Your Genes Shape How Long You Live — Far More Than We Thought

Here is a question scientists have been arguing about for decades: How much does your DNA determine how long you’ll live?

The intuitive answer is a lot. We all know families where people routinely reach their nineties, and families where the men seem to die of heart disease in their sixties. We look at our grandparents and see our own futures. The idea that longevity “runs in families” is as old as medicine itself.

The scientific answer, until very recently, was surprisingly little. About 7 to 25%, depending on the study. Not much more than eye color or musical talent. Your lifestyle, your environment, your access to healthcare — these things mattered far more than the DNA you inherited.

Now a new paper in Science has challenged that consensus head-on — and the challenge is persuasive. When you account for a systematic bias hiding in every previous study, the heritability of human lifespan doesn’t just nudge upward. It more than doubles.

---

The Twin Studies Paradox

The classic tool for measuring heritability is the twin study. The logic is elegant: identical (monozygotic) twins share 100% of their DNA. Fraternal (dizygotic) twins share about 50%. If a trait is heavily genetic, identical twins should be far more similar to each other than fraternal twins are.

For lifespan, studies going back decades have consistently found that identical twins aren’t that much more similar than fraternal twins. The landmark Danish study by Herskind and colleagues in 1996 — following 2,872 twin pairs born between 1870 and 1900, all of whom had died by the time of the follow-up — estimated heritability at about 26% for men and 23% for women. Similar estimates came from Swedish and Finnish twin registries.

Then in 2018, Kaplanis and colleagues at Calico Life Sciences (Google’s aging research company) did something bolder: they analyzed 86 million pedigree records from the genealogy website Ancestry.com. The result was sobering. After accounting for the way genetic relatives tend to choose similar partners and environments (assortative mating), they estimated heritability of lifespan at just 7%.

Seven percent. For a trait that felt so obviously heritable, that was a striking result. It seemed to vindicate the idea that your fate is mostly in your own hands — your diet, your exercise, your luck in avoiding disease. Genes, it seemed, were almost irrelevant.

But a team at the Weizmann Institute of Science in Rehovot, Israel thought something was wrong.

---

The Hidden Flaw: Not All Deaths Are Equal

The insight at the heart of the new study is both simple and, in retrospect, obvious: not everyone dies of old age.

When Herskind followed those 2,872 Danish twin pairs, some of the twins died of cancer at 68. Some died of heart disease at 74. And some died in accidents, from infectious disease, from violence — deaths that have nothing to do with how well their bodies manage the slow biological processes of aging. All of these deaths were treated identically in the heritability calculation.

This matters enormously. If an identical twin is killed in a car accident at 40, his death gives you essentially no information about his genes’ effects on aging. His twin might have lived to 90. By including that pair in your analysis as “twins who died at different ages,” you artificially inflate the dissimilarity in the dataset. That makes identical twins look less similar to each other than they really are — which pulls the heritability estimate down.

In statistical terms, extrinsic deaths (from accidents, infections, violence, and other random causes) act as noise. They dilute the genetic signal. And because twins — identical or fraternal — have roughly equal exposure to extrinsic mortality, this noise affects both types of pairs similarly, muddying the comparison.

This is what Ben Shenhar, Uri Alon (a professor famous for his work on the mathematics of biological networks), and their collaborators set out to correct.

---

The Fix: Mathematical Modeling Plus the Right Twin Data

The Weizmann team, working with collaborators at the Karolinska Institutet in Stockholm and Leiden University Medical Center in the Netherlands, developed a mathematical model to separate the two components of mortality: intrinsic mortality (the accumulation of biological damage that underlies aging) and extrinsic mortality (deaths from causes that are essentially random with respect to genes).

They then applied this model to twin cohort data — crucially including twins who were raised apart as well as twins raised together. This matters because twins raised together share not just genes but also environment: the same house, the same food, the same parents’ smoking habits. Using twins raised apart lets you tease out the purely genetic component more cleanly.

The result? When you isolate intrinsic mortality and recalculate, the heritability of human lifespan climbs to above 50%.

The paper, published in Science on January 29, 2026, also identified a second, subtler confound: a “cutoff age” effect. Most heritability studies set a minimum age of death — say, they only include twins who died after age 30. But which deaths you include turns out to matter nonlinearly. The younger the deaths you include, the more accidents and infections dominate, and the more heritability gets diluted. Earlier studies that included many young deaths were systematically underestimating genetic contributions.

