Purpose: To investigate the possible roles of season of birth and perinatal duration of daylight hours (photoperiod) in the development of myopia.
Design: Retrospective, population-based, epidemiological study.
Participants: A total of 276 911 adolescents (157 663 male, 119 248 female) 16 to 22 years old. All were Israeli-born conscripts to the Israeli Defense Forces who were examined during the 5-year period 2000 through 2004.
Methods: Noncycloplegic refraction was determined by autorefractometer and validated by qualified optometrists. Myopia, defined on the basis of right eye spherical equivalence, was classified as mild (-0.75 to -2.99 diopters [D]), moderate (-3.0 to -5.99 D), or severe (-6.0 D or worse). The photoperiod was recorded from astronomical tables and classified into 4 categories. Using multivariate logistic regression models, we calculated odds ratios (ORs) for several risk factors of myopia including season of birth.
Main outcome measure: The OR for photoperiod categories as risk factors for myopia.
Results: Overall prevalences of mild, moderate, and severe myopia were 18.8%, 8.7%, and 2.4%, respectively. There were seasonal variations in moderate and severe myopia according to birth month, with prevalence highest for June/July births and lowest for December/January. On multivariate logistic regression, the ORs of photoperiod categories for moderate and severe myopia were highly significant and demonstrated a dose-response pattern. Odds ratios for severe myopia were highest for the shortest versus the longest photoperiods (1.24; 95% confidence interval, 1.15-1.33; P<0.001). Mild myopia was not associated with season of birth or perinatal light exposure. Other risk factors were gender (1.14 for female), education level (1.32 for age above 12), and father's origin (1.31 for Eastern vs. Israeli origin).
Conclusion: Myopia in this population is associated with birth during summer months. The exact associating mechanism is not known but might be related to exposure to natural light during the early perinatal period.