Guide

The Comprehensive Guide to Health and Longevity

What the evidence says about living longer in good health: sleep, exercise, food, checkups, mind and the frontier, with one habit to track in each chapter.

November 13, 2026 · 47 min read


The honest summary of longevity science fits in one sentence. Do not smoke, stay fit and strong, sleep enough at regular hours, eat mostly whole foods, drink little, keep your blood pressure and cholesterol in check, and stay close to people. Everything else is either small, unproven or still in the lab.

This guide goes through each of those areas in turn: what the best studies available in late 2026 found, how sure we can be, and what to do about it. It also weighs four of the loudest voices in the field, Peter Attia's Outlive, Andrew Huberman's Protocols, Bryan Johnson's Blueprint and Ray Kurzweil's predictions, against that evidence. Each chapter ends with a habit worth tracking, because knowing all this changes nothing until it shows up in your week.

One note before we start. This is general information for healthy adults, not medical advice. Studies describe averages across thousands of people, and you are one person with your own history. For anything involving a diagnosis, a medication or a test, talk to your doctor.

The short version

  1. 1Fitness, sleep, food, no smoking and little alcohol do most of the work.
  2. 2A few numbers matter early: blood pressure, cholesterol and blood sugar. Silent, and treatable.
  3. 3No supplement or anti-aging drug has yet been shown to extend the life of healthy people.

The big picture: what a long life is made of

The goal
More healthy years, not only more years
What decides it
Mostly habits kept for decades
Evidence
Strong for the basics

Lifespan and healthspan

Lifespan is how long you live. Healthspan is how long you live well: able to move, think and take care of yourself. The two have drifted apart. Across 183 countries, people lived an average of 9.6 years in poor health in 2019, and 12.4 years in the United States, the widest gap of any country (Garmany and Terzic, JAMA Network Open, 2024).

So the aim of this guide is not to reach 100. It is to push illness and frailty as late as possible, so that the curve of your health stays high and then drops quickly, instead of sagging for twenty years.

Two ways to age

Health
405060708090

Age

  • Fit and active
  • Typical
Illustrative shapes, not data. The goal is the upper curve: health that stays high for longer and declines late.

What actually ends lives

In rich countries most people no longer die of infections or accidents. They die of slow diseases. In Outlive, the physician Peter Attia calls them the Four Horsemen: heart disease and stroke, cancer, neurodegenerative disease such as Alzheimer's, and type 2 diabetes with its related metabolic problems.

All four have something in common. They take decades to develop and give no symptoms for most of that time. That is why the same handful of habits protects against all of them, and why starting at 40 beats starting at 70.

How much is in your hands

A Harvard analysis followed 78,865 women and 44,354 men for up to 34 years (Li, Circulation, 2018). It counted five habits: never smoking, a healthy weight, at least 30 minutes of activity a day, moderate alcohol and a good diet. At age 50, women with all five could expect to live 14.0 years longer than women with none, and men 12.2 years longer. A follow-up found that women with four or five of the habits had 10.7 more years free of diabetes, heart disease and cancer than women with none, and men had 7.6 more (Li, BMJ, 2020).

Years gained at 50, healthy habits or none

Women, total life
14 years
Men, total life
12.2 years
Women, disease-free
10.7 years
Men, disease-free
7.6 years
From Li et al., Circulation 2018 and BMJ 2020. Observational cohorts of US health professionals. Total life compares people with all five low-risk habits to people with none. Disease-free years compare people with four or five habits to people with none.

These are observational numbers, and they compare two extremes. People with five good habits also differ in income, education and luck. But even if the true effect is half that size, nothing in a pharmacy comes close.

What about genes? Estimates vary a lot. A study of 400 million people in family trees put the heritability of lifespan under 10% (Ruby, Genetics, 2018). Classic twin studies say about 25%. A 2026 paper in Science argued for about 50% once deaths from accidents and infections are modelled out (Shenhar). Whichever is right, a large share is not written at birth, and it is the only share you can act on.

How far it can go

Habits buy healthy years, not immortality. Life expectancy gains in the longest-lived countries have slowed since 1990, and one analysis concluded that survival to 100 is unlikely to exceed 15% of women and 5% of men this century without a real breakthrough in the biology of aging (Olshansky, Nature Aging, 2024). The verified record is still Jeanne Calment, who died in 1997 at 122.

Be careful with legends, too. The famous "Blue Zones" of extreme longevity have been challenged by a preprint arguing that reports of extreme old age cluster where birth records are poor (Newman, preprint, updated 2024). The Blue Zones researchers dispute this. Either way, the habits in this guide rest on much larger studies than a few villages.

How to read the rest of this guide

Three questions sort good claims from bad ones:

  • Was it a trial or a cohort? A randomized trial can show cause. An observational cohort shows that two things travel together. Most longevity evidence is the second kind.
  • Was it in people? Many "anti-aging" results come from mice or from cells in a dish.
  • Did it measure life and disease, or a marker? A better blood value or "biological age" score is not the same as a longer life.

Where the evidence is strongest

Not smokingStrongDecades of consistent data. Quitting before 40 restores most of the lost years.
Fitness and regular exerciseStrongHuge cohorts, plus randomized trials including one in cancer survivors.
Treating blood pressure and LDLStrongLarge randomized trials with fewer heart attacks, strokes and deaths.
Enough sleep at regular timesGoodConsistent cohorts with wearables. Few trials with hard outcomes.
Mediterranean-style eatingGoodOne large randomized trial plus many cohorts.
Drinking less alcoholGoodClear for cancer. The old benefit of light drinking has faded.
Close relationshipsGoodA pooled analysis of 148 studies. Cause and effect are hard to separate.
SaunaEarlyMostly one cohort of Finnish men.
Longevity supplementsSpeculativeNo human trial shows a longer life.
Biological age testsSpeculativeUnreliable for one person. No proof that a better score means less disease.
Our reading of the research cited in this guide, as of late 2026. The levels describe how well a claim is supported in humans, not how large the effect is.

