Insights / Human & Science

How to live longer: what survives the evidence (and what's folklore)

Longevity has become an industry with a supplement aisle, a conference circuit, and a billionaire wing. Underneath the noise, the actual evidence is almost embarrassingly stable: genes account for maybe a fifth of your lifespan, behavior runs most of the rest, and the behaviors that matter fit on an index card. Here's the card — ranked — plus an honest look at what the blue-zone stories got wrong.

By Seçil Sayhan11 min readJune 2026
The short version
  • Genes set roughly 20–25% of your lifespan — likely less. The rest is mostly behavior, environment, and luck. That ratio is the best news in this entire field.
  • The ranked list: don't smoke, build cardiorespiratory fitness, keep muscle, stay connected to people, sleep on a schedule, eat a mostly-plants pattern without chronic excess, go easy on alcohol.
  • Fitness is the strongest biomarker we can measure. Low VO2max carries mortality risk in the same league as smoking — and it's trainable at any age.
  • Loneliness is a biological variable, not a mood. Its mortality risk rivals smoking in meta-analyses.
  • The supplement aisle has no entry on this list. Nothing in a capsule currently has human evidence in the same league as any item above it.

The 20% myth-buster: genes vs. behavior

Start with the number that should reorganize how you think about your own old age. Large twin studies — the classic Danish cohort work — put the heritability of human lifespan at roughly 20–25%. And when researchers later analyzed family trees covering hundreds of millions of people, even that figure looked inflated: spouses' lifespans correlated almost as much as siblings', which means a chunk of what looked like genetics was actually shared kitchens, shared habits, shared streets. True heritability may sit well under 20%.

Read that against the fatalism most people carry — my grandfather died at 60, so my number's drawn. The data says otherwise: for the overwhelming majority of us, the majority of the outcome is behavioral. Your grandfather's heart attack at 60 is information about risk, not a reservation.

One honest asterisk: extreme longevity — the 100+ club — does appear meaningfully more genetic. Getting from 95 to 110 is mostly inheritance. But getting to a strong, lucid 90? That project belongs mostly to you, which is precisely why it's worth knowing what actually funds it.

The ranked list: what actually moves lifespan

1. Don't smoke — still, boringly, first

Roughly a decade of life separates lifelong smokers from never-smokers, and quitting pays at every age — quit by 40 and you recover almost all of it. Nothing else on this list, and nothing in any supplement aisle on earth, moves the needle this far. If this one's handled, move on.

2. Cardiorespiratory fitness — the strongest signal we can measure

Here's the finding that should be on billboards. In a JAMA-published cohort of 122,000 patients given treadmill tests, low cardiorespiratory fitness carried a mortality risk comparable to smoking — and the benefit of being fitter kept climbing with no upper plateau. The elite-fit had dramatically lower mortality than the merely active, who in turn beat the sedentary by a wide margin.

VO2max sounds like athlete vocabulary; it's actually a measure of how much functional engine you'll still have at 80. And it's trainable at any age: the dose that captures most of the benefit is unglamorous — regular brisk-to-hard aerobic work most days, with some genuinely breathless intervals weekly. (If you're starting from exhausted, fix the energy system first.)

3. Muscle — the retirement account nobody opens

From your thirties onward you lose muscle by default, and the loss accelerates. Grip strength predicts mortality across dozens of cohorts; leg strength predicts whether a fall at 78 is an anecdote or the beginning of the end. The mechanism isn't exotic: muscle is glucose storage, metabolic reserve, and the physical buffer between an aging body and the ground. Two resistance sessions a week is the minimum effective dose, and starting at 60 still works — the trainability never expires.

4. People — the variable that embarrasses the biohackers

Julianne Holt-Lunstad's meta-analyses (300,000+ participants) found that weak social connection carries mortality risk comparable to smoking up to 15 cigarettes a day — heavier than obesity, heavier than inactivity. The Harvard adult development study, eight decades in, keeps returning the same headline: relationship quality at 50 predicts health at 80 better than cholesterol does. I wrote about the well-being side of this here — for longevity purposes, just note that your calendar's social column is a clinical document.

5. Sleep, on a schedule

Short sleep and — underrated — irregular sleep both track with cardiovascular disease, metabolic dysfunction, and dementia risk. The full mechanics are in the sleep guide; the longevity summary is one line: same wake time, seven days, for decades.

6. A mostly-plants pattern, without chronic excess

Strip the diet wars down to what survives meta-analysis and you get something nobody can sell a book about: mostly plants, adequate protein (more than you think after 60), minimal ultra-processed food, and not chronically overeating. The Okinawan habit of hara hachi bu — stopping at about 80% full — is the cultural version of the only dietary intervention that reliably extends lifespan in animal models: eating somewhat less, sustainably. Pattern beats protocol. Adherence beats both.

7. Alcohol — the demoted guest

The old "red wine is cardioprotective" curve has not aged well; better-controlled analyses keep shrinking the safe dose toward zero. The honest current position: less is better, none is fine, and the Mediterranean glass was always carrying the credit earned by the vegetables and the friends at the table.

