# Heritability is not the genetic percentage of a person

Draft • Research checked 4 October 2026 • AI-assisted DomDNA editorial content. No independent clinical review or publication is claimed. General education, not individual medical advice.

“This trait is seventy per cent heritable” can sound like a statement about how much of one person's trait is fixed by DNA. Heritability is instead a statistic about differences among people in a studied population and environment.

That distinction changes how you read a headline. A population estimate does not divide an individual into a genetic portion and a lifestyle portion. It also does not tell you whether a trait is easy to change.

## Translate the sentence before using the number

The [MedlinePlus heritability explanation](https://medlineplus.gov/genetics/understanding/inheritance/heritability/) describes the statistic in terms of variation and notes its dependence on the population and environment. It is not an individual prediction.

Consider a hypothetical study of a measured trait in one group. Researchers estimate that genetic differences account for a substantial share of the variation they observe under their model. This tells you something about differences within that study context.

It does not mean that each person's trait contains the same genetic percentage. Nor does it tell you which specific variants matter or which changes would alter an individual's outcome.

A helpful translation adds three words to the headline: “in this population”. Then add the measured trait, the setting and the method. The number becomes a finding with a scope instead of a universal property.

## Similar families are not automatically a genetic explanation

Relatives can share environments as well as DNA. A habit, language or exposure can run in a family without being explained by inherited sequence differences.

Twin studies use patterns of resemblance to investigate such questions. The [Polderman and colleagues meta-analysis](https://pubmed.ncbi.nlm.nih.gov/25985137/) aggregated published twin-study estimates across many traits. Its abstract describes a large research synthesis; it does not provide one percentage that should be attached to every individual or every condition.

We use the paper here to show the level at which these estimates are produced. Its broad summary should not become a claim that parents, environments or circumstances never matter. The model and the trait remain essential.

For an everyday reader, “it runs in my family” can be a reason to document a pattern. It is not itself a heritability estimate.

## The environment can change what variation is visible

Imagine a hypothetical group whose access to an environmental resource is very similar. Compare it with another group where access varies greatly. Even if the same underlying biology matters, the sources of observed differences can have a different balance.

This example is editorial reasoning, not a claim about a particular population. It explains why a statistic about variation depends on the range of circumstances in the sample. It does not imply that environmental influences disappear when an estimate is high.

A primary [twin study of immune-system variation](https://pubmed.ncbi.nlm.nih.gov/25594173/) examined measured immune features and reported substantial non-heritable influences. That finding belongs to its measured endpoints and participants. It cannot be transferred into a personal immunity score or a general claim that genetics is irrelevant.

The two research examples address different sets of traits and questions. Their differences are not a contradiction requiring one to be discarded.

## Heritable does not mean unchangeable

A statistic explaining differences under one set of circumstances is not an intervention trial. It does not test whether a medicine, training programme or environmental change improves an outcome.

If a headline uses heritability to sell a supposedly DNA-matched routine, look for a separate study of that routine and its relevant outcome. A heritability estimate cannot supply missing evidence of benefit.

It also cannot tell someone how much effort is worthwhile. Translating a population statistic into a moral judgement about personal responsibility is scientifically unhelpful and can be harmful.

## What you can do with this

Your next step is to annotate one heritability headline with its population, trait and study design. If any is missing, keep the claim provisional rather than filling the gaps from a different study.

For a practical decision, identify the evidence that actually tests the proposed action. The question “Does this change help this outcome?” needs a different kind of answer from “What explains variation in this study?”

Keeping those questions separate lets genetics inform curiosity without turning a percentage into a personal limit.

For general lifestyle learning, [DomDNA’s educational quiz](https://domdna.com/quiz) does not estimate heritability or assign a genetic percentage to your answers.