Your DNA May Decide More of Your Lifespan Than You Think — and That Changes Everything

For three decades, the prevailing scientific consensus held that genes account for roughly 20–25% of the variation in human lifespan, leaving the rest to behavior, environment, and chance. That estimate, anchored by a landmark 1996 study of Danish twins, became the intellectual foundation for billions of dollars in public health campaigns, employer wellness programs, and lifestyle medicine — all built on the premise that individual choices are the primary lever for living longer.

A study published in Science in January 2026 has sharply revised that estimate upward, arguing that when confounding factors are stripped away, the heritability of human lifespan is closer to 55% [1]. The implications ripple outward from genetics labs into insurance boardrooms, longevity clinics, and legislatures — raising uncomfortable questions about who benefits from the idea that your health is mainly in your own hands.

The Study: What Changed and Why

The research, led by Ben Shenhar from the laboratory of Uri Alon at the Weizmann Institute of Science in Israel, used mathematical modeling combined with data from three large Scandinavian twin registries — including, for the first time in this context, twins raised apart [2]. The study's central innovation was separating deaths into two categories: intrinsic mortality (deaths from biological aging, genetic mutations, and age-related disease) and extrinsic mortality (deaths from accidents, infections, homicides, and environmental hazards) [3].

Previous estimates, including the 1996 Danish Twin Study by Herskind and colleagues that analyzed 2,872 twin pairs born between 1870 and 1900, had placed heritability at 0.26 for males and 0.23 for females [4]. But those cohorts lived through eras when infectious disease, workplace accidents, and war killed many people before their genes could meaningfully influence lifespan. By lumping all deaths together, earlier studies effectively diluted the genetic signal with environmental noise.

The Weizmann team designed "virtual twin" simulations to model what would happen as extrinsic mortality declined — as it has, dramatically, over the past century. When extrinsic deaths were removed from the analysis, the heritability of remaining (intrinsic) lifespan jumped above 50% [1].

"The number that we got is not out of nowhere," Shenhar told NBC News. "If you look at twin studies on pretty much anything in humans, you get this 50%." [3]

The finding is a significant departure not only from the 1996 Danish estimate but also from a 2018 analysis by Calico Life Sciences (Alphabet's longevity subsidiary). That study, led by Graham Ruby and using pedigree data from over 400 million people on Ancestry.com, concluded heritability was below 10% — and possibly as low as 7% — after controlling for assortative mating, the tendency of people with similar lifespans to marry each other [5].

A Three-Way Scientific Disagreement

The field now has three competing estimates: below 10% (Calico, 2018), 20–25% (Danish Twin Study, 1996), and approximately 55% (Weizmann, 2026). The disagreement is not trivial — it reflects fundamentally different assumptions about what heritability means and what it measures.

Kaare Christensen, who directs the Danish Twin Research Center and contributed to the original 1996 study, called the Weizmann paper "an interesting mathematical exercise" but cautioned that it measures something different from traditional heritability estimates. "In reality, people experience both intrinsic and extrinsic mortality," Christensen noted [6]. Stripping out extrinsic deaths produces a heritability figure for a hypothetical world, not the one people actually live in.

Graham Ruby of Calico acknowledged the work but emphasized it measures "lifespan under specific conditions" rather than absolute lifespan [3].

Eric Verdin, president and CEO of the Buck Institute for Research on Aging, raised a specific methodological concern: genetic susceptibility to infections such as COVID-19 could arguably be classified as intrinsic mortality, since the immune system's response is partly inherited. Reclassifying even some infection-related deaths would lower the heritability estimate [3]. However, when the Weizmann researchers reanalyzed their data to account for age-related vulnerability to infections and falls, the estimate remained near 50% [3].

Source: World Bank / WHO Global Health Observatory

Data as of Dec 31, 2023CSV

The Known Biology — and Its Limits

Despite the higher heritability estimate, the specific genes driving longevity remain largely unidentified. Only a handful of genetic variants have been consistently replicated across studies: FOXO3, a transcription factor regulating cellular stress responses and DNA repair; APOE, a lipid transport protein whose protective ε2 allele appears in 20–25% of centenarians compared to 8% of the general population; and SIRT6, a sirtuin involved in DNA repair whose rare variants in centenarians show enhanced activity [7] [8].

These genes do not operate as simple on/off switches. FOXO3 influences energy metabolism, oxidative stress response, inflammation, and cell death pathways. Centenarians do not universally carry favorable variants of these genes, suggesting that multiple genes interact in ways researchers have not yet mapped [3].

