First, Jez Guggenheim summarized current data on myopia prevalence. The global prevalence of myopia is increasing rapidly, particularly in South East Asia, with about 70-80% by the age of 16-18. In Singapore and the big universities in China, more than 90% are reached. In Europe, 40% of the population is myopic (Rudnicka et al., 2016). High myopia increases the risk of visual impairment, due to myopic macular degeneration (MMD), retinal detachment, early cataract and glaucoma.

Second, he reported on possible risk factors of myopia that have been identified by large epidemiological studies. Number 1 is education. There is a clear correlation between myopia and the years of education across several studies. On average, one more year of education and studies causes 0.27 D more myopia. Number 2 is little exposure to bright light (as it is outdoors). Number 3 is too much close work or near work, like reading. It is not sure whether smartphones contribute to this development; if yes, their effect is rather small. Number 4 is genetics: twin studies and molecular studies showed a strong influences of genes in myopia and were suggestive that 70% of myopia is due to genes. But these figures are typical for twin studies (where twins have largely shared visual environments) and for studies in cohorts of children that had similar visual exposure, so that the environmental input was similar. In this case, genes largely determine the differences in myopia among individuals. If there are differences in visual exposure they become the major determining factor of myopia. Furthermore, myopia is caused by an interaction between many genetic and environmental factors (such as near work and outdoor exposure).

Does education cause myopia and why?

A perfect study would be to split a large cohort of children into two random subgroups, one receiving full time education, and the other none - a certainly unethical approach. But such a study would probably show lots of myopia in one group and almost none in the other. Causes for the association between years of education and myopia across studies could be that myopic children are more studious or that children who are genetically predisposed to study more become myopic for this reason. Furthermore, there is the option that socioeconomic factors simultaneously lead to myopia and a higher level of education. In his lecture, Jez Guggenheim described and discussed possible study designs to answer these questions. He explained a number of statistical methods and their strengths and pitfalls.

Luckily, a “natural” experiment was started in 1972, since the legal school-leaving age in England and Wales changed from 15 to 16 years of age (ROSLA: raising of school leaving age study). Study samples were taken from the UK biobank to compare the refractive error in children which were not affected by the law change with children who went to school for an additional year. A limitation of the study was that the sample size was not very high, since refractive error was measured in only 25% of the biobank samples. A regression discontinuity analyses was performed. And the results of the study showed that one additional year of school caused about 0.8D more myopia (0.0 to -1.5D), but the p-value just reached significance.

Mendelian randomization was also used to examine the causal effect of a risk factor (years of education) on a disease (myopia) in observational studies. This approach reduces bias from confounding and reverse causation, to which observational epidemiology studies are susceptible. The sample size for the Mendelian randomization study was larger than in the ROSLA study. The genetic data were taken from large scale genome-wide association studies (GWAS) for educational attainment and myopia, together with the genotypes of participants in the UK Biobank. 74 genetic variants were associated with education in the Biobank samples. Using the Mendelian randomization approach it could be shown that more time spent in education caused myopia, in particular each year of education led to ‑0.27D more myopia” (p-value < 0.001).

Genetics:

Genome-wide association studies are used to identify common genetic variants associated with myopia and refractive error. Each dot in a Manhattan plots represents the results of one statistical test (1 genetic variant). Routinely geneticists assume that “effect size” is the same in every individual, but this is not true since variants may have no effect in “unexposed” individuals. This means that the effect is not the same in everybody. To solve this problem, composite quantile regression (CQR) can be used. Using this approach it was shown that most (88%) of the genetic variants associated with refractive errors have different effects in different individuals, e.g. a higher effect in myopes and a lower in emmetropes and high hyperopes. A potential causes for this difference can be gene-environment or gene-gene interactions. In fact it was shown that education is adding to the genetic risk.

Can genetics be used to predict myopia?

There are millions of genetic variants. Current GWAS studies for refractive errors can explain 5.5% of the risk of becoming myopic, the age of onset of spectacle wear 5.5% and years of education 0.1%. In summary, if all factors are combined, about 9-10% of the risk can be explained up to date. Taking the polygenetic risk score into account the ODDs ratio (risk) in myopes (top 25%) is increased 3fold and in high myopes (top 25%) 4.2fold.

Summary

In conclusion, Jez Guggenheim stated that two lines of evidence suggest that education causes myopia.

He demonstrated that genetic variants associated with myopia have different effects depending on the refractive error: higher effects in myopes and lower effects in emmetropes. Genetics can be used to detect children at high risk.

(Summary by Andrea Carrillo Aleman, Sandra Gisbert, Marita Feldkaemper, Frank Schaeffel)

About Prof. Jez Guggenheim

Director of Research, School of Optometry and Vision Sciences, University of Cardiff, UK

The research of Prof. Jez Guggenheim is aimed at discovering the causes of short-sightedness (myopia). Approximately 1 in 3 people in the United Kingdom suffer from myopia, and the condition increases the risk of a number of serious eye disorders. Prof. Jez Guggenheim has contributed to collaborative efforts with researchers from across the world, leading to the discovery of more than a hundred genetic variants that increase susceptibility to myopia. Nobody yet knows how most of these genetic variants play a role in myopia development, nor how best they can be used to target and improve treatment.