Four guides: Attia, Huberman, Johnson and Kurzweil

Peter Attia
Prevent early, train for old age
Andrew Huberman
Time your light, coffee and rest
Bryan Johnson
Measure everything, automate the day
Ray Kurzweil
Stay alive until medicine catches up

Four people shape most conversations about health and longevity today. They disagree on almost everything except the foundation, which is the interesting part.

Peter Attia and Outlive

Outlive: The Science and Art of Longevity (2023, written with Bill Gifford) is the most complete popular book on the subject. Its argument:

  • Medicine 2.0 against Medicine 3.0. Today's medicine is superb at acute problems and waits too long with chronic ones. Attia's "Medicine 3.0" means acting decades before disease shows, weighing the risk of doing nothing, and caring about healthspan as much as lifespan.
  • The Four Horsemen. Heart disease, cancer, neurodegenerative disease and metabolic dysfunction cause most slow deaths. Prevention has to start before any of them is visible.
  • The Marginal Decade. Picture your last ten years and decide what you want to be able to do in them. His "Centenarian Decathlon" is that list: hike a hilly trail, get up off the floor with one arm for support, pick up a child, carry your groceries. Then train backwards from it.
  • Five tools. Exercise, nutrition, sleep, emotional health and what he calls exogenous molecules (drugs and supplements). He calls exercise "by far the most potent longevity drug".

The book's benchmarks are demanding: about three hours of easy cardio a week as a minimum, one hard interval session, a dead hang of two minutes for a man of 40 (90 seconds for a woman), and aerobic fitness in the top few percent for your age.

Where critics push back: Attia goes well beyond medical guidelines. He argues for driving ApoB, the protein on the particles behind artery disease, far below standard targets, for at least 1.6 g of protein per kg of body weight and for colonoscopies from 40. He also uses whole-body MRI scans with patients, while warning that they produce many false alarms. The American College of Radiology says there is not enough evidence to recommend such scans for people without symptoms or risk factors. His fitness claims lean on observational data. And his style of medicine is expensive.

Andrew Huberman and Protocols

Andrew Huberman is a neuroscientist at Stanford School of Medicine and the host of the Huberman Lab podcast. His book Protocols: An Operating Manual for the Human Body (September 2026) turns years of episodes into a reference manual of more than 500 pages, organized into seven areas: sleep, exercise, stress control, nutrition, light, focus and learning, and personal growth.

Where Attia thinks in decades, Huberman thinks in days. In practice a protocol is a specific behavior with a time attached, and much of the book is about timing: when you see light, drink coffee, eat, train and rest. Its best known tools:

  • Morning light. Get outside soon after waking for several minutes, longer when it is overcast, to set the body clock. Then a second dose of daylight late in the afternoon.
  • Caffeine timing. Wait an hour or more after waking for the first coffee, and stop 8 to 12 hours before bed.
  • Non-sleep deep rest. A guided relaxation of 10 to 20 minutes, lying down, to recover during the day.
  • Breathing. Cyclic sighing, a double inhale followed by a long exhale, to bring stress down within minutes.
  • Temperature. Deliberate cold exposure and sauna.
  • Movement after meals. A short walk after eating.

The book leads with behavior and treats supplements as secondary, and most of what it asks for is free.

Where critics push back: reviewers, including a long review in Slate, note that confident instructions often rest on thin evidence. Often that means small or short studies, or a mechanism that makes sense but was never tested as a daily routine. Delayed coffee is the standard example, since no controlled trial has tested it. Much of the research on cold exposure is small: one often cited study compared eight winter swimmers with eight other men. Huberman also has paid advisory and sponsor relationships with companies in this field, which he discloses. And the sheer number of protocols is its own problem: nobody can run dozens of routines at once.

Protocols, sorted by evidence

Daylight early, regular sleepGoodStrong lab data on the body clock, plus large wearable cohorts on light and timing.
A short walk after mealsGoodLight walking breaks lower blood sugar after meals in pooled small trials.
Slow breathing with long exhalesEarlyOne month-long trial of 108 adults, co-run by his lab.
Non-sleep deep restEarlySmall studies of guided relaxation. Harmless to try.
Deliberate cold exposureEarlySmall studies. Nothing on disease or lifespan.
Delaying morning coffeeSpeculativeA plausible mechanism. No controlled trial.
Our reading of the research behind six well known Huberman tools, as of late 2026. The levels describe support in humans for a health benefit.

Bryan Johnson and Blueprint

Bryan Johnson sold his payments company Braintree, which owned Venmo, to PayPal in 2013 for 800 million dollars. In 2021 he launched Project Blueprint: measure every organ he can, and let the data set his routine. He says he spends about 2 million dollars a year on it. A Netflix documentary about him, Don't Die: The Man Who Wants to Live Forever, was released in January 2025.

His day is rigid. He calls sleep his number one priority, eats his last meal many hours before bed, exercises about an hour every day, eats an almost entirely plant-based diet and in 2023 was reported to take 111 pills a day.

The most useful part of Blueprint is what he reports, including what he dropped. After six plasma exchanges, one using his son's plasma, he reported "no benefits detected" in 2023, and later switched to a different procedure. In 2024 he stopped taking rapamycin, citing infections and worse blood sugar and lipids. In 2023 he also had a follistatin gene therapy on the island of Roatan in Honduras, which the US Food and Drug Administration has not approved.

The scientific problems are plain. It is one person with no control group, changing dozens of things at once, scored by "biological age" tests that are not reliable for one person, and he sells the products he tests. Nothing shows that he will outlive anyone.

And yet the basics he lists first are ordinary: sleep, exercise and diet. The expensive part is the unproven part.

Ray Kurzweil and the bridges

Ray Kurzweil, born in 1948, is an inventor and futurist who joined Google in 2012. His father died of heart disease at 58, and he was diagnosed with type 2 diabetes at 35. He wrote about extending life in Fantastic Voyage (2004, with the physician Terry Grossman), and most recently in The Singularity Is Nearer (2024).

His strategy is to cross a series of bridges:

Kurzweil's four bridges

  1. 11. Today's knowledgeDiet, exercise, prevention
  2. 22. Biotech plus AIPredicted for the 2020s
  3. 33. NanotechnologyMedical nanobots, 2030s
  4. 44. Backing up the mindA 2040s technology
The bridges as described in The Singularity Is Nearer (2024). Only the first exists today. The dates on the others are his predictions.

