schoolchildren in multiple areas of Taiwan, 1-year follow-up

Multi-area cluster-randomized trial of the Recess Outside Classroom (ROC) programme targeting up to 11 hours outdoors weekly. Graded C rather than B because of a cluster-randomization failure: 24 schools were randomized but 5 intervention and 3 control schools withdrew before enrollment, leaving badly unbalanced arms (267 vs 426) and breaking the protection randomization was meant to provide. Light exposure was objectively measured with light-meter recorders. No academic outcome measured.

Intervention arm showed less myopic shift (0.35 D vs 0.47 D, P = .002) and less axial elongation (0.28 mm vs 0.33 mm, P = .003) at 1 year, and a 54% lower risk of rapid myopia progression (OR 0.46, 95% CI 0.28-0.77). Unlike most of the literature, protective effects were reported in both already-myopic and non-myopic children. Light-meter data indicate high-intensity sunlight is not required: >=200 min/day at >=1000 lux (hallways, shade under trees) was associated with less myopic shift, which matters practically because it means covered outdoor space counts.

School-level randomization in principle removes genetic confounding, but differential pre-enrollment school withdrawal (5 vs 3 schools) reintroduces the possibility that surviving clusters differ systematically. The lux dose-response analyses are observational.