Separate from inherited DNA sequence, biological aging is tracked through several molecular markers. Telomere length — the repetitive DNA caps on chromosomes that shorten with each cell division — correlates with biological age, though its predictive value for mortality is limited [9]. Epigenetic clocks, which measure patterns of DNA methylation (chemical modifications that change gene expression without altering the underlying sequence), have proven more robust predictors. The GrimAge epigenetic clock, in particular, outperforms telomere length in predicting mortality across multiple US cohorts [9]. Mitochondrial function — the efficiency of cells' energy-producing organelles — also declines with age and is partly heritable [10].

The Weizmann study does not identify which of these mechanisms dominates at specific life stages. That gap matters: if epigenetic drift is the primary driver in middle age while telomere attrition matters more in late life, interventions would need to be staged accordingly.

The Counterargument: Lifestyle Still Moves the Needle

The strongest pushback comes from a 2025 study published in Nature Medicine by Argentieri and colleagues, who analyzed proteomic data — measurements of how proteins change over time — from nearly 500,000 UK Biobank participants [11]. Their conclusion pointed in the opposite direction: environmental factors accounted for approximately 17% of lifespan variation, while genetic factors contributed less than 2%. Age and sex explained about 50% [12].

Among 164 environmental exposures assessed, smoking, socioeconomic status, physical activity, and living conditions showed the strongest associations with premature death and accelerated biological aging. Smoking alone was linked to 21 age-related diseases; socioeconomic factors were associated with 19 [12].

A 2024 analysis in BMJ Evidence Based Medicine found that a healthy lifestyle may offset the effects of life-shortening genes by more than 60%, adding approximately 5.5 years of life expectancy at age 40 for those at high genetic risk [13]. A 2023 study in The Lancet Healthy Longevity reported similar findings in Chinese older adults, with healthy lifestyles adding 4.35 years at age 65 even in the highest genetic risk group [14].

The four lifestyle factors most consistently associated with longevity gains across studies: not smoking, regular physical activity, adequate sleep, and a healthy diet [13].

The Weizmann researchers do not dispute these findings. "The message of our paper is not that lifestyle, exercise and diet are not important. That is not our message, not at all," Shenhar said [3]. If genetics explain 55% of intrinsic lifespan, the remaining 45% attributable to lifestyle and environment is "a not insignificant proportion" — one that can translate to a decade or more of additional life [3].

Source: Herskind et al. 1996 / Ruby et al. 2018 / Shenhar & Alon 2026

Data as of Jan 29, 2026CSV

An $80 Billion Industry Built on Personal Choice

The study arrives at a moment when the global anti-aging and longevity market is valued at approximately $80 billion and projected to exceed $1.1 trillion by 2035 [15]. The US longevity supplements market alone reached $6.35 billion in 2025 [16]. The broader wellness economy exceeds $6 trillion, with the longevity-focused segment forecast to reach $610 billion by 2026 [17].

Employer wellness programs constitute an $8 billion domestic industry, covering more than 63 million American workers through roughly 90% of large employers [18]. These programs — which offer financial incentives for health screenings, exercise commitments, and smoking cessation — are built on the assumption that behavior meaningfully predicts health costs.

Yet a large randomized trial tracking 33,000 employees at BJ's Wholesale Club found that wellness program participants self-reported healthier behaviors but showed no significant differences in health measures, healthcare spending, or absenteeism after three years [19]. If the genetic contribution to health outcomes is substantially larger than previously assumed, the already-questionable return on investment for these programs becomes harder to justify.

Who Gets Hurt by Genetic Fatalism

The history of "your genes determine your health" arguments carries baggage. The tobacco industry spent decades using personal responsibility framing — casting smoking-related illness as a matter of individual choice rather than product design — to deflect regulation and litigation [20]. The rhetorical structure of genetic determinism can serve a similar function: if biology is destiny, then corporate responsibility for environmental toxins, unhealthy food, or workplace hazards diminishes.

No evidence has surfaced linking the Weizmann study's authors or funders to industry groups with a financial interest in undermining behavioral health interventions. The study was conducted at a publicly funded academic institution. But the framing risk is real: research showing that genes matter more than behavior can be selectively cited by industries seeking to weaken regulation.

The equity implications are equally significant. Life expectancy varies by roughly 20 years between the wealthiest and poorest US counties [21]. Black Americans have a life expectancy approximately five years shorter than white Americans, a gap driven overwhelmingly by social determinants — access to healthcare, environmental exposures, chronic stress, income — rather than genetic difference [22]. Life expectancy in Jamaica and Barbados exceeds that of Black Americans despite much lower national incomes, a finding that undercuts genetic explanations for racial health disparities [22].

