W e read with interest the recent article by Wang et al. evaluating global and regional projections of open-angle glaucoma (OAG) while incorporating the rising prevalence of myopia. The incorporation of exposure dynamics into burden estimation is a useful extension of prior models. Our comments do not dispute the qualitative conclusion that increasing myopia is likely to raise the future burden of OAG. Rather, we emphasize that key modeling assumptions may affect the magnitude, precision, regional distribution, and age-specific attribution of the projected estimates.
The projection framework appears to estimate future OAG prevalence by combining baseline prevalence, projected myopia prevalence, and a pooled relative risk (RR). This approach assumes stable baseline non-myopic risk, transportability of the RR across target populations, and a common myopia–OAG risk function. These assumptions concern parameter, structural, and population-level uncertainty and therefore affect interpretation of the numerical projections. ,
First, stable baseline risk among non-myopic individuals is a structural assumption. Changes in screening uptake, diagnostic criteria or technologies, and healthcare access may alter observed OAG prevalence independently of myopia trends. Scenario analyses allowing baseline risk to vary over time would clarify how much the absolute burden and myopia-attributable estimates depend on this assumption.
Second, using one pooled RR across regions, age groups, and periods assumes that the myopia–OAG association is transportable across target populations. This assumption may be strong if the association differs by age, ethnicity, refractive error severity, diagnostic criteria, or study setting. ,,, Sensitivity analyses allowing the RR to vary, or, where data permit, severity-specific or dose–response estimates, would better characterize this uncertainty.
Third, the model does not appear to consider age-related effect modification. Because age strongly determines baseline OAG risk and may modify the effect of myopia, particularly high myopia, a homogeneous RR could distort age-specific burden and subgroup attribution. , Age-stratified or interaction-based scenarios would help assess the robustness of these estimates.
In light of these considerations, the most informative refinement would be to quantify how sensitive the projections are to key assumptions. At minimum, scenario analyses allowing baseline non-myopic risk and the RR to vary would clarify the dependence of absolute burden and myopia-attributable estimates on these inputs. Where data permit, severity-specific, dose–response, age-stratified, or region-stratified risk estimates could better capture heterogeneity in the myopia–OAG association. , These analyses would not remove uncertainty, but would make its contribution to the projected estimates more explicit.
Overall, Wang et al. provide an important projection of future OAG burden in the context of rising myopia. The issues raised here are unlikely to negate the qualitative conclusion that increasing myopia will contribute to future OAG burden. However, they may affect the magnitude, precision, regional distribution, age-specific burden, and estimated myopia-attributable fraction. Readers should therefore interpret the numerical estimates as model-based projections under specific assumptions, rather than as precise etiologic attribution.
CRediT authorship contribution statement
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
