Abstract
Purpose :
To investigate the effect of age of myopia onset, environmental factors and a myopia genetic risk score on refractive error, axial length and high myopia
Methods :
Studies in the Consortium for Refractive Error and Myopia (CREAM) with data on age of myopia onset, refractive error, and ocular biometry at adolescent or early adult examination, together with environmental factors and genetic loci for refractive error were included. Data from individuals with conditions that could alter refraction were excluded. The effect of the above factors were considered separately and collectively in multivariable regression models:
Dependent variable = β0 + β1 Age of myopia onset + β2 Age at visit + β3 gender + β4 genetic risk score + β5 outdoor + β6 nearwork
Three outcomes were considered - spherical equivalent (SE), axial length (AL) adjusted for height, and the dichotomous outcome of high (SE <-5.0D) vs mild to moderate myopia (SE <-0.5D). A fixed-effect meta-analysis was performed with sub-analysis by age at visit and ethnicity.
Results :
Complete phenotypic and genotypic data from four European and two Asian studies were included in the meta-analysis (n=1947). In a basic model age of myopia onset, adjusted for age at visit and sex, was significantly associated with both SE (β 0.13, 95% CI 0.11–0.15)and high myopia (odds ratio (OR) 0.90, 95% CI 0.87–0.90). When the genetic risk score and environmental factors were included in the full model, age of onset remained significant with a comparable effect for both SE (β 0.10, 95% CI 0.08-0.12) and high myopia (OR 0.92, 95% CI 0.88–0.96). Total SE variance explained ranged from 0.06 to 0.54 in the basic and 0.09 to 0.54 in the full model. Mean area under the curve statistics in high myopia models ranged from 0.55 to 0.92 in the basic and 0.61 to 0.93 in the full model, with higher estimates in studies comprised of older participants. Age of myopia onset was significantly associated with axial length in all studies but this became non-significant in the meta-analysis
Conclusions :
Age of myopia onset was a significant determinant of SE, AL and high myopia. In explanatory models, only moderate benefit was gained incorporating genetic and environmental factors compared to age of myopia onset alone. This is clinically relevant - early onset myopia is most likely to progress to high myopia, andpotentially, visually disabling pathologic myopia in later life.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.