Age of myopia onset and severity of refractive error in later childhood among Singapore children in the SCORM cohort

Abstract

Purpose : To investigate the effect of age of myopia onset and other predictors on the severity of refractive error later in life in the Singapore Cohort Of the Risk factors for Myopia (SCORM) cohort.

Methods : A cohort study, SCORM was conducted in school children aged 7 to 9 years in three schools. Children who were followed up till 11 years and with age of myopia onset 10 years and younger were included in this analysis (n=928). Age of myopia onset was defined either through parent-administered questionnaire at baseline or during eye examinations at subsequent annual follow-up visits. A comprehensive questionnaire determined the child’s near work and parental myopia; while height was measured in the children. Cycloplegic refraction and A-scan ultrasound biometry were measured at every annual eye examination. High myopia was defined as right eye spherical equivalent (SE) of ≤ -5.0D.

Results : In multivariable regression models, children with younger age of myopia onset (per year decrease) were more likely to have higher odds of high myopia (OR= 2.86; 95% CI: 2.39 to 3.43; p<0.001), more myopic SE (β= -0.86D; 95% CI: -0.93 to -0.80; p<0.001) and longer axial length (AL) (β = 0.28mm; 95% CI: 0.24 to 0.32mm; p<0.001) at aged 11 years, adjusting for gender, race, school, books per week and parental myopia. Inclusion of age of myopia onset predicted high myopia by 85% (area under the curve (AUC) of 0.85) and; including other factors of gender, race, school, books per week and parental myopia only marginally improved this prediction (AUC increased to 0.87). A child with all three risk factors such as age of myopia onset 8 years old and younger, who read more than 2 books per week and had at least one myopic parent was defined as high-risk. Compared to children who were not in the high-risk group, those in the high risk group were associated with higher risk of high myopia (OR= 3.66; 95% CI: 2.46 to 5.44; p<0.001), more myopic SE (β= -1.45D; 95% CI: -1.76 to -1.13D; p<0.001) and longer AL (β = 0.53mm; 95% CI: 0.38 to 0.69mm; p<0.001), after adjusting for gender, race and school.

Conclusions : Age of myopia onset was the most important predictor of high myopia in later childhood. Thus, future trials targeting treatments to retard the progression of myopia to high myopia could focus on children with younger onset of myopia and other risk factors.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.