The idea behind it is "longevity escape velocity", popularized by Aubrey de Grey, who set it out in a 2004 paper: the moment when medicine adds more than one year of remaining life expectancy per year that passes. In his 2024 book Kurzweil wrote that the most diligent and informed people would reach it "by around 2030". In interviews in 2026 he says 2032.

For decades he also took enormous numbers of supplements, about 250 pills a day in the mid 2000s and around 100 later.

The pushback: his forecasts about computing have often held up, while his forecasts about biology have run early. No treatment has ever extended the maximum human lifespan. And there is no good evidence that megadoses of supplements beat a placebo.

What to borrow from each

Peter AttiaPhysician, author of Outlive

Prevent the slow diseases early and train for your last decade.

Borrow Exercise first. Know your numbers decades before trouble.

Skip Scans and targets that go beyond the evidence.

Andrew HubermanNeuroscientist, author of Protocols

Small daily behaviors, timed well, steer sleep, energy and stress.

Borrow Morning daylight, regular sleep, a walk after meals.

Skip Treating every protocol as settled science.

Bryan JohnsonFounder of Blueprint

Measure the body and let data, not willpower, run the day.

Borrow Sleep first. Fixed routines. Dropping what does not work.

Skip A hundred pills and unapproved therapies.

Ray KurzweilInventor and futurist

Stay healthy long enough for medicine to catch up.

Borrow Bridge one: the proven basics, done now.

Skip Betting your health on a predicted date.

Our summary. The ideas are theirs, the sorting into borrow and skip is ours.

What they agree on

Strip away the price tags and the predictions, and all four stand on the same floor. Kurzweil's first bridge, Johnson's first priorities, Huberman's first protocols and Attia's first tool are the same short list: sleep, movement, food, and catching disease early.

Bets on Costs Evidence in humans
Attia Prevention and training High, if you buy all the testing Strong for exercise and blood pressure, thinner for the extras
Huberman Daily timing and behavior Low, mostly free Good for light, sleep and movement, thin for many specifics
Johnson Measurement and protocol Extreme One person, no controls
Kurzweil Future technology Low for bridge one None yet for bridges two to four

The rest of this guide is that shared floor, one chapter at a time.

Sleep: regular first, then long enough

Aim for
7 or more hours, most nights
Biggest lever
The same times every day
Evidence
Good, mostly observational

Sleep is where most longevity routines start, for a practical reason: a bad night makes every other habit in this guide harder the next day. You move less, eat worse and have less patience for people.

How much is enough

A dose-response meta-analysis of sleep and mortality (Yin, Journal of the American Heart Association, 2017) found a U shape with the lowest risk at about 7 hours. Each hour below 7 went with about 6% higher mortality, and each hour above 7 with about 13% higher. The long-sleep side probably reflects illness more than harm from sleep itself.

The American Academy of Sleep Medicine and the Sleep Research Society recommend that adults sleep 7 or more hours on a regular basis. That is a floor, not a magic number. If you wake without an alarm and feel alert through the afternoon, you are likely getting enough.

Short sleep in midlife also shows up in brain data. In the Whitehall II study (Sabia, Nature Communications, 2021), which followed 7,959 British civil servants for 25 years, people sleeping 6 hours or less at age 50 had a 22% higher risk of dementia than those sleeping 7, and 37% higher at age 60. This is an association. Poor sleep can also be an early symptom.

Why regular beats long

The most useful sleep finding of recent years is about timing. Researchers tracked 60,977 UK Biobank adults with wrist accelerometers (Windred, Sleep, 2024). Compared with the least regular fifth of sleepers, every more regular group had 20 to 48% lower all-cause mortality, and regularity predicted mortality better than sleep duration did.

It was one week of measurement in an observational cohort, and irregular sleep can be a sign of illness or shift work. Still, the practical reading is cheap to try: keep your wake time within about half an hour, weekends included, and let bedtime follow.

Habit to track

Up at the same time (within 30 min)

A fixed wake time anchors your body clock, and bedtime follows on its own. Regular sleepers had the lowest mortality in wearable data.

Every day

Light, caffeine, alcohol, heat

Light. In 88,905 UK adults who wore light sensors for a week (Windred, PNAS, 2024), people with the brightest nights had 21 to 34% higher mortality, and people with the brightest days had 17 to 34% lower. Laboratory work offers a mechanism: morning light pulls the body clock earlier, evening light pushes it later. Get outside in the first half of the day, and dim the house in the last hours.

This is the first sleep protocol in Huberman's book: daylight outdoors soon after waking, for several minutes on a clear day and longer under cloud, because outdoor light is far brighter than any lamp. One caveat: the cohort studies measured daytime light as a whole, so nobody has shown that the morning is uniquely powerful. It is simply the easiest time to make it a habit.

Habit to track

Daylight outside within an hour of waking

Daytime light sets the body clock, and people with bright days had lower mortality in sensor data. Tie it to your coffee or commute.

Every day

Caffeine. Its half-life averages about 5 hours in healthy adults, with a wide range from 1.5 to 9.5 hours. In a small crossover study (Drake, Journal of Clinical Sleep Medicine, 2013), a 400 mg dose (about four cups of coffee) taken 6 hours before bed still measurably cut sleep, and the 12 participants did not notice the loss. Protocols adds two timing rules. Stopping 8 to 12 hours before bed fits the half-life well. Delaying the first cup for an hour or more after waking is popular, and no controlled trial has tested it.

How much caffeine is still in you

75 mgLeft after 5 hours
38 mgLeft after 10 hours
Pure arithmetic with the average half-life of about 5 hours. Your own half-life may be anywhere from 1.5 to 9.5 hours.

Alcohol. A meta-analysis of controlled studies in healthy adults (Gardiner, Sleep Medicine Reviews, 2025) found that alcohol delays and reduces REM sleep in a dose-dependent way, even at low doses (up to half a gram of alcohol per kg of body weight, which is two to three drinks for most adults). The nightcap that helps you fall asleep costs you later in the night.

