Abstract
Purpose
To investigate whether premyopia is a risk factor for myopia onset and whether outdoor activities can protect against myopia development in premyopic children in the Recess Outside Classroom (ROC) study.
Methods
Nonmyopic schoolchildren aged 7–11 years were recruited from two schools in Taiwan. One school implemented the ROC program, which encouraged children to go outdoors during recess. The control school maintained its usual schedule. A cycloplegic autorefraction was performed. Premyopia was defined as spherical equivalent refraction ≤ +0.75 diopters (D) and > –0.50 D.
Results
After one year of follow-up, multivariate logistic analysis revealed that the ROC program reduced the risk of myopia onset by 61 % (odds ratio [OR] = 0.39, 95 % confidence interval [CI]: 0.21–0.70, P = 0.002). However, premyopia status increased the risk of myopia onset by 14 times compared to hyperopic status (OR = 14.0, 95 % CI: 1.86–105.3, P = 0.010). In the subgroup analysis of premyopic children, the myopic shift was also significantly lower in the ROC group than in the control group (–0.20 ± 0.60 D/year vs. –0.40 ± 0.66 D/year, P = 0.017). Myopia incidence in premyopic children was significantly lower in the ROC group than in the control group (19.6 % vs. 37.8 %, P = 0.001). Multivariate regression analysis showed that participation in the ROC program was significantly associated with a lower myopic shift in premyopic children (–0.22 D/year, 95 % CI: –0.39 to –0.06, P = 0.008)
Conclusions
Premyopia is a risk factor for myopia onset. A school policy that includes more outdoor time can effectively prevent myopia onset and shift in premyopic children.
1
Introduction
Myopia, or nearsightedness, is a refractive error characterized by difficulty seeing distant objects clearly and a pathologic change in eyeball elongation. School myopia or childhood-onset myopia progresses rapidly until the end of adolescence. Once myopia reaches ≥ –5 diopters (D), it is associated with an increased risk of sight-threatening conditions such as retinal detachment, myopic maculopathy and glaucoma. The prevalence of myopia is increasing, particularly in East Asian countries but also globally, especially during the COVID-19 pandemic. It raises concerns about potential long-term ocular health implications.
Recent research has identified a stage called premyopia. It is defined as a refractive state of an eye of ≤ +0.75 D and > –0.50 D in children. This state, combined with other risk factors, provides a sufficient likelihood of future myopia development to warrant preventive measures. The importance of this issue regarding premyopia is also recently emphasized by the Yilan Myopia Prevention and Vision Improvement Program (YMVIP) cohort study in preschoolers, which also conducted further subgroup analyses specifically targeting premyopia. However, the YMVIP lacked a control group for comparison. Understanding the progression from premyopia to myopia and identifying effective interventions at this stage is crucial for myopia prevention strategies.
Several studies have suggested that increased outdoor time may protect against myopia onset and progression. The Recess Outside Classroom (ROC) study, implemented in Taiwan, was one of the first interventions to use an educational policy to increase outdoor time to reduce myopia incidence by up to 50 %. However, there is still no clear evidence that increasing outdoor time could prevent myopia onset in premyopic children. This study aimed to address two primary questions through a subgroup analysis of the ROC study: 1) Is premyopia a risk factor for myopia onset? and 2) Can outdoor activities protect against myopia development in premyopic children?
2
Methods
This study is a subgroup analysis of the prospective, comparative, consecutive, interventional ROC study published previously. We recruited elementary school students aged 7–11 years from two nearby schools located in a suburban area of southern Taiwan. The children from one school participated in the ROC program (intervention group), while those from the other school served as the control group. The interventions involved implementing a ROC program, where classroom lights were turned off during recess, classrooms were emptied, and teachers required all children to go outside for outdoor activities. The ROC program increased outdoor time by approximately 80 minutes per day (10, 20 and 10 minutes during recess periods in both the morning and afternoon), resulting in a total weekly recess time of approximately 6.7 hours. In contrast, the control school did not implement any special program during recess, and most children remained in the classroom to practice academic work. Both schools allocated 2 hours per week for outdoor physical education.
