Abstract
Purpose :
High myopia is characterized by a disproportional axial elongation in childhood and teenage years. To help identify high risk groups, we developed a model based on non-invasive measures to predict early onset myopia.
Methods :
In the population-based birth-cohort study Generation R 4,636 children underwent ophthalmic examination at 6 and 10 years of age, including measurement of axial length and corneal curvature. Variables related to nearwork and outdoor exposure, SES, demographic characteristics were assessed by questionnaire. 2,360 children underwent cycloplegic refractive error measurements, and myopia was defined as average spherical equivalent of ≤-0.5D in ODS. A risk score was created using beta’s from linear regression models of associated variables. The accuracy of the risk score (area under the curve) of the discrimination between myopes and non myopes was estimated using receiver operation curves.
Results :
Average axial elongation was 0.21 mm/year. Seven parameters were independently associated with axial elongation: myopic parents, books read per week, time spent reading, time spent outdoors, sports, ethnicity, and AL/CR ratio at baseline. Prevalence of myopia increased from 2.3% at 6 years to 11.3% at 10. The risk score ranged from 3 – 30. Children with a score of ≤8 had a risk of 1.2% to develop myopia; children with a risk score of ≥15.5 had a risk of 55%. The discriminative accuracy of the model was 0.76.
Conclusions :
This model based on easy to obtain variables is an accurate tool to identify children at risk for high myopia. It can be used to recommend early life style changes and intervention in high risk groups to prevent high myopia.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.