Heat. A hot bedroom clearly harms sleep. The often quoted ideal of about 18 C (65 F) is expert convention, not a hard finding, and older adults may sleep best a little warmer. Cool, dark and quiet is the direction.

Screens. The light from a phone delays sleep by about 10 minutes at most (Bauducco, Sleep Medicine Reviews, 2024). The real cost is the hour you stay up scrolling. The fix is a bedtime, not orange glasses. If the phone is the problem, see how to stop doomscrolling.

Sleep folklore against the data

You often hearThe evidence says

Everyone needs 8 hours7 or more, lowest risk near 7

Blue light ruins sleepStaying up late does the damage

A nightcap helps you sleepLess REM, even at low doses

Hours are all that countRegular timing predicts better

Four common beliefs and what the studies cited in this chapter support.

A sample wind-down, 90 minutes

1015203015
  • Last screen, set alarm10 min
  • Tidy, prepare tomorrow15 min
  • Dim lights, wash20 min
  • Read or stretch30 min
  • Lights out15 min
An example to adapt, not a prescription. The point is the same sequence at the same time each night.

When habits are not enough

For chronic insomnia, the American College of Physicians recommends cognitive behavioral therapy for insomnia (CBT-I) as the first treatment for all adults, ahead of medication (Qaseem, Annals of Internal Medicine, 2016). It works by changing the habits and thoughts that keep insomnia going.

Obstructive sleep apnea is common and mostly undiagnosed. One modelled estimate puts mild to severe cases at 936 million adults aged 30 to 69 worldwide (Benjafield, Lancet Respiratory Medicine, 2019). Loud snoring, pauses in breathing that a partner notices and heavy daytime sleepiness are reasons to see a doctor, not to buy a gadget.

Two cautions about sleep culture. The popular story that sleep "washes" the brain comes from mouse studies and is under active dispute (Miao, Nature Neuroscience, 2024). And sleep trackers can backfire: clinicians coined the word orthosomnia for patients who lost sleep chasing a perfect score (Baron, Journal of Clinical Sleep Medicine, 2017). Watch the trend over weeks, not last night's number.

Habit to track

In bed by my set bedtime

Work nights are where sleep gets cut. A bedtime you tick off is easier to defend than a vague plan to sleep more.

Sunday to Thursday

More on this in how to go to bed earlier and our evening routine for better sleep.

Exercise: the strongest lever you have

Aim for
150+ min of cardio, 2 strength days
Biggest lever
Going from nothing to something
Evidence
Strong

If this guide had to shrink to one chapter, it would be this one. Few things you can change have an association with survival as large as physical fitness.

How big the effect is

At the Cleveland Clinic, 122,007 patients took a treadmill test and were followed for a median of 8.4 years (Mandsager, JAMA Network Open, 2018). The least fit group died at about five times the rate of the elite group (hazard ratio 5.04), and there was no upper limit where more fitness stopped helping. In the same statistical model, smoking carried a hazard ratio of 1.41 and diabetes 1.40.

A study of 750,302 US veterans found the same shape: the least fit had about four times the mortality of the extremely fit, and each step up in fitness (one MET) went with 14% lower risk (Kokkinos, Journal of the American College of Cardiology, 2022).

Mortality risk in one model

Low fitness (vs elite)
5 x
Smoking
1.4 x
Diabetes
1.4 x
Coronary artery disease
1.3 x
Hazard ratios for death from Mandsager et al., JAMA Network Open, 2018 (122,007 patients). Low fitness is compared with elite fitness, the others with not having the condition. Observational data.

Read these numbers with care. They are observational, fitness is partly inherited, and illness lowers fitness before it kills. Nobody becomes five times less likely to die by buying running shoes. But the direction is the same in nearly every cohort, and one large randomized trial now points the same way. In the CHALLENGE trial (Courneya, New England Journal of Medicine, 2025), 889 people treated for colon cancer were assigned to a three year structured exercise program or to health education. Estimated survival at eight years was 90.3% in the exercise group against 83.2% in the other, although survival was a secondary outcome of the trial.

The biggest gain is at the bottom

The step from doing nothing to doing a little is worth more than any later step. An analysis of device-measured activity in up to 135,046 adults (Ekelund, The Lancet, 2026) estimated that just 5 more minutes a day of moderate to vigorous movement might prevent 6% of deaths among the least active. It is a model, not a trial, but it matches every dose-response curve in this field: steep at first, then flattening.

The World Health Organization's 2020 guideline is a sensible target: 150 to 300 minutes of moderate aerobic activity a week (or 75 to 150 vigorous), plus muscle strengthening on 2 or more days. Packing most of it into one or two days seems to protect against disease about as well as spreading it out (Kany, Circulation, 2024).

What your daily minutes add up to

140Minutes a week
93 %Percent of the 150 minute target
122Hours in a year
Arithmetic only: minutes a day times 7, compared with the lower WHO target of 150 minutes of moderate activity a week.

Four things to train

In Outlive, Peter Attia splits exercise into four parts. It is a useful map even if you never follow his volumes.

1. Easy cardio (zone 2). Steady effort just below the point where breathing gets hard: you can still talk in full sentences. It builds your aerobic base and you recover from it quickly, so you can do a lot of it. It is not magic, though. A 2025 review titled "Much Ado About Zone 2" (Storoschuk, Sports Medicine) found no good evidence that this intensity is uniquely best for mitochondria, and concluded that harder work matters more when your time is short.

2. Hard cardio (VO2 max). Your peak aerobic capacity drops with age, by 3 to 6% per decade in your twenties and thirties and by more than 20% per decade after 70 (Fleg, Circulation, 2005). Intervals are an efficient way to push it back up. The classic format is four hard efforts of 4 minutes with easy minutes between them, which raised VO2 max more than the same amount of slow running in a Norwegian trial (Helgerud, Medicine and Science in Sports and Exercise, 2007). One honest footnote: in the Generation 100 trial of 1,567 adults in their seventies, five years of intervals did not significantly lower mortality compared with following the national guidelines (Stensvold, BMJ, 2020). The comparison group was also asked to stay active, so the trial compared kinds of exercise, not exercise with none.

