📄 genetic and genomic medicine

Heritability of Age-Related Macular Degeneration in the Amish

This study demonstrates that Age-Related Macular Degeneration (AMD) has a substantial genetic heritability of approximately 50% in the Amish population, confirming a significant polygenic effect beyond known risk variants through both SNP and pedigree-based analyses.

Original authors: Moore, N. C., Song, Y. E., Gulyayev, A. V., Miskimen, K., Miron, P., Laux, R. A., Lynn, A., Fuzzell, S. L., Hochstetler, S. D., Miller, D., Caywood, L. J., Clouse, J. E., Herington, S. D., Wang, P., L

Published 2026-08-06

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Moore, N. C., Song, Y. E., Gulyayev, A. V., Miskimen, K., Miron, P., Laux, R. A., Lynn, A., Fuzzell, S. L., Hochstetler, S. D., Miller, D., Caywood, L. J., Clouse, J. E., Herington, S. D., Wang, P., Liu, Y., Dorfsman, D. A., Vance, J. M., Nittala, M. G., Sadda, S. R., Stambolian, D., Scott, W. K., Pericak-Vance, M. A., Haines, J. L.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine your body as a massive, bustling city. Sometimes, parts of that city start to wear down, like old streetlights flickering out or roads developing potholes. One specific neighborhood that can get damaged as we get older is the center of our eyes, a tiny spot called the macula. When this spot gets sick, it's called Age-Related Macular Degeneration, or AMD. It's like the central hub of a camera lens getting cloudy, making it hard to see faces or read signs. Scientists have long known that while things like smoking or getting older can make this damage worse, there's also a hidden blueprint inside our DNA that plays a huge role. Think of this blueprint as a set of instructions passed down from parents to children. For a long time, researchers tried to figure out exactly how much of the "damage" is caused by these genetic instructions versus the environment. They used to look at twins—like two copies of the same blueprint—to guess the answer, but that method can sometimes be tricky because twins often share very similar lives, not just genes. To get a clearer picture, scientists are now looking at larger families and using new tools to read the genetic code directly, trying to separate the "nature" part from the "nurture" part of the story.

This paper takes a fresh look at that genetic puzzle by studying a very special group of people: the Amish. Imagine the Amish community as a giant, extended family tree where everyone is related to everyone else, kind of like a massive game of "Six Degrees of Kevin Bacon" but with real cousins, aunts, and grandparents stretching back generations. Because they have kept detailed family records for centuries and live in a way that keeps their environment very consistent (like eating similar foods and avoiding smoking), they are the perfect "laboratory" for scientists to see how much of AMD is written in the genes. The researchers gathered 525 people from this community, checked their eyes to see who had the disease and who didn't, and then used two different methods to measure the genetic influence. One method looked at the actual DNA letters (called SNPs) to see how much they matched up with the eye problems, while the other method used the family tree records to see how the disease ran in the branches of the family.

The results are like finding two different maps that point to the same treasure. The DNA-based map suggested that about 55% of the variation in who gets AMD is due to genetics, while the family-tree map suggested about 49%. Both numbers are very close, and both are statistically significant, meaning it's highly unlikely these results happened by pure chance. The authors found that even when they accounted for the few known "super-risk" genes that are famous for causing AMD, the numbers didn't change much. This suggests that the story isn't just about a few bad apples; instead, it's a "polygenic" effect, meaning hundreds or thousands of tiny genetic nudges, each doing a little bit of work, add up to a big influence. The study also checked if the different ways they found people (some in Pennsylvania, some in Ohio/Indiana) changed the results, and it didn't.

One interesting twist the paper explores is why the family-tree number wasn't higher than the DNA number. Usually, family trees show a bigger genetic effect because they include shared environments. But here, the authors suspect the Amish lifestyle is so uniform—everyone eats well, doesn't smoke, and lives on family land for generations—that there isn't much "environmental noise" to confuse the signal. This actually makes the family tree a very clean way to measure genetics. The study concludes that AMD is indeed highly heritable, supporting what we thought from twin studies but confirming it with a different, perhaps more reliable, approach. While the sample size of 525 people is a bit small for a definitive final answer, the findings strongly suggest that there is a massive amount of genetic risk waiting to be discovered beyond the few big genes we already know about. The authors are careful to say this "suggests" a large polygenic effect rather than proving it beyond a shadow of a doubt, but the evidence points clearly toward a future where we need to look at the whole genetic landscape, not just the peaks, to understand this eye disease.

