There are very few population-based longitudinal studies that can accurately assess the ten-year incidence of glaucoma and its risk factors. Moreover, with the increasing worldwide prevalence of high myopia, it is of particular interest to determine whether high myopia is a risk factor for the development of glaucoma in an unbiased, population-based sample. This important study by Wang et al. filled this gap by re-examining 2695 of the 4430 original participants of the Beijing Eye Study (2001-2011). Specifically, they estimated the ten-year incidence of open-angle glaucoma (OAG) to be 3% in adult Chinese participants in the population-based Beijing Eye Study. The strongest risk factor for the development of OAG was high myopia, with a 7.3-fold increased incidence compared to emmetropic eyes. Other risk factors included higher age, IOP, vertical cup disc ratio and thinner central corneal thickness. The incidence estimates are similar to the 3.4% glaucoma incidence rate in the Blue Mountain Study, while slightly lower than the nine-year incidence rate of the Barbardos Eye Study (4.4%)1 and (accounting for shorter follow-up) lower than the Los Angeles Latino Eye Study (four-year incidence 2.3%).2 The association between high myopia and incident glaucoma is consistent with other cross-sectional reports on risk factors associated with the prevalence of glaucoma.
The strongest risk factor for the development of OAG was high myopia, with a 7.3-fold increased incidence compared to emmetropic eyes
Strengths of the study include its unique ten-year follow-up of a large population-based sample, large number of variables assessed in univariable analysis, appropriate and clearly reported multivariable modeling to identify risk factors for incident glaucoma, and its ability to assess high myopia and the experts involved in the assessment of the photograph- based endpoint. A weakness of this report is that some details regarding the definition, reproducibility and comparability of the incident OAG endpoint were not included. The differentiation between highly myopic eyes with and without glaucoma is often challenging, so clear reproducible criteria for defining glaucoma are important. To exclude Beijing Eye Study participants with OAG at the baseline exam from the assessment of OAG incidence, fundus photographs were reviewed by two expert graders to detect glaucomatous optic neuropathy (GON) defined using both absolute criteria, which is similar to the definition recommended by the International Society of Geographical and Epidemiological Ophthalmology, and relative criteria based on retinal nerve fiber layer and neuroretinal rim thinning and retinal caliber, which has not previously been reported. Flicker chronoscopy was used to detect incident glaucomatous optic neuropathy by a panel of three experts. In case of a disagreement, photographs were reassessed up to three times until consensus was reached. Although visual fields were used to define glaucoma in the original the Beijing Eye Study glaucoma prevalence report,3 the authors note as a limitation that visual field test results could not be included in the definition of glaucoma for this ten-year incidence study. Given that glaucoma was defined was photographs alone, additional details on the masked flicker chronoscopy implementation would have been helpful to better interpret these results. For example: (1) was the temporal sequence of the photographs masked (i.e., order of photographs randomized)?; (2) what was the initial agreement among graders for flicker chronoscopy and what proportion of eyes required three reassessments before consensus was reached?; (3) were high myopes over-represented in the eyes requiring consensus?; and (4) how this GON definition may affect comparison with other population-based estimates of the incidence of OAG? In addition, it is unclear why the relative criteria for defining GON was initiated and how many subjects were excluded at baseline due to the relative criteria for GON alone. Given that flicker chronoscopy has been shown to be reproducible and accurate,4,5 it is likely that documentation of these details would lead to a strengthening of the manuscript conclusions.The authors acknowledge another limitation to the study, the participation rate of 66.4%, which is lower than other population-based studies (Blue Mountain 75.6%, and Beaver Dam 82.9%), in part due to the construction of a new airport which led to the relocation of a large number of participants; something that researchers cannot anticipate.In conclusion, this study provides important new information on the ten-year incidence of glaucoma in a Chinese population, and highlights the importance of high myopia as a risk factor for the development of glaucoma. As the prevalence of high myopia increases worldwide, this study has important implications for public health planning of ophthalmic services for glaucoma detection.