3. Strength. We lose muscle slowly from midlife, roughly half a percent a year, and strength two to five times faster than mass (Mitchell, Frontiers in Physiology, 2012). A meta-analysis of 16 cohorts linked muscle strengthening to 10 to 17% lower risk of death, heart disease, cancer and diabetes, with the largest reduction at about 30 to 60 minutes a week (Momma, British Journal of Sports Medicine, 2022). In the PURE study of 139,691 adults, every 5 kg less grip strength went with 16% higher mortality (Leong, The Lancet, 2015). Grip is a marker of whole body strength, not something to train alone.

4. Stability and balance. In the United States, falls are the leading cause of injury death after 65, with about 41,400 deaths in 2023 (CDC). In a Brazilian cohort of 1,702 adults aged 51 to 75, those who could not stand on one leg for 10 seconds were 84% more likely to die over the next 7 years (Araujo, British Journal of Sports Medicine, 2022). Balance is trainable at any age.

A sample training week

5 h
  • Easy cardio150 min
  • Strength90 min
  • Hard intervals30 min
  • Balance and mobility30 min
An example of how the four parts can share about five hours. Beginners should start with far less and build slowly.

Habit to track

Strength training

In cohort studies the largest drop in mortality shows up at about 30 to 60 minutes a week. Two short sessions cover it.

2 times a week

Steps and everyday movement

A 2025 systematic review in The Lancet Public Health (Ding) covered 57 studies and pooled 31 of them. Compared with 2,000 steps a day, 7,000 steps went with 47% lower all-cause mortality and, in the two studies that measured it, 38% lower dementia risk. For several outcomes the curve flattens between 5,000 and 7,000.

The famous 10,000 comes from the name of a Japanese pedometer launched in the mid 1960s, around the Tokyo Olympics: the Manpo-kei. It is a fine goal and it was never a research finding.

Short bursts count too. Among 25,241 UK adults who did no formal exercise, those whose days included about three bursts of vigorous effort lasting a minute or two (stairs, running for a bus) had 38 to 40% lower all-cause and cancer mortality than those with none (Stamatakis, Nature Medicine, 2022). And if you sit all day, activity offsets it: in a pooled analysis of over a million adults, 60 to 75 minutes of moderate activity a day erased the extra risk of sitting more than 8 hours (Ekelund, The Lancet, 2016).

One of the better supported tools in Huberman's Protocols fits here: a short walk after meals. In a meta-analysis of small trials, breaking up sitting with short bouts of light walking lowered blood sugar after meals compared with staying seated (Buffey, Sports Medicine, 2022).

Habit to track

Walk 7,000 steps

Most of the benefit seen in step studies arrives by about 7,000 a day. A phone or watch counts them for you.

Every day

Test yourself

Attia's "Centenarian Decathlon" asks a good question: what do you want to be able to do in your last decade? Carry groceries, get up off the floor, lift a suitcase overhead. Train now for those, with a margin, because capacity only goes down.

Six simple self-checks

0 / 6

Markers used in the studies above or everyday versions of them, not medical tests. Try them near a wall or chair, and skip any that feel unsafe.

Four realistic weeks

10 days in a row, at bestTrained: 15 · Missed: 1 · Rest day: 12
  • Trained
  • Missed
  • Rest day
An example month: three or four sessions a week, planned rest days and one missed session. Consistency over years is what the studies measure.

Habit to track

Cardio that leaves me out of breath

Low fitness predicted death as strongly as smoking or diabetes, or more so, in the large treadmill studies. Easy and hard sessions both count.

3 times a week

To get going, see how to make exercise a habit, how to walk every day and how to start running.

Nutrition: a pattern, not a diet

Aim for
Mostly whole foods, mostly plants
Biggest lever
Less ultra-processed food
Evidence
Good, one large randomized trial

Nutrition is the noisiest part of longevity, and the evidence is weaker than the shouting suggests. Diet trials are short, people misreport what they eat, and no one can be randomized to a way of eating for 40 years. What survives all that is a pattern, not a brand of diet.

The pattern with the best evidence

The strongest single piece of evidence is PREDIMED, a Spanish trial that assigned 7,447 adults at high cardiovascular risk to a Mediterranean diet with extra olive oil, the same diet with nuts, or low-fat advice. After about five years, both Mediterranean groups had roughly 30% fewer heart attacks, strokes and cardiovascular deaths (Estruch, New England Journal of Medicine, 2013, retracted and republished in 2018 after problems with randomization at some sites). The benefit came mainly from fewer strokes, and overall mortality did not differ.

Modelling fills in the long view. Using pooled cohort data, one team estimated that moving from a typical Western diet to an optimal one at age 20 would add 10.7 years for women and 13.0 for men, with most of the gain from more legumes, whole grains and nuts and less red and processed meat (Fadnes, PLOS Medicine, 2022). Starting at 60 still added about 8 years for women and almost 9 for men, and at 80 about 3 for both. A model assumes the associations are causal, so treat the figures as an upper bound.

It is not too late to eat better

Starting at 20
10.7 years
Starting at 60
8 years
Starting at 80
3.4 years
Modelled years of life gained by women who move from a typical Western diet to an optimal one, by starting age (Fadnes et al., PLOS Medicine, 2022). Estimates, not trial results.

A plate that matches the evidence

Most days
  • Vegetables and fruit
  • Legumes
  • Whole grains
  • Nuts
  • Olive oil
Some days
  • Fish
  • Eggs and dairy
  • Poultry
Now and then
  • Red meat
  • Processed meat
  • Sugary drinks
  • Packaged snacks
The food groups that carry the gains and losses in the studies cited in this chapter. A direction, not a meal plan.

Ultra-processed food

The clearest experiment here is small and strict. Twenty adults lived in a research ward for a month and ate an ultra-processed diet for two weeks and an unprocessed one for two weeks, matched for calories on offer, sugar, fat and fibre. On the ultra-processed diet they ate about 500 calories more a day and gained 0.9 kg. On the unprocessed diet they lost 0.9 kg (Hall, Cell Metabolism, 2019).

