What is Longevity?

Longevity went from a technical term to an everyday word in a few years. This guide covers what the word means, where it comes from, what the research currently supports and where the market runs ahead of the evidence. Every central number below carries a source, and every model estimate is labelled as one.

Longevity means long life. In current usage it carries more than that. The word now stands for a research field and a set of lifestyle decisions. Both are built around healthspan: the years someone lives without serious chronic disease or a lasting loss of independence. A short definition lives in the longevity glossary.

Outside health the word just means long duration: the longevity of a career, a marriage or a washing machine. Close synonyms are long life, long-livedness and durability. The spelling trips people up too: it is longevity, not longetivity or longivity.

The word is Latin in origin: longus means long, aevum means lifetime. Late Latin longaevitas, from longaevus meaning long-lived, became longevity in English. The pronunciation is [lɒnˈdʒɛvᵻti], stressed on the second syllable, with a soft j sound as in journalist rather than a hard g.

For some years now longevity has carried an extra layer of meaning: healthy years rather than years alone, measurable markers, prevention rather than repair. That is the shift the plain phrase long life does not capture, which is why the technical word keeps showing up in ordinary writing.

Two neighbouring terms show up in the same conversations. Geroscience describes the biology of ageing at the cellular level. Preventive medicine describes the clinical side, meaning early detection and risk factors. Longevity is used colloquially as an umbrella over both, which is why the word is useful in a headline and imprecise in a study.

Two terms carry the whole debate. Lifespan is how many years someone lives. Healthspan is how many of those years pass without serious chronic disease or lasting limitation. Life expectancy is the third term and a statistical one: the average number of years people of a given age can expect to live.

The global picture is well quantified. An analysis of 183 WHO member states put the global gap between life expectancy and healthy life expectancy at 9.6 years in 2019, up from 8.5 years in 2000. The United States recorded the largest gap of any member state at 12.4 years, almost three years above the global figure ^[2]. Those are the years people spend on average with cardiovascular disease, diabetes, dementia or reduced mobility. Women's gap averaged 2.4 years wider than men's ^[2].

Where	Healthspan-lifespan gap, 2019
Worldwide	9.6 years
United States	12.4 years
Australia	12.1 years
New Zealand	11.8 years
United Kingdom	11.3 years
Norway	11.2 years

The five largest gaps among the 183 states, next to the global figure ^[2].

The upper end is narrower than the trend suggests. The highest fully documented lifespan is 122 years, reached by Jeanne Calment, who died in 1997 ^[17]. A 2016 analysis in Nature argues that the maximum reported age at death has plateaued at around 115 years since the 1990s. On that reading, 122 years is a statistical outlier ^[8]. The paper is contested, and several technical comments in Nature dispute the method. A study of the longest-lived populations reaches a sober conclusion: the probability of reaching 100 is unlikely to exceed 15 percent for women and 5 percent for men, unless biological ageing itself can be slowed ^[7].

Longevity escape velocity names the hope that medicine will one day add more than a year of remaining life expectancy for every year that passes. The data point the other way for now: gains in the longest-lived countries have slowed since 1990 ^[7].

So the interesting target is the gap, not the record. In the literature this is called compression of morbidity: pushing the sick years into a shorter window at the end rather than adding years at the top.

Read enough longevity writing and the same three claims keep surfacing, however different the packaging.

Healthspan before lifespan. The stated goal is not the maximum number of years but the maximum number of functional years. The term has a known weakness. Matt Kaeberlein argued in 2018 that there is still no single agreed way to measure healthspan, which makes studies hard to compare ^[1]. Careful writing therefore states which definition it is using.

Genes and lifestyle both count. Heritability is the share of the differences between people that genes explain. For lifespan, the answer depends heavily on method:

• About a quarter. The classic Danish study of 2,872 twin pairs put it at 0.26 in men and 0.23 in women ^[3].
• Well below 10 percent. A far larger family-tree (pedigree) analysis in 2018 accounted for people choosing partners similar to themselves, and the estimate fell well below 10 percent ^[4].
• Near 50 percent. A 2026 paper in Science stripped out outside causes of death such as accidents and infection. For this intrinsic lifespan, heritability lands near 50 percent ^[5]. Even at that upper estimate, half of the variation is left for behaviour, environment and chance.

The range is wide, but the practical conclusion is stable: a large share of the difference comes from behaviour and environment. Two US cohorts with more than 120,000 participants show how large that share can be. The study used a life-table projection. Women combining five low-risk factors had a life expectancy at age 50 that was 14.0 years higher than women with none of them, and men 12.2 years higher ^[13]. The five factors were never smoking, a healthy body weight, regular activity, moderate alcohol intake and a high-quality diet. Moderate alcohol counted as low risk in that 2018 analysis. The German Nutrition Society concluded in 2024 that no amount of alcohol is safe and advises avoiding it ^[19].

Evidence over hype. The scientific frame is the twelve Hallmarks of Aging, updated in Cell in 2023 ^[6]. They describe linked processes, from DNA damage and shortening telomeres (the protective caps on chromosome ends) to chronic inflammation. Along the way come cellular senescence, where cells stop dividing but do not die, and faltering mitochondria, the cell's power plants. Demography supplies the counterweight. Olshansky and colleagues show that gains in life expectancy have decelerated in the longest-lived countries since 1990, making radical life extension in this century implausible ^[7]. Both belong together: a clear research programme without a promise of rescue.

We group the evidence into six pillars: movement, nutrition, sleep, social connection, stress management and prevention. Different sources count anywhere from three to nine, from a lean three-way split to the Power 9 of the Blue Zones literature. The number is an editorial choice, not a constant. We use six because each of these six behavioural fields has its own large cohort study or meta-analysis behind it. The order follows the strength of the data, not a priority for daily life.

