This prospective cohort study was conducted across 26 elementary schools from September 2015 to January 2017 in Chiayi, Taiwan. The 5417 participants were aged 7 to 11 years and underwent examinations for refraction and eye health as well as interviews about lifestyle factors. The cohort included students without myopia. Refraction and eye health examinations were repeated the following year to determine the incidence of new cases of myopia and associated risk factors. Furthermore, ocular alignments were also measured the following year to reveal the factors associated with horizontal heterophoria.
Informed consent form
The local administration of the Education and School Board were contacted to request their cooperation. This study was approved by the Institutional Review Board of Chang Gung Foundation (Number: 201700887B0C501). It followed the tenets of Declaration of Helsinki. Informed consent was obtained from the parents or guardians of all children before the examination.
All students were screened in the same environment: a classroom with blackout curtains in the selected school. To reduce individual error, the same examiner team conducted all questionnaires and examinations. The questionnaires collected data on students’ gender, weight, height, sleep duration, eye-related symptoms, and history of exposure to atropine. Noncycloplegic refractive errors were assessed with an autorefractor (Autorefractometer ARK-1, Nidek Co., LTD., Aichi, Japan).6,18 Students with a history of ocular and physical pathologies, strabismus, or amblyopia were excluded. Refractive errors were defined by spherical equivalent (SE) refraction, calculated by adding the spherical diopter to one half of the cylindrical diopter. Myopia was defined as an SE of −1.0 diopter or greater in one or both eyes. Myopia was categorized as low myopia (−1.0 to −3.0 diopters) and high myopia (−3.0 diopters or less). Ocular alignment was assessed by observing corneal reflexes (Hirschberg test) and using the monocular cover-uncover test, in which each eye is briefly covered while the examiner watches for any movement in the opposite, uncovered eye that would indicate heterotropia. If no movement in the uncovered eye is noted, movement in the covered eye when the cover is applied and movement in the opposite direction (fusional movement) when the cover is removed indicates heterophoria. If the patient has heterophoria, the eye remains straight before and after the cover-uncover test; the deviation appears during the test as the result of an interruption of binocular vision. A participant with heterotropia, however, starts and ends the test with the same eye deviated, or in case of alternating heterotropia, ends the test with the opposite eye deviated.
This study included patients with SEs of less than −1.0 diopter. New myopia was defined as myopia of greater than −1.0 diopter. Patients were further analyzed regarding demographic data and risk factors with particular focus on ocular alignment. Adjusted odds ratios for all risk factors are presented.
The statistical analysis was conducted using SPSS Statistics 20.0. The data were presented as numbers (%) for fractions and as means with standard deviations for continuous variables where appropriate. A univariate logistic regression was used to assess the effects of the categorical variables on new cases of myopia. A multinomial logistic regression was then performed. The division of new myopia cases in terms of ocular alignment was compared using the chi-squared test. A Cox hazard ratio model was then used to validate the associated factors for changes in ocular alignment. A P value of <.05 was considered significant.