Stage-specific associations of outdoor time, reading, and screen exposure with myopia onset and progression in 404,297 Chinese schoolchildren: a multistate cohort study

To determine how outdoor time, reading duration, and screen exposure, individually and in combination, affect stage-specific transitions in myopia development among schoolchildren. The longitudinal study included 404,297 primary school students (grades 1–6) participating in the Tianjin Child and Adolescent Research of Eye (TCARE) program from 2021 to 2024. Non-cycloplegic spherical equivalent (SE) was used to classify refractive states (non-myopia, mild, moderate, and high myopia). Cox proportional hazards models were used to quantify the associations with first incident mild, moderate, and high myopia. Multistate Markov models (MSMs) further estimated transition-specific associations for non-myopia, mild to moderate, and moderate to high myopia. The models were adjusted for age, sex, baseline SE, parental myopia, socioeconomic status, and NDVI. Over three years, 148,952 students progressed from non-myopia to mild myopia, 61,480 from mild to moderate myopia, and 7,847 from moderate to high myopia. Outdoor time exceeding 2 h per day was consistently associated with a lower risk of transition from non-myopia to mild myopia (HR, 0.90; 95% CI, 0.86–0.91), corresponding to a population attributable fraction (PAF) of 8.6% (95% CI, 7.2%–10.0%). In contrast, ≥ 2 h per day of reading (HR, 1.05; 95% CI, 1.03–1.06) and screen exposure (HR, 1.03; 95% CI, 1.01–1.04) at baseline were associated with the higher risk of this early transition. Although these associations were statistically significant, their magnitudes were modest and should be interpreted cautiously in terms of individual-level clinical relevance. None of these behaviors showed statistically significant associations with transitions between later myopia stages in multistate analyses. However, complementary analyses of continuous refractive change suggested that baseline outdoor time and screen exposure remained modestly associated with subsequent myopic progression among children with established myopia. All findings were robust across stratified and sensitivity analyses. Behavioral exposures showed stronger associations with the transition from non-myopia to mild myopia than with later transitions. This pattern should be interpreted cautiously because non-cycloplegic refraction may introduce misclassification around the onset threshold and may not necessarily reflect biological differences between myopia onset and progression. These findings underscore the potential relevance of behavioral factors during early refractive development, while suggesting that the strength of behavioral associations may vary across refractive transition stages.