Disordered sleep and myopia among adolescents: a propensity score matching analysis

Purpose: Observational studies have suggested a possible relationship between disordered sleep and myopia, but the association may be subject to selection biases. We aimed to minimize selection biases and assess the association by applying a propensity score matching (PSM) approach.Methods: The study was designed as a school-based cross-sectional study on a Chinese cohort aged 13–14 years in rural areas of China. The Chinese version of the Children's Sleep Habits Questionnaire was used for the measurement of sleep quality and a score of 41 or more was used to define the presence of disordered sleep. Each participant's refractive status was measured after cycloplegia using an autorefractor and myopia was defined as spherical equivalent <−0.50 D. The propensity scores for disordered sleep were formulated using nine potential confounders. We matched the propensity scores for subjects with and without disordered sleep within a caliper of 0.01 of logit function of propensity scores.Results: In this study, 474 pairs (1 subject with vs. 1 subject without disordered sleep) were successfully matched based on propensity scores. The odds ratio (OR) of myopia for disordered sleep before PSM reached conventional levels of statistical significance (OR: 1.43 95% confidence interval [CI] 1.05, 2.58, p = 0.01). After matching, the magnitude of association was reduced and the OR of myopia was not significant (OR: 1.54 95% CI 0.90, 2.57, p = 0.46).Conclusions: At current stage, there are insufficient evidence indicating that disordered sleep could affect the development of myopia in adolescents.