Longevity headlines often ask for one percentage. Science rarely gives us the luxury of one percentage that survives every definition.
Twin research shows why: the answer changes depending on whether we model all deaths together or try to separate ageing-related mortality from accidents, infections and other external causes.
Quick answer
Is lifespan 25% genetic or more than 50% genetic?
Those numbers refer to different models. Classic twin studies estimated total observed lifespan variation at roughly one quarter genetic, while a 2026 analysis modelled 'intrinsic' lifespan after accounting for extrinsic mortality and obtained a higher estimate.
Key takeaways
- A classic Danish study estimated lifespan heritability around 23–26%.
- A 2026 Science analysis argues that intrinsic lifespan heritability exceeds 50% after modelling extrinsic mortality.
- The studies answer different statistical questions and should not be treated as a simple contradiction.
- Neither estimate tells an individual how long they will live.
The classic result: environment dominates observed lifespan
Historical Danish twin cohorts found that most variation in age at death was not explained by genetic differences.
The non-shared part was especially large — a reminder that even siblings raised together accumulate very different lives.
The 2026 challenge: what if accidents dilute ageing genetics?
The newer Science paper argues that deaths from strongly extrinsic causes can mask the heritability of intrinsic survival processes.
Its higher estimate is therefore not simply a re-run of the old analysis with a bigger sample; it relies on a different conceptual separation.
TwinPare perspective: uncertainty is part of the story
The responsible takeaway is not 'genes matter more than we thought' or 'lifestyle decides everything'. It is that lifespan is a layered outcome and its heritability depends on what exactly we define as lifespan.
That makes personal prevention and health behaviour relevant without pretending they can guarantee longevity.
Source notes
The sources have been verified and editorially reviewed for this article. The limitations below show which level of conclusion the sources support.
- [herskind-1996] The heritability of human longevity: a population-based study of 2872 Danish twin pairs born 1870-1900 A. M. Herskind et al.. Human Genetics, 1996. Evidence type: Population-based Danish twin lifespan analysis Limitation: Classic twin modelling estimated moderate longevity heritability around one quarter in this historical cohort; the estimate depends on model and definition. PubMed DOI
- [shenhar-2026] Heritability of intrinsic human life span is about 50% when confounding factors are addressed Ben Shenhar et al.. Science, 2026. Evidence type: Mathematical modelling and twin-cohort analyses separating intrinsic from extrinsic mortality Limitation: This newer estimate addresses intrinsic lifespan after modelling extrinsic mortality and should not be presented as directly overturning every earlier lifespan estimate; it answers a narrower modelling question. PubMed DOI
Editorial source review
This section shows how the article's key factual claims are linked to the source.
Phrasings that require caution
- Do not mix total lifespan heritability with intrinsic-lifespan modelling.
- Do not present heritability as a personal life-expectancy percentage.
| ID | Claim | Source support | Caution |
|---|