What is heritability in the context of disease risk?

In a given population, different people vary in their predisposition to develop a certain disease. The proportion of this variation between people that can be attributed to genetic factors, as opposed to environmental ones, is called “heritability.” Heritability is a statistical concept that describes the extent to which genetic factors explain these risk variations between people in a specific population. To be clear, it does not explain the genetic risk for an individual to develop a certain disease.

When we talk about a disease having a certain heritability, we are describing the balance between genetic risk factors, and environmental ones - both of which can contribute to the development of a given disease. As more risk factors accumulate, that individual will pass a threshold making them more susceptible and likely to develop that disease.

How do we measure disease heritability?

Heritability estimates are single numbers ranging from 0.0 to 1.0. An estimate close to 0.0 indicates very little contribution from genetics, while a heritability estimate close to 1.0 suggests that nearly all the variability in risk can be attributed to genetics, with little contribution from environmental factors.

Heritability extremes are infrequent, and most diseases fall somewhere in the middle between 0.0 and 1.0.

Consider Crohn’s disease. The heritability of Crohn’s disease is estimated at 0.75, or 75%. This essentially tells us that 75% of the variations in risk between individuals in a population is explained by genetic factors. To be clear: this does not tell us that 75% of Crohn’s disease is determined by genetics, nor that an individual’s risk for getting Crohn’s disease is 75% if their parents have it, too. It does suggest that between people, genetic factors influence the variations in risk for Crohn’s disease more than environmental factors.

Most complex diseases, like Alzheimer’s, breast cancer, obesity, or inflammatory bowel disease (IBD) each have a certain degree of disease heritability. Obesity, for example, is highly heritable, indicating a stronger genetic contribution to variations in risk for this disease than environmental factors. But we also know that an individual’s body mass index can be reduced by good nutrition, exercise, and healthy habits. So, a disease with high heritability doesn’t necessarily predetermine one’s fate.

Heritability estimates apply on a population level, not to you as an individual

Heritability is a complex topic to grasp, and it is fraught with misconceptions. The most common misconception is that it describes the magnitude to which genetic and environmental factors dictate an individual’s risk. But, heritability estimates don’t work this way. They function only at a population level.

A disease with a 50% heritability does not mean that 50% of the reason an individual develops that disease is determined by that person’s genetics, while the other 50% is determined by their environment. It simply reflects how much variation between people with that disease is due to underlying genetic factors. To take it further, we can’t assume that a parent with a severe case of Crohn’s disease will also have a child with a severe case. While the likelihood might be higher given a higher heritability estimate, we still have to factor in that child’s nutrition and lifestyle, which we know strongly influences the ability to mitigate the severity of the disease.

Heritability estimates are not constant and depend on the specific population and geography

Another misconception is that heritability is fixed, but heritability estimates can actually change over time, and between people and places.

For example, the heritability of Crohn’s disease in the Ashkenazi Jewish population is two to four times higher than that of non-Jewish European ancestry. Intuitively, this makes sense: different ethnic populations have different genetic backgrounds and live in different environments, which can tip the scale. Imagine if we looked at the heritability of Crohn’s disease 100 years ago in the Ashkenazi Jewish population, it’s likely that even within the same population that the heritability of the disease would be different from what it is now. Since heritability estimates are predicated on the balance between environmental and genetic factors, it follows that a changing environment will also change the heritability of a given disease.

How do we study heritability?

There are a few approaches to estimating heritability, but twin studies have been a cornerstone of this research for decades. Identical twins share 100% of their DNA, while fraternal twins only share 50% of their DNA on average – similar to regular siblings.

All twin studies ask the following question: “Do we expect individuals who are more genetically similar to also be similar in developing some disease?” If yes, then it’s likely due to genetics. If not, then any difference can be attributed to differences in either their environment or pure chance. By correlating the genetic similarity between identical twins to a trait or disease, one can isolate the genetic effect from the environmental one.

