Fig. 2

Heatmap of the weightings of all questions contributing to each component from principal component analyses in kindergarten, elementary school, and high school, respectively. All questions in the questionnaire on parental behavior in myopia control and parental attitude about children’s time distribution after class were included. Cells in black indicate that these questions were not asked at the kindergarten level. Multivariate logistic regression analyses for children’s myopia or high myopia were performed for all components at each school level, adjusted for age, parental myopia, parental education level, and SES. Components with significant associations in the analyses are marked with asterisks, daggers, and double daggers. * At the kindergarten level, a significant negative association was observed between component 5 and children’s myopia (adj. OR: 0.71, 95% CI: 0.51–0.98, P = 0.04). This component was majorly contributed by the parental attitude in regulating outdoor time and reading time. † At the elementary school level, a significant positive association was observed between component 1 and children’s myopia (adj. OR: 1.19, 95% CI: 1.05–1.34, P = 0.005), and a negative association was observed between component 3 and children’s myopia (adj. OR: 0.69, 95% CI: 0.61–0.77, P < 0.001). Component 3 was negatively affected by parental behavior in the medical control of myopia. ‡ At the high school level, no significant association was observed between the components and myopia in children. However, when we targeted moderate and high myopia (n = 963) subgroups, a negative association at a borderline significance level was observed between component 1 and children’s high myopia (adj. OR: 0.87, 95% CI: 0.75–1.0, P = 0.05)