July 2019
Volume 60, Issue 9
Free
ARVO Annual Meeting Abstract  |   July 2019
Improving the genetic prediction of myopia and refractive error using educational attainment
Author Affiliations & Notes
  • Neema Ghorbani Mojarrad
    School of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales, United Kingdom
  • Cathy Williams
    Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, England, United Kingdom
  • Jeremy A Guggenheim
    School of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales, United Kingdom
  • Footnotes
    Commercial Relationships   Neema Ghorbani Mojarrad, None; Cathy Williams, None; Jeremy Guggenheim, None
  • Footnotes
    Support  PhD Scholarship, College of Optometrists
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 4816. doi:https://doi.org/
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Neema Ghorbani Mojarrad, Cathy Williams, Jeremy A Guggenheim; Improving the genetic prediction of myopia and refractive error using educational attainment. Invest. Ophthalmol. Vis. Sci. 2019;60(9):4816. doi: https://doi.org/.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Genetic prediction models derived from Genome Wide Association Studies (GWAS) for refractive error (RE) can be used to predict children at risk of developing myopia and who would benefit most from intervention. Currently, the best genetic prediction model can explain 7.8% of the variance in RE (Tedja et al. 2018 Nature Genetics:50;834) when tested in an independent sample. In this population-based study, we hypothesised that combining GWAS summary statistics for RE with GWAS summary statistics for educational attainment, which is genetically correlated to RE, would improve the accuracy of genetic prediction.

Methods : GWAS summary statistics for RE and educational attainment from the following GWAS analyses of European individuals were combined using MTAG software: (1) GWAS for ‘True RE’ (measured by autorefraction) in 95,619 participants from UK Biobank; (2) GWAS for ‘Predicted RE’ (inferred from self-reported age-of-onset-spectacle-wear) in 287,448 participants from UK Biobank; (3) Social Science Genetic Association Consortium GWAS for educational attainment (‘EduYears’) in 328,917 participants reported by Okbay et al. (2016 Nature:533;539). RE prediction models were generated with LDpred software, and tested in an independent sample of 1,516 adult participants (ALSPAC Mothers cohort) to calculate the variance explained (R2). For prediction of myopia, area under the receiver operating characteristic curves (AUROC) were calculated for myopia levels of ≤-0.75D and ≤-3.00D.

Results : The variance explained (R2) by genetic prediction models for each trait alone, or combined using MTAG, was: True RE=5.6%; Predicted RE=5.5%; EduYears=0.1%; combined True RE and Predicted RE=8.9%; combined True RE, Predicted RE and EduYears=9.4%. Variance explained was improved by inclusion of EduYears (8.9% vs. 9.4%; P=0.004). For prediction of myopia (≤-0.75D), the AUROC was: combined True RE and Predicted RE=0.65; combined True RE, Predicted RE and EduYears=0.66 (the latter AUROC was significantly higher, P=0.02). For predicting moderate myopia (≤-3.00D) inclusion of EduYears did not improve prediction (AUROC=0.72 with and without EduYears).

Conclusions : Our results show that combining GWAS data for RE and educational attainment significantly improves the genetic prediction of RE. Future studies may benefit from combining information from genetically correlated traits to improve prediction accuracy.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×