Technical Summary: Heritability of Age-Related Macular Degeneration in the Amish

Problem Statement
Age-related Macular Degeneration (AMD) is a leading cause of central vision loss with a strong genetic component. While heritability estimates have historically relied on twin studies, these designs are susceptible to ascertainment biases, such as the overrepresentation of monozygotic, female, and concordant twins, potentially leading to inflated estimates of genetic influence. Furthermore, existing estimates vary widely (ranging from 45% to 71%) depending on disease stage and study design. To clarify the genetic burden of AMD and identify the extent of undiscovered genetic susceptibility, alternative approaches utilizing broader pedigree structures and relationship-informed genome-wide analyses are necessary.

Methodology
This study utilized data from the Amish Eye Study, a founder population characterized by high endogamy, extensive genealogical documentation, and reduced environmental variability. The analysis focused on 525 individuals (285 controls, 240 cases) aged 50 or older from Amish communities in Ohio, Indiana, and Pennsylvania.

Phenotyping: AMD status was determined via clinical eye exams and imaging graded according to the Beckman scale. Cases included both intermediate and advanced AMD; controls had no AMD.
Genotyping: DNA was genotyped using Illumina arrays (MEGAex or GSA), resulting in 338,575 high-quality SNPs after quality control.
Heritability Estimation: The study employed two parallel approaches:
1. SNP-based Heritability ( $h^2_{SNP}$ ): Genetic relationships were identified using KING software (up to second-degree relatives). The LDAK TetraHer software was used to estimate heritability on the liability scale, adjusting for population prevalence (12.3%), age, sex, and the first principal component (PC1) to account for population structure.
2. Pedigree-based Heritability ( $h^2_{ped}$ ): Genealogical data from the Anabaptist Genealogical Database (AGDB) were used to construct three-generation pedigrees. These were analyzed using Sequential Oligogenic Linkage Analysis Routines (SOLAR) software, with results post-hoc transformed to the liability scale.
Sensitivity Analyses: Models were constructed to test robustness by including binary carrier statuses for three known high-risk AMD variants (A69S, Y402H, and R1210C) as covariates in various combinations. Stratified analyses were also performed for intermediate and advanced AMD stages.

Key Results

Primary Estimates: The study estimated SNP-based heritability at 55% ± 13% ( $p = 9.87 \times 10^{-6}$ ) and pedigree-based heritability at 49% ± 18% ( $p = 3.06 \times 10^{-4}$ ).
Model Comparison: A likelihood ratio test comparing a reduced model (SNP-based only) to a full model (SNP + pedigree kinship matrix) yielded a significant result ( $p = 1.03 \times 10^{-4}$ ), indicating that the inclusion of the pedigree kinship matrix improved model fit and captured additional phenotypic variance.
Impact of Known Variants: Including the major risk variants (A69S, Y402H, R1210C) as covariates did not substantially alter the heritability estimates (SNP estimates ranged 54–61%; pedigree estimates ranged 48–50%). This suggests that these large-effect variants do not overshadow the aggregate contribution of other genetic factors.
Stratified Analysis: Estimates for intermediate AMD were higher ( $h^2_{SNP} = 63\%$ , $h^2_{ped} = 62\%$ ) compared to advanced AMD ( $h^2_{SNP} = 37\%$ , $h^2_{ped} = 0.97\%$ ), though the authors note that large standard errors in these stratified groups reduce confidence in these specific figures.

Significance and Claims
The paper claims that these heritability estimates corroborate existing twin and SNP-based studies, confirming a substantial genetic contribution to AMD (approximately 50–55%). The study highlights several specific contributions:

Validation of Methods: It demonstrates that SNP-based and pedigree-based methods are effective, accessible alternatives to twin studies for estimating heritability in founder populations.
Polygenic Nature: The finding that heritability estimates remain robust even after adjusting for known large-effect variants supports the hypothesis that AMD is driven by a large polygenic effect, where numerous small-effect variants collectively contribute more to heritability than individual rare or common high-risk variants.
Population Representativeness: The results from the Amish population are strikingly similar to those reported for the general Non-Hispanic White population, suggesting that the Amish are a representative model for studying AMD genetics and that findings may be generalizable to the wider population.
Environmental Control: The study notes that the stable lifestyle of the Amish (diet, lack of smoking/alcohol) likely reduces environmental variability, potentially limiting environmental bias in pedigree-based estimates.

The authors conclude that while the sample size is a limitation, the study provides updated heritability predictions and reinforces the necessity of searching for additional genetic risk loci beyond the currently known major variants.

Technical Summary: Heritability of Age-Related Macular Degeneration in the Amish

More like this