For example, if the correlations are stronger between identical twins who developed Crohn’s disease compared to fraternal twins, then we can infer that their risk of developing the disease was largely influenced by genetics. We can then use these correlations between identical and fraternal twins to estimate the heritability of Crohn’s disease.

From twin studies to GWAS

More recently, scientific advances in Genome-wide Association Studies (GWASs) have allowed us to take a twin-free approach to estimating heritability by delving more deeply into the human genome. These genome-wide studies have revolutionized our understanding of complex diseases and our ability to estimate their heritabilities.

The goal of a GWAS is to identify genetic markers that are associated with a given disease. The most common type of genetic marker is called a single nucleotide polymorphism (SNPs, pronounced “snips”) - a SNP is a variation at any single position in our DNA sequence. Individual SNPs contribute tiny amounts to genetic risk, but the cumulative effects of many SNPs added together can have a large impact on disease risk.

So in GWASs, scientists scour the genomes of thousands of people in search of SNPs unique to a population with a given disease and compare them to a population without it. Basically, SNP data provides novel information about how genetically similar individuals are with a certain disease, which in turn can be used to estimate heritability. This “twin-free” heritability approach based on sets of SNPs found in GWASs is called SNP-based heritability.

Unlike disease heritability which infers genetic similarity to disease risk through closely related relatives, SNP-based heritability is estimated from unrelated individuals who participate in GWASs. This is what makes it such a powerful approach.

Since GWASs are the new gold standard, it’s important to understand SNP-based heritability and how we use these estimates in the context of your genetic risk to disease.

What is SNP-based heritability?

SNP-based heritability represents how much of the variation in disease risk is captured by common SNPs in a specific population. It is measured in the same way as disease heritability, with a number between 0.0 and 1.0.

Let’s go back to our Crohn’s disease example. We know that the disease heritability of Crohn’s disease is 0.75 (75%), but the SNP-based heritability is only 0.37 (37%). We can interpret this as “37% of the variation in risk for Crohn’s disease can be attributed to a common set of SNPs that this population shares.”

You also probably noticed that the SNP-based heritability was less than half of the disease heritability. This discrepancy exists for almost all complex diseases, and SNP-based heritabilities are almost always lower than disease heritability. The gap between these two numbers simply means that we don’t yet understand all of the genetic risk factors involved in a specific disease. Given the speed at which the field of genetics is evolving, we should be able to close that gap and refine our heritability estimates as scientists uncover new information.

Why is this important for you?

One of the most exciting advances that have stemmed from GWASs and SNP-based heritabilities are polygenic risk scores (PRS). A PRS is an estimate of an individual’s genetic risk and can help predict whether someone, like yourself, is at risk of developing a given disease (here we mean you, and not the population!).

Since most diseases have a genetic component, it’s important to assess what percentage of a disease outcome can be explained by PRS. Some conditions are less heritable, so the ceiling for utility is lower. Other conditions have moderate to high heritability, and a PRS is able to capture a meaningful portion of the genetic component. Recognizing this, we carefully curate genetic risk predictors for chronic diseases with high heritability and a PRS risk distribution in which the highest percentiles of risk are at least 2.5 times the baseline rate of developing the disease. With these risk predictors, we can make informed choices about our families, personalize preventative measures, and attempt to overcome certain predispositions with lifestyle alterations and better practices.

If you’re interested in diving deeper into some of the science, we suggest reading the following articles

• Visscher, Peter M., William G. Hill, and Naomi R. Wray. "Heritability in the genomics era—concepts and misconceptions." Nature reviews genetics 9.4 (2008): 255-266.
• Yang, Jian, et al. "Concepts, estimation and interpretation of SNP-based heritability." Nature genetics 49.9 (2017): 1304.
• Baselmans, Bart ML, et al. "Risk in relatives, heritability, SNP-based heritability and genetic correlations in psychiatric disorders: a review." Biological Psychiatry (2020).

‍