Purchase this article with an account.
Yeunjoo E. Song, Omar García Rodríguez, Kristy Miskimen, Muneeswar Gupta Nittala, Srinivas R. Sadda, William Scott, Dwight Stambolian, Jonathan Haines; Genome-wide Association Study of Choroidal Thickness in the Amish. Invest. Ophthalmol. Vis. Sci. 2018;59(9):1418. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Choroidal thickness (CT) is a promising biomarker for evaluating age-related macular degeneration (AMD) status. We previously confirmed that CT was negatively correlated with the severity of AMD phenotypically and found that it was significantly heritable. However, only one of 52 known AMD risk SNPs showed significant association with CT suggesting that most of the known AMD risk loci are not strongly associated with CT. Therefore, we carried out a genome-wide association study (GWAS) of CT in Amish from Ohio, Indiana, and Pennsylvania to identify new loci for CT.
We performed exome chip genotyping (Illumina MEGAex with custom content) and analyzed common SNPs (>1%) after extensive QC. For the baseline CT analysis, we excluded the subjects with AMD grade ≥ 2 in at least one eye, thus included 629 normal or early AMD subjects. The mean CT of right and left eyes was used with age at exam, sex, spherical equivalent refraction, and the first principle component as covariates. For change in CT from baseline to the two-year follow-up exam, a random eye of right and left eyes in 248 subjects with available follow-up data were used with an additional covariate, CT at baseline. A linear mixed model implemented in the ‘GWAS including families’ (GWAF) program was used to test for association between each CT trait and the exome chip data.
The QQ plot did not suggest evidence of stratification, technical bias, or inflated type I error (λ=1.08). For the baseline CT analysis, no SNPs reached genome-wide significance. However, we identified 6 loci with suggestive associations (p<1.0x10-5). They are either in known genes or downstream variants of known genes (PRDM5; p=5.7x10-6, HEXB; p=4.9x10-6, RP11-445N20.3; p=8.8x10-6, BUB3; p=1.5x10-6, CHST15; p=7.9x10-7, ARL4C; p=5.4x10-6). With change in CT, we found 4 loci across all autosomes, two on chromosome 12, one on chromosome 17, and one on chromosome 19 exhibit genome-wide significance (SLCO1C1; p=2.3x10-08) or suggestive (FAM222A; p=8.4x10-8, LOC102723505; p=5.9x10-8, FPR3; p=9.1x10-8).
We identified 6 loci potentially associated with the baseline CT. Interestingly, 4 variants including the one with the strongest signal are all located on chromosome 10 near the known AMD susceptibility loci HTRA1/ARMS2. We also identified 4 novel loci associated with the change in CT over time. These findings will help us to advance our knowledge of the physiology in AMD.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only