Abstract
Purpose :
To investigate for evidence of effect measure modification (EMM) by race/ethnicity on the association between myopia and primary open angle glaucoma (POAG); in other words, to determine whether race/ethnicity modifies the association between myopia and POAG.
Methods :
A cross-sectional study was conducted in the 100% sample of 2019 California (CA) Medicare beneficiaries from the Centers for Medicare & Medicaid Services. Inclusion criteria were: age 65 or older, CA residence, and active coverage with Medicare Part B. The outcome of interest was POAG and the primary exposure was myopia, both defined by having International Classification of Diseases, Tenth Revision diagnosis codes for their respective conditions. To test for statistically significant EMM, a multivariable logistic regression model was created with myopia, POAG, and the following covariates: age, sex, race/ethnicity, systemic disease burden defined by Carlson Comorbidity Index (CCI) score, and an interaction term between myopia and race/ethnicity. To compare the effect of myopia on POAG modified by race/ethnicity, separate multivariable models were fitted for the effect of myopia on POAG, stratified by race/ethnicity, and adjusted for age, sex and CCI.
Results :
The study population included 2,717,346 CA Medicare beneficiaries. About 1.9% (n = 52,517) of beneficiaries had a diagnosis of myopia and 37.2% (n = 1,010,539) identified as non-White. Subjects with myopia had 2.15 times the adjusted odds of POAG (95% confidence interval [CI], 2.09-2.21) compared to subjects without myopia, and there was evidence of statistical interaction between myopia and race/ethnicity (p<0.0001). In our multivariable models stratified by race/ethnicity, myopia was associated with greater adjusted odds of POAG than those without myopia within Black (odds ratio [OR]: 2.49, 95% CI, 2.19-2.82), Hispanic (OR: 3.13, 95% CI, 2.93-3.34), and Asian subjects (OR: 2.22, 95% CI, 2.05-2.41), as compared to White counterparts (OR: 1.92, 95% CI, 1.86-1.99).
Conclusions :
In the CA Medicare population, myopia is associated with greater odds of POAG among minorities as compared to White subjects, after adjusting for covariates. There is significant EMM between race/ethnicity and myopia on the likelihood of having POAG. We cannot discern whether this is completely due to health disparities or other factors, such as gene-environment interactions.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.