The record that has not moved in three decades
Any conversation about the maximum human lifespan has to start with a stubborn empirical fact. Jeanne Calment died in 1997 at 122 years and 164 days. Nobody has come close since. In 2016, Dong, Milholland and Vijg analysed global demographic data and reported two patterns that are easy to state and hard to argue with: improvements in survival tend to decline after age 100, and the age at death of the world's oldest person has not increased since the 1990s.[4] Life expectancy at birth has risen dramatically over the same period; the extreme tail has not moved at all.
That is the backdrop against which a new modelling paper landed this summer, and it is why the coverage — "humans could live to 194" — inverted the paper's actual mood. The study, Somatic mutations impose an entropic upper bound on human lifespan, was published in npj Aging by Evgeniy Efimov, Vlad Fedotov, Leonid Malaev, Ekaterina Khrameeva and Dmitrii Kriukov of the Skolkovo Institute of Science and Technology and the Artificial Intelligence Research Institute in Moscow.[1] It does not predict how long anyone will live. It asks a narrower and more useful question: of all the things that kill us slowly, how much of the ceiling is set by DNA damage alone?
What the lifespan model actually did
The design is what the authors call an incremental modelling framework. Rather than fitting one equation to human mortality, they built a model of population survival dynamics and then switched aging mechanisms on one at a time, measuring what each addition costs in years. The scenario they report is a counterfactual: assume every hallmark of aging has been solved — inflammation, proteostasis, stem-cell exhaustion, the lot — and leave only the irreversible accumulation of somatic mutations, the DNA errors that arise in ordinary body cells during a lifetime and are never inherited.[1]
Their baseline is the strangest number in the paper and the one most often misread. An organism whose mortality risk does not rise with age still dies eventually, from accident and background hazard; in their framework that produces a median lifespan of 1,759 years. It is not a prediction. It is a yardstick — the length of the ruler against which each aging mechanism's cost is measured. Somatic mutations alone shorten it to roughly 156 years.[1]
Two conclusions follow, and they point in opposite directions. Mutations are enormously costly: they remove more than 90% of the theoretical span. And mutations are surprisingly weak: 156 is still nearly double the observed human median of around 80, which means something other than DNA damage is doing roughly half the work of killing us. The authors state it plainly — the reduction is "substantial yet incomplete," implying comparable contributions from other hallmarks.[1]
The asymmetry that matters: replaceable versus irreplaceable cells
The most defensible finding in the paper is not the headline number but the organ asymmetry underneath it. Tissues do not age at one rate, and the model makes the reason mechanical rather than mysterious.
| Tissue type | Cell turnover | Modelled outcome under mutation-only aging[1] |
|---|---|---|
| Neurons, cardiomyocytes | Post-mitotic — essentially never replaced | Critical bottleneck; median lifespan ≈ 156 years |
| Liver and other proliferating tissue | Continuous replacement | Functional for thousands of years; mutation-driven decline effectively neutralised |
| Multi-organ integration | Mixed | Median 146–194 years |
The logic is simple enough to check by hand. A liver cell that acquires a catastrophic mutation dies or is outcompeted, and a neighbour divides to replace it; the tissue launders its own damage. A cortical neuron has no such option. The cell firing in your prefrontal cortex today is, with few exceptions, the cell that was there when you were three, carrying every error it has ever accumulated. Renewal is itself a repair mechanism, and post-mitotic tissue has opted out of it.
This is also why the model's brain result should feel familiar to anyone following aging neuroscience. The same irreplaceability underlies why structural change in the midlife brain is so consequential, and why damaged nuclear DNA leaking into the cytoplasm sets off the cGAS-STING alarm that feeds chronic inflammation instead of being quietly cleared.
The measured data underneath the model
A model is only as good as the mutation rates fed into it, so it is worth looking at what has actually been counted inside human cells.
The reference dataset for neurons comes from Lodato and colleagues, published in Science in 2018. Using single-cell whole-genome sequencing on 161 individual neurons from 15 neurologically normal people aged four months to 82 years, they found somatic single-nucleotide variants accumulating approximately linearly with age — at roughly 23 variants per year in prefrontal cortex neurons and about 40 per year in the hippocampal dentate gyrus.[2] Neurons were not clean slates at birth either: within a year of birth they already carried an estimated 300–900 such variants. Neurons from individuals with Cockayne syndrome and xeroderma pigmentosum — inherited defects in DNA repair — carried a 2.3-fold and 2.5-fold excess respectively, exactly the direction the mutation hypothesis of aging predicts.[2] The authors coined the term genosenium for this slow, inexorable drift.
The comparative evidence is, if anything, more striking. Cagan and colleagues sequenced 208 intestinal crypts from 56 individuals across 16 mammal species for a 2022 Nature paper. Across species differing about 30-fold in lifespan and 40,000-fold in body mass, the per-year somatic mutation rate varied enormously and correlated inversely with lifespan — yet the mutation burden accumulated by the end of life varied only about threefold.[3] Long-lived species are not spared mutation; they are slower at accruing it, arriving at a broadly similar total burden when they die. That near-constant end-of-life burden is the strongest empirical argument that mutation load does not merely correlate with aging but constrains it.
Where the model is weakest
Four caveats deserve to be stated in the paper's own spirit rather than buried.