The Weizmann study drew its twin data from Scandinavian registries — populations that are relatively ethnically homogeneous and benefit from universal healthcare and strong social safety nets. Whether the 55% heritability estimate holds in populations facing structural disadvantage is an open question. Heritability is not a fixed biological constant; it is a population-level statistic that changes with environmental conditions. In a society where everyone has equal access to nutrition, healthcare, and clean air, genetic differences explain more of the remaining variation. In a society with vast inequalities, environment dominates [22].

The Race to Hack Biology

If the ceiling imposed by genetics is higher than thought, the commercial incentive to breach that ceiling grows proportionally. In January 2026, the FDA cleared Life Biosciences — cofounded by Harvard geneticist David Sinclair — to begin the first human clinical trial of a partial epigenetic reprogramming therapy. The treatment, ER-100, uses three Yamanaka factors (Oct4, Sox2, and Klf4) to reset age-associated epigenetic markers in cells without causing them to lose their specialized identity [23]. The Phase 1 trial targets vision restoration in patients with glaucoma and non-arteritic anterior ischemic optic neuropathy [24].

Other programs are close behind. NewLimit, backed by Coinbase co-founder Brian Armstrong, claims to be nearing a clinic-ready epigenetic reprogramming therapy targeting liver cells [25]. YouthBio completed a productive FDA meeting regarding its Alzheimer's gene therapy, YB002 [26]. Junevity plans first-in-human trials of its JUN_01 therapy in the second half of 2026 [26].

Source: OpenAlex

Data as of Apr 1, 2026CSV

The cost and accessibility of these therapies remain unknown. Gene therapies already on the market for other conditions — such as Zolgensma for spinal muscular atrophy — carry price tags exceeding $2 million per treatment. If longevity interventions follow a similar trajectory, they will be available first to the wealthy, widening the gap between those who can and cannot buy additional years of life.

Sofiya Milman, a professor at the Albert Einstein College of Medicine who studies centenarian biology, framed the goal as making longevity-promoting genetic mechanisms "accessible to people who didn't win the genetic lottery" [6]. Whether that aspiration survives contact with pharmaceutical pricing remains to be seen.

What Changes If Policymakers Listen

If the 55% heritability estimate gains traction in policy circles, several structures built on the assumption of personal responsibility for health face pressure.

Insurance underwriting: Life insurers already price policies partly on behavioral factors like smoking status, with premiums for smokers running up to four times higher [27]. The Affordable Care Act permits health insurers to impose tobacco surcharges of up to 50% [28]. If genetic factors explain more variance than behavior, the actuarial basis for behavioral pricing weakens — while pressure to incorporate genetic testing into underwriting grows. Most countries currently prohibit genetic test results in insurance underwriting, but this regulatory wall faces increasing strain as direct-to-consumer genetic testing proliferates [27].

Public health campaigns: Government spending on behavioral health messaging — anti-smoking campaigns, dietary guidelines, physical activity promotion — rests on the premise that changing behavior changes outcomes at scale. A higher genetic contribution doesn't eliminate the value of these campaigns, but it could reduce their expected effect size and, by extension, the political will to fund them.

Employer wellness mandates: The legal framework allowing employers to incentivize or penalize health behaviors — upheld under the ACA's wellness program provisions — assumes those behaviors are meaningfully within individual control. Genetic determinism arguments could support legal challenges to wellness program penalties, particularly from employees who face higher health costs despite compliant behavior.

Tort liability: If genetics explain more of disease risk, plaintiffs suing corporations for health damages (from pollution, product defects, or occupational hazards) may face arguments that their conditions were genetically predetermined. This echo of the tobacco industry's decades-long defense strategy is perhaps the most legally consequential implication [20].

The Limits of the Evidence

Several caveats bear emphasis. The Weizmann study is a modeling exercise based on existing twin data, not a new empirical cohort study. Its conclusions depend on the validity of its assumptions about how to separate intrinsic from extrinsic mortality — assumptions that critics like Christensen view as debatable [6].

Heritability estimates describe population-level variance, not individual destiny. A 55% heritability figure does not mean 55% of any one person's lifespan is genetically determined. It means that, across a population, 55% of the differences in lifespan can be statistically attributed to genetic differences — under the specific environmental conditions of that population.

The study's Scandinavian twin registries, while high-quality, represent populations with relatively low socioeconomic inequality and universal healthcare. Extrapolating to the United States, where health outcomes diverge sharply by race and income, requires caution.

Finally, the study does not identify actionable genetic targets. Knowing that genes matter more does not yet translate into knowing which genes, in which combinations, at which life stages, could be modified to extend life.

What the study does accomplish is reopening a question many researchers thought was settled. The answer — still unresolved — will shape how societies allocate responsibility for health, how industries price risk, and who gets access to the therapies designed to rewrite biological fate.