A 2025 trial in 55 adults compared two diets that both met UK dietary guidelines. People lost about twice as much weight on the minimally processed version (Dicken, Nature Medicine). And an umbrella review covering nearly 10 million people linked higher intake to 21% higher mortality, though most of that evidence is graded low or very low (Lane, BMJ, 2024).

You do not need to fear a slice of bread. The practical point is that food designed to be eaten fast makes you eat more.

Fibre and protein

Fibre. A review of 185 cohort studies and 58 trials for the World Health Organization found 15 to 30% lower mortality, heart disease, stroke, type 2 diabetes and colorectal cancer in people who ate the most fibre, with the largest benefit at 25 to 29 g a day (Reynolds, The Lancet, 2019). Most people eat far less.

Protein. The long-standing official minimum (the RDA) is 0.8 g per kg of body weight a day. The 2025 to 2030 US dietary guidelines now set a higher goal of 1.2 to 1.6 g. Expert groups on aging suggest at least 1.0 to 1.2 g for healthy people over 65, because older muscle responds less to protein. In people who lift weights, extra muscle gain levels off around 1.6 g (Morton, British Journal of Sports Medicine, 2018). Attia recommends 1.6 to 2.2 g, which starts at the top of that range, and no trial has tested high protein against lifespan. People with kidney disease need medical advice before raising protein.

Protein ranges for your body weight

56 g a dayMinimum (RDA)
84 g a dayOlder adults
112 g a dayUpper useful range
Arithmetic only: body weight times 0.8 (the RDA, a minimum), 1.2 (suggested for older adults) and 1.6 (where muscle gain levels off in training studies). Not a prescription.

Fasting and eating windows

Fasting had its moment. The trials since have been sobering. In a year-long trial of 139 adults with obesity, eating only between 8 am and 4 pm added nothing significant to plain calorie restriction (Liu, New England Journal of Medicine, 2022). A 2026 Cochrane review concluded that intermittent fasting may make little to no difference to weight loss compared with standard dietary advice, based on low-certainty evidence. Attia, who once did regular multi-day water fasts, has backed away from them and now worries about the muscle they cost.

Eating less overall does move markers of aging. In CALERIE, healthy adults who cut calories by about 12% for two years slowed one epigenetic measure of aging by 2 to 3% (Waziry, Nature Aging, 2023). That is a marker, not a longer life.

An eating window is a fine tool if it helps you eat less junk at night. It is not a longevity therapy.

What you drink

Coffee looks safe and possibly helpful: in US survey data, people who drank it in the morning had 16% lower mortality than non-drinkers (Wang, European Heart Journal, 2025, observational). Sugary drinks go the other way, with 21% higher mortality at two or more a day compared with less than one a month (Malik, Circulation, 2019). Alcohol gets its own chapter next.

Swaps that follow the evidence

Instead ofTry

Packaged snack at 4 pmA handful of nuts and fruit

White bread or riceA whole grain version

Processed meat in sandwichesBeans, eggs, fish or chicken

Soft drink with lunchWater, coffee or tea

Examples of small changes in the direction of the studies above.

Habit to track

Vegetables or legumes at two meals

Fibre, legumes and whole plants carry most of the benefit in diet studies. Counting meals is easier than counting grams.

Every day

Habit to track

No packaged snacks

Ultra-processed food led people to eat about 500 more calories a day in a controlled trial. Weekdays are where routines live.

Weekdays

If you have a medical condition, are pregnant or have a history of disordered eating, get individual advice before changing how you eat. For the habit side, see healthy eating habits and how to cook at home more.

Alcohol and tobacco: the two big subtractions

Tobacco
None. Quitting pays at any age
Alcohol
Less is better
Evidence
Strong for tobacco, good for alcohol

Most of this guide is about adding things. Two subtractions are worth more than almost all of them.

Smoking

In a study of more than 200,000 US adults, smokers lost at least 10 years of life compared with people who never smoked (Jha, New England Journal of Medicine, 2013). The same study carries the best news in this guide: people who quit between 25 and 34 got about 10 of those years back, and quitting before 40 avoided about 90% of the excess risk. Even quitting at 45 to 54 returned about 6 years.

Years of life regained by quitting

Quit at 25 to 34
10 years
Quit at 35 to 44
9 years
Quit at 45 to 54
6 years
Approximate years regained compared with people who kept smoking, by age at quitting (Jha et al., New England Journal of Medicine, 2013). Observational data from US adults.

If you smoke, nothing else in this guide comes close to stopping. Doctors and pharmacists have treatments that raise the odds, and most people need several attempts.

Alcohol

For years, studies suggested that light drinkers outlive non-drinkers. Much of that turned out to be a flaw in the comparison: the non-drinking group included people who had quit because they were ill. A meta-analysis of 107 cohorts and 4.8 million people that corrected for this found no significant survival benefit from low-volume drinking, and clearly higher mortality from about two drinks a day for women and three for men (Zhao, JAMA Network Open, 2023).

On cancer the picture is firmer. The World Health Organization stated in 2023 that no level of drinking is safe for health, and the US Surgeon General's 2025 advisory called alcohol the third leading preventable cause of cancer in the United States, linked to about 100,000 cases and 20,000 deaths a year.

The debate is not over. A 2024 US National Academies review still found lower overall mortality in moderate drinkers, and the 2025 to 2030 US dietary guidelines, released in January 2026, dropped the numeric daily limits and simply tell people to consume less alcohol. A fair reading: the risk from a few drinks a week is small, the benefit is probably zero, and the dose matters. Alcohol also cuts into the REM sleep from the previous chapter.

What a weekly habit adds up to

364Drinks in a year
73Same as bottles of wine a year
Arithmetic only. A 750 ml bottle of wine at 12% holds about five US standard drinks of 14 g of alcohol each.

Habit to track

Alcohol-free day

Counting dry days is simpler than counting drinks, and it protects your sleep on work nights.

5 days a week

Cutting back is a classic habit problem of cues and replacements. See how to break bad habits.

Know your numbers: the quiet risks you can treat

Check
Blood pressure, cholesterol, blood sugar
Biggest lever
Finding high blood pressure early
Evidence
Strong, from randomized trials

Heart disease and stroke kill more people than anything else, and they are built over decades by forces you cannot feel. This is the part of longevity where modern medicine is strongest, and where Attia's central point, act early, is hardest to argue with.