Nothing on this list is secret. That's the finding. Longevity isn't hidden knowledge — it's unhidden knowledge that's hard to do for forty years.

The blue zones: what holds, what doesn't

The blue-zone story — five regions where people supposedly live extraordinarily long, studied for their shared habits — built half the modern longevity industry. As a behavioral scientist I owe you both halves of the truth about it.

What doesn't hold: the demography. Researcher Saul Newman's audits found that claimed extreme ages cluster suspiciously in places with poor historical birth records and pension incentives — and when Costa Rica and others re-checked records, significant chunks of reported extreme longevity evaporated. Some of the world's "supercentenarian hotspots" look less like fountains of youth and more like filing errors with good marketing.

What holds anyway: the behaviors. Movement woven into daily life rather than outsourced to a gym hour. Tight social fabric. Mostly-plant eating. A reason to get up — the ikigai variable. Every one of these is independently supported by research that has nothing to do with the zones. The postcard may be retouched; the habits on it are real.

The supplement question, answered honestly

The pattern repeats every few years, so learn the pattern rather than the molecule. Resveratrol: spectacular headlines, failed replications, a discredited lab. NAD-boosters (NMN, NR): genuinely raise blood NAD; have shown no lifespan or major healthspan outcome in humans. Metformin and rapamycin: legitimate, interesting research programs — and unproven in healthy people, with real trade-offs (rapamycin is an immunosuppressant; metformin may blunt exercise adaptations, which would be a darkly funny way to lose the thing that actually works).

The tell, always: the people selling cellular immortality are quiet about exercise, because exercise is free and outperforms their product in every dataset we have. When a longevity intervention beats brisk walking in a human outcomes trial, you'll hear about it from the medical journals before the podcast ads. Until then, the index card stands.

The reframe that changes everything

Stop optimizing for more years and start optimizing for the slope of your decline. The real prize isn't 95 instead of 88 — it's carrying your own luggage, your own name, and your own decisions through your eighties. Fitness, muscle, people, sleep: every item on the list buys slope before it buys length.

Lifespan was never the real goal

The clinical term is healthspan — the years you're functional, lucid, autonomous — and the gap between it and lifespan currently runs about a decade. Ten years, on average, of being alive but not well. Everything ranked above compresses that gap; almost nothing in the longevity marketing complex does.

And the deeper pattern, the one this whole site keeps returning to: a longer life is a behavior-change problem wearing a biology costume. Knowing the list has never been the bottleneck — you could have written most of it before reading. The bottleneck is running it for decades inside a real, busy, stressful life. Which is a systems problem. That problem has a science too.

The index card

LeverMinimum effective doseWhat it buys
No smokingZero, or quit now — payoff at any age~A decade
Cardio fitnessBrisk work most days + breathless intervals weeklyLargest measurable mortality lever
Strength2 resistance sessions/week, foreverThe slope: function, autonomy, fall-proofing
PeopleReal contact, weekly, protected like a meetingRisk reduction rivaling smoking cessation
SleepSame wake time × 7 days, ~7–8h opportunityCardiovascular, metabolic, cognitive protection
FoodMostly plants, enough protein, ~80% fullMetabolic headroom without dieting theater
AlcoholLess; none is fineHonest accounting, better sleep

The list is free. The system that makes you run it is the product.

The Longevity Protocol turns this evidence into a daily operating system — energy, sleep, food, movement — with 3 months of Marsa Coach included.

See the Longevity Protocol →

Frequently asked questions

What actually increases lifespan the most?

Ranked: not smoking, cardiorespiratory fitness, maintained muscle strength, social connection, consistent sleep, a mostly-plants pattern without chronic excess, minimal alcohol. No supplement has human evidence in the same league.

How much of lifespan is genetic?

Twin studies say ~20–25%, and family-tree analyses suggest even less. Behavior and environment run most of the outcome — though extreme longevity (100+) is more strongly inherited. A healthy 90 is mostly yours to build.

Are the blue zones real?

The behaviors hold — daily movement, social fabric, plants, purpose — because they're supported by independent research. The demography is shakier: audits found extreme-age claims clustering where birth records are poor and pensions reward exaggeration. Trust the habits, not the mythology.

Do longevity supplements like resveratrol or NMN work?

Not on current human evidence. Resveratrol failed replication; NAD-boosters raise blood markers without proven outcomes; metformin and rapamycin are unproven in healthy people with real trade-offs. Exercise outperforms all of them in every dataset we have.

About the author

Seçil Sayhan is a behavioral scientist and the founder of MARSA.AI. Trained on both sides of her field — a BA in Business Management, an MSc in Clinical Health Psychology & Wellbeing, a diploma in neuroplasticity, advanced training in Lifestyle Medicine from Harvard University, and an ICF coaching credential — she has spent the past decade helping 7,000+ people across 12 countries rewire the systems running their lives. That decade produced the conviction MARSA is built on: behavior is one science — whether it moves a person, a market, or a machine. Her work draws on the clinical literature throughout: see the full bibliography.