What appears in none of the lists is a supplement: the pillars describe behaviour, not products. Each one has its own body of numbers, from cardiorespiratory fitness through sleep duration to social relationships. The 6 pillars in detail takes each one apart, with the studies, the effect sizes and the limits of the data behind them.

Chronological age counts birthdays. Biological age tries to estimate how far ageing has actually progressed in the body. The best known method is the epigenetic clock. Steve Horvath showed in 2013 that age can be estimated with surprising accuracy from DNA methylation patterns (chemical tags on the DNA that shift with age) across many tissues ^[14]. Many more clocks have followed. They respond differently to health trajectories and are not interchangeable.

In practice that means a single number is an estimate, not a diagnosis. The methods are covered in the biological age guide. If you want to try one, we run a free age test and a photo-based face age estimate. Both are educational, both return an estimate and neither replaces a medical examination.

Blue Zones are five regions with unusually high numbers of very old residents: Okinawa in Japan, the Barbagia region of Sardinia, Ikaria in Greece, the Nicoya peninsula in Costa Rica and Loma Linda in California. The shared traits drawn from them sound familiar: constant everyday movement, a largely plant-based diet, close social ties, a reason to get up.

The underlying data has been under scrutiny for several years. The demographer Saul Justin Newman showed that regions reporting exceptional numbers of very old people frequently coincide with incomplete birth registration, high poverty and short average life expectancy. His proposed explanation: in the United States, missing vital registration (official birth and death records) predicts where these records turn up. Where states introduced birth certificates, the number of records fell by 69 to 82 percent. In Italy, England and France, regional poverty, low incomes and shorter life expectancy predict them. On top of that, reported birthdates cluster on days divisible by five, a pattern the paper links to clerical errors and pension fraud ^[15]. The work is a preprint (not yet peer reviewed) and has not gone unchallenged, but it moved the debate.

What survives is less dramatic and still useful. The lifestyle patterns described in the Blue Zones broadly match what large cohort studies find independently of them ^[13]. It is the record ages that deserve caution, not the patterns. The full case is in the Blue Zones guide.

The word is heavily marketed, so it is worth drawing three boundaries.

Not a supplement programme. The most visible part of the longevity market is capsules. For most of these products there is no human lifespan data. There is cell work, animal work and short studies on surrogate markers (lab values such as blood sugar that stand in for real outcomes). A review in Nature Reviews Drug Discovery sets out the state of the search ^[16]. Candidates such as rapamycin, metformin and NAD precursors, building blocks the body turns into the coenzyme NAD+, are scientifically interesting. Yet none of them is supported by outcome trials in healthy people, meaning trials that measure disease or death rather than a lab value ^[16]. What is known about individual compounds sits in NMN and rapamycin. The term longevity shots is sometimes used for GLP-1 drugs such as semaglutide. These drugs are approved for conditions such as diabetes and obesity, not for ageing. What the data show is in GLP-1 and longevity.

Not anti-aging cosmetics. Anti-aging in retail usually means appearance: skin, wrinkles, hair. Longevity means function: cardiovascular system, metabolism, muscle, cognition. The terms overlap in marketing, not in research.

Not the Bryan Johnson protocol. The US entrepreneur has documented an extensive self-experiment since 2021 and did a great deal to popularise the topic. A single case without a control group is not a study design that generalises. As a media phenomenon it is relevant, as evidence it is not.

How to spot unsupported claims is collected in the guide to longevity pseudoscience.

Longevity medicine is not a protected term and not a recognised medical specialty. In practice the phrase describes an offering: clinics and practices that bundle extensive diagnostics, lab panels, imaging and lifestyle coaching, usually paid out of pocket. Courses and certificates in longevity medicine exist, but they are private qualifications, not recognised specialist training.

There is no defined treatment called a longevity therapy. What gets sold under that label splits into three groups. First, established prevention that ordinary healthcare already provides: blood pressure, lipids, glucose, vaccination status, cancer screening. Second, diagnostics with no demonstrated benefit in people without symptoms. Third, experimental approaches, from infusions to off-label prescribing (drugs used outside their approved purpose), whose benefit-risk balance in healthy people is unresolved ^[16].

The evidence-backed core is unglamorous. Expert consensus treats LDL cholesterol as a settled cause of atherosclerotic disease (plaque build-up in the artery walls). That verdict rests on genetic, epidemiological and randomised evidence ^[12]. The link between fitness and mortality is also well documented ^[9]. Both are standard medicine rather than longevity medicine.

This section describes how the market is structured and recommends no specific service. What a clinic check-up typically costs is broken down in longevity clinic costs. Which offerings are typically self-funded is covered in self-pay longevity, and the line to non-medical practitioners in practitioners and longevity.

Longevity has a price problem. The visible part of the field is tests, capsules and subscriptions. The evidenced part is mostly free. Three routes that cost nothing:

Meet people. Cities across our network run regular meetups where researchers, clinicians and curious people show up in the same room. Entry is free, and the list of cities sits under chapters. Given the strength of the social-connection data, this is the pillar with the best ratio of effort to evidence ^[11].

Run the numbers. Our tools are free and need no account, from energy requirements through a VO2max estimate to grip strength. They replace no diagnostic work, but they put your own numbers in context.

Use the check-ups you already have. In Germany statutory health insurance covers a general health check every three years from age 35, and once between 18 and 35. It includes blood pressure, lipids and fasting glucose ^[18]. Many other systems fund something comparable, so it is worth checking what yours covers. That already covers part of the sixth pillar, and anything beyond it is a conversation with a clinician.

If you take one thing from this page: the difference between people in the large cohort studies was never a product. It was a handful of habits held for decades ^[13].