Cycloplegic refraction measurements were taken at baseline and one year later using an autorefractometer (KR-7000/8100, Topcon; Tokyo, Japan). Children with amblyopia or those using orthokeratology were excluded from this study. A questionnaire for parents was completed (assessing factors such as parental myopia, near-work activities and outdoor activities).
Only nonmyopic children with a spherical equivalent refraction (SER) > –0.5 D were included in this analysis. These children were divided into two groups: premyopic (SER ≤ +0.75 D and > –0.50 D) and hyperopic (SER >0.75 D). Myopia was defined as an SER of at least –0.5 D.
The SER data of the right eye were analyzed. Myopic shift was calculated as the change in SER within one year. The χ 2 and Student t tests were used to compare categorical and numerical data, respectively, in independent groups. We employed multivariate logistic regression with all variables in Table 1 to analyze variables associated with myopia onset in nonmyopic schoolchildren, aiming to adjust for confounders and identify independent predictors. For premyopic children, we used multivariate regression analysis to investigate factors associated with a myopic shift. Statistical significance was set at P < 0.05. All data analyses were performed using commercially available software SAS V 9.1.3 (SAS Institute, Inc., Cary, NC, USA).
| OR (95 % CI) N = 295 | |||||
|---|---|---|---|---|---|
| adjusted estimate | P value | ||||
| ROC program (yes vs. no) | 0.39 (0.21–0.70) | 0.002* | |||
| Premyopia (premyopia vs. hyperopia) | 14.0(1.86–105.3) | 0.010* | |||
| Gender (boys vs. girls) | 0.88(0.49–1.60) | 0.678 | |||
| Age (in years) | 0.99(0.79–1.23) | 0.925 | |||
| Myopic parent (yes vs. no) | 1.15(0.62–2.11) | 0.664 | |||
| Reading/writing (frequent vs. seldom or none) | 1.17(0.63–2.18) | 0.610 | |||
| Computer (frequent vs. seldom or none) | 0.66(0.31–1.40) | 0.279 | |||
| Other near work # (frequent vs. seldom or none) | 1.13(0.54–2.35) | 0.745 | |||
| TV (frequent vs. seldom or none) | 1.11(0.35–3.49) | 0.855 | |||
| Outdoor activity after school (frequent vs. seldom or none) | 0.87(0.48–1.58) | 0.654 | |||
3
Results
A total of 295 nonmyopic children, 146 boys and 149 girls, were included in this study. Of these, 249 were classified as premyopic and 46 as hyperopic. The ROC and control groups consisted of 174 and 121 children, respectively ( Supplementary Table 1 ). The mean age of the participants was 8.58 ± 1.33 years, and the mean baseline SER was 0.38 ± 0.58 D.
Seventy children, 28 and 42 in the ROC and control groups, respectively, developed myopia after 1 year. Multivariate logistic analysis revealed two significant factors associated with myopia onset in these nonmyopic schoolchildren, namely participation in the ROC program (odds ratio [OR] = 0.39, 95 % confidence interval [CI]: 0.21–0.70, P = 0.002) and premyopia status (OR = 14.0, 95 % CI: 1.86–105.3, P = 0.010, Table 1 ). The ROC program reduced the risk of myopia onset by 61 %. However, premyopia status increased the risk of myopia onset by 14 times compared to hyperopic status.
We further analyzed the premyopic subgroup. A total of 249 premyopic children were included, with 138 in the ROC group and 111 in the control groups, respectively ( Table 2 ). The sex distribution was comparable between the two groups, with the ROC group comprising 68 boys (49 %) and 70 girls (51 %), and the control group comprising 54 boys (49 %) and 57 girls (51 %) ( P = 0.922). The mean ages of participants in the ROC and control groups were 8.59 ± 1.32 and 8.82 ± 1.36 years, respectively, with no statistically significant difference between the groups ( P = 0.174). At baseline, the mean SER values were 0.22 ± 0.33 D and 0.17 ± 0.33 D in the ROC and control groups, respectively, with no significant difference between the two groups ( P = 0.190).