Blood pressure

High blood pressure is the largest single risk factor for lost healthy years worldwide (Global Burden of Disease 2023). It has no symptoms.

In the SPRINT trial, 9,361 adults at high cardiovascular risk and without diabetes were treated to a systolic pressure under 120 or under 140. The lower target meant 25% fewer major cardiovascular events and 27% fewer deaths, at the cost of more fainting, low blood pressure and kidney problems (New England Journal of Medicine, 2015). The 2025 American guideline sets a goal under 130/80 for adults and encourages measuring at home.

A validated upper arm monitor costs little. Sit quietly for five minutes, measure twice, and bring the readings to your doctor.

Habit to track

Measure my blood pressure

High blood pressure is silent and is the world's largest risk factor. Home readings show your doctor far more than one reading in a clinic.

Once a week

Cholesterol, ApoB and Lp(a)

The particles that carry LDL cholesterol do not just travel with artery disease, they cause it. A consensus review of more than 200 studies and 2 million people concluded that LDL causes artery disease, and in statin trials each 1 mmol/L reduction in LDL cholesterol meant about 22% fewer major vascular events (Ference, European Heart Journal, 2017). Exposure adds up over a lifetime, which is the case for knowing your level in your thirties, not your sixties.

Two terms you will hear:

  • ApoB counts the harmful particles directly. Attia argues for pushing it far below standard targets, as early as possible. The 2026 American guideline brings back LDL goals as the main target and gives ApoB a supporting role. The gap between those two positions is a judgment call for you and your doctor.
  • Lp(a) is an inherited particle that is high in about one person in five. Guidelines now say to measure it once in adulthood. There is no proven treatment yet: in a large trial of people who already had cardiovascular disease, reported in full in November 2026, a drug that lowered Lp(a) by about 78% did not reduce heart attacks or strokes. A high result is still useful, because it is a reason to be stricter about everything else.

Blood sugar and waist

Type 2 diabetes develops over years of rising blood sugar. HbA1c, a simple blood test, shows your average over about three months: under 5.7% is normal, 5.7 to 6.4% is prediabetes, and 6.5% or more is diabetes. By one strict definition, only 6.8% of US adults had optimal cardiometabolic health in 2017 to 2018 (O'Hearn, Journal of the American College of Cardiology, 2022).

A tape measure adds information that weight alone misses. Continuous glucose monitors are popular among healthy people, but reviews have found no evidence that they improve health in people without diabetes.

What to ask about, and when

Every year or two
  • Blood pressure
  • Weight and waist
Every few years
  • Cholesterol panel
  • HbA1c or glucose
Once as an adult
  • Lp(a)
By age and risk
  • Colorectal from 45
  • Breast from 40
  • Cervical 21 to 65
  • Bone density, women 65+
Stay current
  • Vaccines
  • Dental and eye checks
  • Hearing test
A general map for adults based on US recommendations. Ages and intervals differ by country and by personal risk, so use it as a list of questions for your doctor.

Screening: more is not always better

Screening finds cancers early, and the standard programs have earned their place: in the United States, colorectal screening from 45, breast screening every two years from 40 to 74, cervical screening from 21 to 65, and lung scans for heavy smokers.

Beyond those, be careful. Whole-body MRI scans find something in a large share of healthy people, most of it harmless, and the American College of Radiology says there is no evidence that they prolong life. The largest trial of a multi-cancer blood test, in about 142,000 people in England, missed its main goal in 2026. Every test has false alarms, and false alarms have costs.

When habits are not enough

Lifestyle comes first, and for many people it is not sufficient. That is not a failure.

  • Statins. For adults at raised risk who have not yet had a heart attack or stroke, pooled trials show fewer of both and about 8% lower overall mortality (Chou, JAMA, 2022). US advice applies this to ages 40 to 75.
  • Aspirin. No longer advised as a routine step for people without heart disease. US advice is not to start it at 60 or older, and to treat it as an individual choice for some higher-risk people aged 40 to 59, because bleeding offsets much of the benefit.
  • GLP-1 drugs. In the SELECT trial of 17,604 adults with excess weight and existing heart disease, semaglutide cut major cardiovascular events by 20% (Lincoff, New England Journal of Medicine, 2023). A notable share of the weight lost on these drugs is lean tissue, which makes strength training and protein matter more. Two large trials of semaglutide against early Alzheimer's failed in 2025.

All of these are decisions to make with a doctor who knows you.

Questions for your next checkup

0 / 7

Tick the ones you have an answer to. The rest are worth one appointment.

Brain: protection starts in midlife

Preventable share
About 45% of dementia, in theory
Biggest levers
Hearing, cholesterol, learning, contact
Evidence
Good, mostly observational

Dementia is the horseman people fear most, and the one with the fewest treatments. The encouraging news is how much of the risk is shared with everything else in this guide.

Fourteen risks you can change

The Lancet Commission on dementia (Livingston, 2024) estimated that about 45% of cases worldwide are linked to 14 modifiable risk factors spread across life. The estimate assumes each link is causal and that the risk could be removed completely, so it is a ceiling. The list itself is practical.

Largest modifiable dementia risks

Hearing loss
7 %
High LDL cholesterol
7 %
Less education
5 %
Social isolation
5 %
Air pollution
3 %
Share of dementia cases linked to each factor in the 2024 Lancet Commission. The other nine are depression, head injury, inactivity, diabetes, smoking, high blood pressure, vision loss, obesity and excess alcohol.

Three things stand out. Hearing loss shares the top place with high LDL cholesterol, and it is fixable. High LDL cholesterol was added in 2024, which ties brain health to the previous chapter. And most of the factors act in midlife, long before memory slips.

