Association Between Sleep Status and Myopia in Children and Adolescents: A Cross-sectional Study in Shenzhen

Purpose: The aim of this cross-sectional study was to assess the association between sleep status and myopia in children and adolescents. Methods: Stratied cluster sampling a total of 30,188 schoolchildren from 14 schools located in 6 streets (Xinxiang, Songgang, Shiyan, Fuyong, Shajing and Xinan) of Baoan District in Shenzhen, China. The demographic characteristics of schoolchildren, information of self-reported myopia and their sleep status in the last 1 month were collected through using a standardized questionnaire. χ2-test was used to compare differences in myopia rates among children and adolescents in different sleep status. Logistic regression models were used to analyze the association of sleep status with myopia in children and adolescents. Results: The self-reported myopia rates was 49.8% among children and adolescents in Shenzhen. In primary, junior high and senior high students, the myopia rates were 25.6%, 62.4%, and 75.7%, respectively. After controlling for confounders, the results of multiple logistic analysis showed that night sleep duration < 7h/d (OR = 2.02, 95%CI: 1.87-2.19), falling asleep at a xed time (no) (OR = 1.14, 95%CI: 1.08-1.20), getting up at a xed time in the morning (no) (OR = 1.28, 95%CI: 1.19-1.37), putting off sleeping on weekends ≥ 2 h/d (OR = 1.28, 95%CI: 1.17-1.41) and Putting off getting up on weekends ≥ 2 h/d were risk factors (OR = 1.23, 95%CI: 1.13-1.34). After stratied by grade, social jet lag was not associated with myopia among senior high students. Conclusion: Insucient sleep duration was positively associated with self-reported myopia, irregular patterns of sleep and social jet lag increased the risk of myopia among schoolchildren. Myopia prevention and control among children and adolescents should not only ensure adequate sleep, but also develop a habit of regular sleep.


Introduction
Myopia is a common disease that occurs mainly in childhood and early adulthood, and it is also known as short-sightedness or near-sightedness 1 . Because prevalence of myopia has rapidly risen globally over the past three decades, it has been described as an epidemic disease and has become an important public health problem 2 . A large number of studies have shown that myopia is associated with many factors, such as outdoor activities, near-work activities, educational pressure and parental myopia [3][4][5][6] . Considering the high incidence of myopia, there is an increasing need to explore other environmental factors that may in uence the development of myopia.
Sleep is an important physiological process of the human body, and good sleep is a necessary condition to improve physical and mental health of children and adolescents. Several recent studies suggest that sleep may be potentially associated with myopia in children and adolescents. For instance, a crosssectional study of 3, 625 Korean adolescents aged 12 to 19 years indicated that per 1 hour increase in sleep duration was associated with a 10% lower risk of myopia 7 . Difference, similar to jet lag between sleep pattern on weekends and school days are known as "social jet lag" 8, 9 . In children and adolescents, it is common that weekend sleep patterns are not as same as the study days 10,11 . Objective measurements of sleep and light collected every 6 months for 14 days in Australian children aged 10 to 15 years revealed that non-myopic children had more regular sleep status (seldom staying up late and putting off getting up), while myopic children showed greater variability, whether it was the study day or weekend, summer or winter 12 .
Therefore, we propose a hypothesis that poor sleep status, such as lack of sleep time, and social jet lag (mainly manifested in the inconsistency between OR values and 95% CI, with P < 0.05 considered statistically signi cant, using nighttime sleep duration, falling asleep at a xed time in the night, getting up at xed time in the morning, putting off sleeping on weekends, and putting off getting up at weekends as independent variables. Table 1 showed the basic characteristics of this sampling. The self-reported myopia rate of children and adolescents (mean age = 12.44, SD = 3.47) in Shenzhen was 49.8% (15, 026/30, 188),,and the myopia rate of female students (54.2%) was higher than that of male students (46.3%), and the difference was statistically signi cant; the self-reported myopia rates among primary, junior high and senior high students were 25.6%, 62.4%, 75.7%, respectively, which increased with the grade, and the difference was statistically signi cant; the myopia rate of students with higher father and mother education level was lower, and the myopia rate of students with poor family economic level was lower, and the difference was statistically signi cant. Note: a is for χ 2 , b is for χ 2 -trend Table 2 showed the outcome of univariate analysis for sleep status and self-reported myopia. The self-reported myopia rates of students with night sleep duration ≥ 9, 8-9h/d, 7-8 and < 7h/d were 26.0%, 27.8%, 43.2%, and 68.6%, respectively, which increased with the decrease of night sleep duration (P < 0,001).

Results
And the effect of social jet lag on myopia was signi cant (all P value < 0.001). After subgroup analysis were performed by grade, in the primary school group, not falling asleep at a xed time at night, not getting up at a xed time in the morning, putting off falling asleep ≥ 2h/d on weekends, and putting off getting up ≥ 2h/d on weekend was associated with self-reported myopia (all P values < 0.001); in the junior high group, social jet lag was not associated with myopia (all P values > 0.05); in the senior high group, not getting up at a xed time was related to self-reported myopia (P < 0.05).  Table 3 and Fig. 1 showed the association between sleep status and the risk of self-reported myopia: Controlling of factors such as sex, age (years), outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia, the results of logistic analysis showed that the OR (95% CI) for myopia was 2.  Note. Model 1: adjusted for sex, age (years), outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia.
Model 2: strati ed by grade, adjusted for sex, age (years), outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia.
After strati cation by grade (Table 3, Fig. 2-4 Figure 1 Forest plot of the association between sleep status and myopia in children and adolescence, adjusted for sex, outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia.

Figure 2
Forest plot of the association between sleep status and myopia in primary group, adjusted for sex, outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia.

Figure 3
Forest plot of the association between sleep status and myopia in junior high group, adjusted for sex, outdoor activity, academic record, study burden, family economic level, father's education, mother's education, father myopia, mother myopia. Figure 4