AMD Family History, Genetics Influence Progression Risk

In age-related macular degeneration (AMD), family history and genetics can be used to determine an individual’s composite risk score for disease progression, according to a research team at the University of Massachusetts and Harvard University who published study results in the American Journal of Ophthalmology.

The researchers estimated hazard ratios (HR) for progression from AMD to advanced AMD (AAMD), geographic atrophy, or neovascularization in participants from the Age-Related Eye Disease Study (AREDS), graded at baseline with AREDS severity rubric. Researchers differentiated progression from non-progression with area under the receiver operating characteristic curve (AUC-ROC) and net reclassification improvement (NRI) analyses. Of 4910 eyes, 863 developed AAMD during an up to 12-year follow-up.

Multivariate results revealed those of older age, White race, individuals who smoke, and those with a body mass index of 30 or higher tended toward progression. Greater starting severity and a fellow eye with advanced disease were likely to progress.

AMD family history was an independent predictor — HR for individuals who had 1 relative with AMD was 1.21 (95% CI, 1.02 to 1.43, P =.03), and for those with 2 or more family members it was 1.55 (95% CI, 1.26 to 1.90, P <.001). The researchers suggest common lifestyle and environment or unidentified genetic variants may also apply.

Composite risk score (CRS) involved 21 predisposing factors — relatives with AMD, 12 genetic variants, and other nongenetic aspects — yielding HR=5.57 (90th vs 10th percentile, AUC=0.92), showing better fit than models with non-genetic influences (AUC, P <.001) or ocular status only (NRI, P <.001).

“The value of such a score is that individuals with higher composite risk scores could be monitored more closely and encouraged to adhere to healthier habits,” according to the study authors.

A limitation of the study relates to self-reported history potentially involving absence of siblings or family members who died young. Also, prediction could possibly be overestimated, that is, in AUC and NRI models.

Disclosures: One study author declared affiliations with the biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.