---

50% — What Does That Actually Mean?

At first, 50% might sound modest. But in the context of human genetics, it’s quite high.

Heritability measures what fraction of the variation in a trait among individuals can be attributed to genetic differences. Height is one of the most heritable human traits we know of — around 79-80% heritable. Intelligence quotient has estimates ranging from 50 to 80%. Body mass index sits around 40-70%. Personality traits like extraversion typically fall in the 40-60% range.

Lifespan at 50%+, if the new estimate holds up, puts it firmly in the mainstream of complex human traits. It suggests that, on average, about half the reason some people live longer than others comes down to the genes they were born with. The other half comes from environment, lifestyle, and chance.

That’s a striking recalibration. It doesn’t mean your habits don’t matter — they clearly do. But it does mean that searching for the genetic levers of longevity is a scientifically serious endeavor, not a futile hunt for a few percent of variance.

---

Why This Changes the Stakes

The practical implications run deep.

If lifespan heritability is only 7%, then genome-wide association studies (GWAS) hunting for longevity genes are chasing a ghost. There’s simply not enough genetic variance to find. The scientific resources would be better spent on lifestyle interventions.

But if heritability is above 50%, the calculus flips. There are real, discoverable genes with meaningful effects on lifespan. And because heritability estimates how much of the variation is genetic, higher heritability means that longevity genes, when found, will have larger and more consistent effects — making them better targets for medicine.

We already know some of these genes exist. APOE — the gene whose ε4 variant dramatically increases Alzheimer’s risk — has well-documented effects on late-life mortality. FOXO3 variants are enriched among centenarians in multiple human populations. CETP and PCSK9 variants affect cardiovascular mortality through cholesterol pathways. But these explain only a tiny fraction of lifespan variation. The new heritability estimate says there is much more to find.

The authors note that their finding aligns lifespan with the rest of the animal kingdom. Heritability of lifespan in model organisms — fruit flies, nematode worms, mice — is well above 50%, often higher. The idea that humans were somehow special in having low heritability always felt biologically implausible. The new work suggests we weren’t special at all; we were just measuring wrong.

---

A Scientific Caution

Like all good science, this result isn’t the end of the conversation.

A preprint critique by Sergey Kornilov, published in April 2026, raises a concern about the mathematical approach. Kornilov argues that the model doesn’t account for heritable susceptibility to extrinsic causes — for example, people with certain genetic variants might be more prone to infectious diseases, or poor impulse control that raises accident risk. If such heritable extrinsic frailty is omitted from the model, Kornilov argues, the intrinsic heritability estimate could be inflated by around 9 percentage points.

This is a legitimate statistical concern, and the scientific community will need to work through it. But even if the critique is partly correct and the true intrinsic heritability lands somewhere around 40-45%, that would still represent a substantial revision upward from the 7-25% consensus. The fundamental message — that genes matter far more than previously thought — would survive.

This is science doing what it should: researchers proposing a bold, well-reasoned correction to a standing result; critics poking at the methods; the truth working its way out through argument and evidence.

---

The Quiet Revolution in Aging Research

Something is shifting in how scientists think about aging. A field that once seemed destined to be dominated by lifestyle advice and incremental medicine is increasingly looking like a field with deep, discoverable molecular structure.

The identification of the hallmarks of aging — senescence, mitochondrial dysfunction, epigenetic drift, chronic inflammation — has given researchers a map. Drugs like rapamycin and senolytic compounds extend lifespan in mice through mechanisms that translate to human biology. Epigenetic clocks can measure biological age more precisely than a birth certificate.

And now, if the Weizmann team’s result stands, we know that roughly half of what makes one person age differently from another is written in their genome. That half is, in principle, findable. And things that are findable can eventually be acted upon.

Your grandparents’ longevity may be telling you something real after all.

---

Source: Shenhar B, Pridham G, De Oliveira TL, Raz N, Yang Y, Deelen J, Hägg S, Alon U. “Heritability of intrinsic human life span is about 50% when confounding factors are addressed.” Science, Vol. 391, Issue 6784, pp. 504–510 (2026). DOI: 10.1126/science.adz1187