Format

Send to

Choose Destination
Genetics. 2017 May;206(1):119-133. doi: 10.1534/genetics.116.196998. Epub 2017 Mar 24.

Bivariate Analysis of Age-Related Macular Degeneration Progression Using Genetic Risk Scores.

Author information

1
Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pennsylvania 15261.
2
Division of Pulmonary Medicine, Allergy and Immunology, Department of Pediatrics, Children's Hospital of Pittsburgh of UPMC, University of Pittsburgh, Pennsylvania 15224.
3
Department of Biostatistics, Center for Statistical Genetics, University of Michigan, Ann Arbor, Michigan 48109.
4
Department of Statistics and Actuarial Science, University of Waterloo, ON, Canada N2L 3G1.
5
The EMMES Corporation, Rockville, Maryland.
6
Neurobiology Neurodegeneration and Repair Laboratory, National Eye Institute, National Institutes of Health, Bethesda, Maryland 20892.
7
Casey Eye Institute, Oregon Health & Science University, Portland, Oregon 97239.
8
Division of Epidemiology and Clinical Applications, National Eye Institute, National Institutes of Health, Bethesda, Maryland.
9
Department of Biostatistics, Graduate School of Public Health, University of Pittsburgh, Pennsylvania 15261 weeks@pitt.edu wei.chen@chp.edu.
10
Department of Human Genetics, Graduate School of Public Health, University of Pittsburgh, Pennsylvania.

Abstract

Age-related macular degeneration (AMD) is a leading cause of blindness in the developed world. While many AMD susceptibility variants have been identified, their influence on AMD progression has not been elucidated. Using data from two large clinical trials, Age-Related Eye Disease Study (AREDS) and AREDS2, we evaluated the effects of 34 known risk variants on disease progression. In doing so, we calculated the eye-level time-to-late AMD and modeled them using a bivariate survival analysis approach, appropriately accounting for between-eye correlation. We then derived a genetic risk score (GRS) based on these 34 risk variants, and analyzed its effect on AMD progression. Finally, we used the AREDS data to fit prediction models of progression based on demographic and environmental factors, eye-level AMD severity scores and the GRS and tested the models using the AREDS2 cohort. We observed that GRS was significantly associated with AMD progression in both cohorts, with a stronger effect in AREDS than in AREDS2 (AREDS: hazard ratio (HR) = 1.34, P = 1.6 × 10-22; AREDS2: HR = 1.11, P = 2.1 × 10-4). For prediction of AMD progression, addition of GRS to the demographic/environmental risk factors considerably improved the prediction performance. However, when the baseline eye-level severity scores were included as the predictors, any other risk factors including the GRS only provided small additional predictive power. Our model for predicting the disease progression risk demonstrated satisfactory performance in both cohorts, and we recommend its use with baseline AMD severity scores plus baseline age, education level, and smoking status, either with or without GRS.

KEYWORDS:

AMD progression; AREDS; bivariate time-to-event; genetic risk score; risk prediction

PMID:
28341650
PMCID:
PMC5419464
DOI:
10.1534/genetics.116.196998
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for HighWire Icon for PubMed Central
Loading ...
Support Center