Abstract
Purpose :
To examine whether associations between a glaucoma polygenic risk score (GPRS) and various glaucoma traits is modified by age.
Methods :
We included UK Biobank participants with demographic information, genetic data, and various glaucoma trait data: intraocular pressure (IOP; n=118,153), macular retinal nerve fiber layer (mRNFL; n=42,132), macular ganglion cell inner plexiform layer (mGCIPL; n=42,042), and prevalent glaucoma status (n=192,283). We generated a GPRS for each individual using 2,673 genetic variants and their estimated effect sizes reported by a recent multi-trait genomic analysis. To assess modification by age of the relation between GPRS and glaucoma traits, we used multivariable linear regression for IOP, mRFNL, and mGCIPL; multivariable logistic regression for prevalent glaucoma risk; and analyzed the associations by age quartiles (Q1: <51, Q2: 51-57, Q3: 58-62, and Q4: ≥63 years), with tests for interactions. All analyses were adjusted for age, age2, sex, ethnicity, social deprivation, body mass index, blood pressure, diabetes, cardiovascular disease, spherical equivalent, and various lifestyle factors. We also sub-stratified interaction effects of GPRS and age for each glaucoma trait by sex.
Results :
For IOP and glaucoma, interactions between GPRS and age were significant (P-interactions<2.3e-4). The association between the GPRS and IOP increased with age (mean differences in IOP (mmHg) per standard deviation (SD) of GPRS = 0.95, 1.03, 1.18, and 1.24 for 1st, 2nd, 3rd, and 4th quartile of age; P<2.5e-25). Similar trends were observed with glaucoma risk (odds ratio per SD of GPRS = 2.39, 2.58, 2.80, and 2.75 for 1st, 2nd, 3rd, and 4th quartile; P≤0.001). However, for mRNFL or mGCIPL, we did not observe any interactions between GPRS and age (P-interactions>0.05). We observed that the interactions of GPRS and age were significant (P-interaction<0.02) within each sex stratum, with no differences by sex (P-interaction of PRS, age, and sex = 0.70 for IOP and 0.68 for glaucoma).
Conclusions :
A glaucoma genetic predisposition score was more strongly associated with higher IOP and greater glaucoma prevalence in older versus younger adults in the UK Biobank. These data may inform how to prioritize glaucoma screening in genetically predisposed individuals.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.