1. It is arithmetic, not evidence
Lead author Dmitrii Kriukov has been explicit that the figures are "a mathematical estimate (though careful), not experimental data," and not a verdict of inevitability.[7] Nobody has tested a mutation-only aging scenario, because nobody can create one.
2. The premise is heroic
The ceiling only applies after every other hallmark of aging has been eliminated. That premise silently includes solving cancer, cardiovascular disease, neurodegeneration, immune decline and chronic inflammaging. Nothing in the model tells you how to do any of that; it merely describes what would still be waiting on the other side.
3. Single-nucleotide variants are not the whole damage picture
Somatic mutation is only one accounting of molecular damage. Structural variants, mitochondrial DNA deletions, transposable-element activity and epigenetic drift are not the same currency, and the clonal expansion of mutated cells — the process behind clonal hematopoiesis — can do harm out of proportion to raw mutation counts.
4. Other models put the ceiling lower
The 146–194 range sits above most competing estimates. Pyrkov and colleagues, analysing longitudinal blood-count data and wearable activity records in Nature Communications in 2021, tracked the progressive loss of physiological resilience — how long the body takes to return to baseline after a perturbation — and found recovery time diverging at a critical point of 120–150 years, which they read as an absolute limit.[5] Demographically the picture is more sober still: Olshansky and colleagues, examining the eight longest-lived national populations plus Hong Kong and the United States from 1990 to 2019, found life-expectancy improvements decelerating and concluded that survival to age 100 is unlikely to exceed 15% for women and 5% for men unless biological aging itself is markedly slowed.[6]
What we would take from it
Read as a lifespan forecast, this paper is noise. Read as a research-prioritisation argument, it is genuinely useful — and it points somewhere unfashionable.
If renewable tissue is effectively immune to mutation-driven decline while post-mitotic tissue is not, then the leverage is not in another antioxidant or a better-dosed supplement. It is in repairing, protecting or replacing cells the body has no mechanism for replacing. That reframes which interventions are worth watching: the DNA-repair advantage observed in APOE2 carriers' neurons, the clearance of damaged cells targeted by senolytics, or the protein-quality-control effects of rapamycin. It also offers a clean explanation for why cardiorespiratory fitness keeps outperforming almost everything else in mortality data: the heart is one of the two organs the model flags as a bottleneck, and VO₂ max is the most direct measure we have of how well it still works.
The honest bottom line is that a 156-year ceiling is not a promise. It is a statement about the shape of the problem: even in a world where everything else had been solved, the cells we cannot replace would still be quietly running out of intact genome. That is a harder engineering problem than any supplement aisle suggests — and, unlike the headline number, it is one that current research can actually attack.
References
Primary studies retrieved and verified via PubMed.
- Efimov E, Fedotov V, Malaev L, Khrameeva EE, Kriukov D. Somatic mutations impose an entropic upper bound on human lifespan. npj Aging. 2026. PubMed · DOI
- Lodato MA, Rodin RE, Bohrson CL, et al. Aging and neurodegeneration are associated with increased mutations in single human neurons. Science. 2018;359(6375):555–559. PubMed · DOI
- Cagan A, Baez-Ortega A, Brzozowska N, et al. Somatic mutation rates scale with lifespan across mammals. Nature. 2022;604(7906):517–524. PubMed · DOI
- Dong X, Milholland B, Vijg J. Evidence for a limit to human lifespan. Nature. 2016;538(7624):257–259. PubMed · DOI
- Pyrkov TV, Avchaciov K, Tarkhov AE, et al. Longitudinal analysis of blood markers reveals progressive loss of resilience and predicts human lifespan limit. Nat Commun. 2021;12(1):2765. PubMed · DOI
- Olshansky SJ, Willcox BJ, Demetrius L, Beltrán-Sánchez H. Implausibility of radical life extension in humans in the twenty-first century. Nat Aging. 2024;4(11):1635–1642. PubMed · DOI
- Skolkovo Institute of Science and Technology. Researchers estimate the upper limit of human lifespan determined by somatic mutations. Press release, July 2026. Skoltech
Common questions
What is the maximum human lifespan?
The longest verified human life is 122 years and 164 days (Jeanne Calment, died 1997), and demographic analysis in Nature found that the age at death of the world's oldest person has not increased since the 1990s.[4] The 146–194 year figure from the 2026 npj Aging model is not an observed maximum — it is a theoretical ceiling that would apply only if every hallmark of aging except DNA mutation had already been eliminated.[1]
Why do neurons limit lifespan more than liver cells?
Because they are not replaced. Neurons and cardiomyocytes are post-mitotic, so every mutation a cell acquires stays with it for life — measured at roughly 23 variants per year in prefrontal cortex neurons and about 40 per year in the hippocampus.[2] Proliferating tissue such as liver continuously replaces damaged cells, diluting the load. In the 2026 model this asymmetry was decisive: liver-type tissue stayed functional for thousands of simulated years while post-mitotic tissue capped median lifespan near 156.[1]
Does this study mean people will soon live to 150?
No. The model estimates what would remain possible after every other aging mechanism was solved; it offers no route to solving them. Its own comparison makes the point — mutations cut the theoretical ceiling from 1,759 years to 156, while real median lifespan is around 80, so mutations account for only part of the gap.[1] Independent demographic work found life-expectancy gains decelerating since 1990 and survival to 100 unlikely to exceed 15% for women and 5% for men.[6]