What trials show so far

  • Hearing aids. In the ACHIEVE trial of 977 adults aged 70 to 84, hearing aids did not slow cognitive decline overall across three years, but in the subgroup at higher risk decline was 48% slower (Lin, The Lancet, 2023). If you ask people to repeat themselves, get a hearing test.
  • Lifestyle programs. In the US POINTER trial, 2,111 at-risk adults aged 60 to 79 followed either a structured program of exercise, diet, mental challenge and health monitoring, or a self-guided version. Both groups improved on cognitive tests across two years, and the structured group improved slightly more (JAMA, 2025). The effect was small and real.
  • Movement. Among 78,430 UK adults wearing trackers, walking about 9,800 steps a day was linked to about half the dementia risk, and 3,800 steps to 25% lower risk (del Pozo Cruz, JAMA Neurology, 2022). Early dementia also makes people walk less, so read this as an upper bound.
  • Vaccination. A natural experiment in Wales, where eligibility for the shingles vaccine depended on an exact birth date, found that vaccination cut new dementia diagnoses by a fifth across seven years (Eyting, Nature, 2025). An Australian study pointed the same way. It is not a randomized trial, but it is unusually clean evidence.
  • Drugs. The antibody lecanemab slowed decline in early Alzheimer's by 27% in its main trial, which amounts to a modest difference on the clinical scale, with a real risk of brain swelling. In 2025, semaglutide failed to slow early Alzheimer's in two large trials.

Sleep belongs here too: recall that short sleep at 50 and 60 went with higher dementia risk in the Whitehall study.

Keep the brain busy

Education and mentally demanding work are linked to lower risk across many cohorts. There is little proof that brain training games transfer to daily life. What you can reasonably do is keep learning things that are hard for you: a language, an instrument, a craft, a subject. See how to learn a language daily and how to read every day.

Protect the hardware as well. Wear a helmet on a bike, and protect your ears from loud noise.

Habit to track

20 minutes of learning something hard

Education and mental challenge are linked to lower dementia risk across cohorts. Pick one skill and show up for it.

5 days a week

Mind: stress, mood and purpose

Aim for
Some recovery in every day
Biggest lever
Treating depression early
Evidence
Moderate

Attia ends Outlive with a chapter on his own emotional health. Its point is blunt: a longer life is worth little if you are unhappy in it. This is the least measurable part of longevity and it may hold the rest together.

Mood and the body

Depression is linked to about 50% higher mortality in a meta-analysis of 293 studies (Cuijpers, American Journal of Psychiatry, 2014). Part of that is physical illness causing low mood, but depression also makes every habit in this guide harder to keep.

It is treatable, and exercise is one of the treatments. A network meta-analysis of 218 trials found moderate reductions in depression from walking or jogging, yoga and strength training, though the authors rated their confidence as low (Noetel, BMJ, 2024), and a 2026 Cochrane review rated exercise as similar in effect to psychotherapy. That makes it a strong addition to professional care, not a replacement for it.

If low mood, anxiety or hopelessness last more than two weeks, or you have thoughts of harming yourself, please contact a doctor, a therapist or a local crisis line. That step is as much a health behavior as any workout.

Stress

Short stress is useful. Stress that never lets up is not. In the INTERHEART study across 52 countries, people reporting permanent stress at work or at home had roughly double the odds of a heart attack (Rosengren, The Lancet, 2004). A pooled analysis of 197,473 European workers found 23% more coronary disease with high job strain (Kivimaki, The Lancet, 2012).

You often cannot remove the source. You can build recovery into the day: sleep, movement, time outside, people, and something that slows your breathing.

What helps, modestly

Meditation. A review of 47 trials found moderate evidence that mindfulness programs ease anxiety, depression and pain, and no evidence that they beat other active treatments such as exercise or therapy (Goyal, JAMA Internal Medicine, 2014). Worth trying, not magic. Start with how to start meditating or everyday mindfulness.

Breathing and rest. Two tools from Protocols belong here. The first is cyclic sighing: a double inhale through the nose followed by a long exhale. In a Stanford trial he ran with the psychiatrist David Spiegel, 108 adults spent five minutes a day for a month on either a breathing exercise or mindfulness meditation. Every group felt less anxious, and the breathing groups, cyclic sighing most of all, gained more in daily positive mood than the meditation group (Balban, Cell Reports Medicine, 2023). It was small, short and based on self-reports, and it costs nothing to try. The second is non-sleep deep rest, a guided relaxation of 10 to 20 minutes lying down. The evidence for it consists of small studies, so treat it as a pleasant way to recover, not as a replacement for sleep.

Nature. In a survey of 19,806 English adults, people who spent at least 120 minutes a week in nature were more likely to report good health and wellbeing (White, Scientific Reports, 2019). It is a snapshot, so healthier people may simply get out more. It is also free.

Purpose and outlook. Among 6,985 US adults over 50, those with the lowest sense of purpose had more than twice the mortality of those with the highest (Alimujiang, JAMA Network Open, 2019). The most optimistic people in two long-running cohorts lived 11 to 15% longer than the least (Lee, PNAS, 2019). Illness can drain both, so cause and effect are tangled. But having a reason to get up is clearly not a luxury.

Two hours a week outside

120 min
  • Weekend walk60 min
  • Lunch breaks in a park40 min
  • Evening stroll20 min
An example of how 120 minutes in nature, the threshold in White et al. 2019, can fit into an ordinary week.

Habit to track

10 quiet minutes, no screen

Meditation, slow breathing or a walk without headphones all count. Recovery works when it is scheduled, not when it is left to chance.

Every day

Relationships: the habit nobody tracks

Aim for
Regular contact with a few people
Biggest lever
Being the one who reaches out
Evidence
Good, observational

No longevity protocol sells friendship, which may be why it gets ignored.

What the studies say

A meta-analysis of 148 studies and 308,849 people found that those with stronger social relationships had a 50% greater likelihood of survival across the study periods (Holt-Lunstad, PLoS Medicine, 2010). A second analysis put the extra mortality risk at 29% for social isolation, 26% for loneliness and 32% for living alone (Holt-Lunstad, Perspectives on Psychological Science, 2015).

You may have read that loneliness is as bad as smoking 15 cigarettes a day. The comparison comes from a chart in that 2010 paper and the press release that went with it, and it is stretched: the paper measured social relationships in general, not loneliness. A later meta-analysis of 90 cohorts and about 2.2 million people found a smaller effect for loneliness, about 14% higher mortality, and a similar one for isolation, about 32% (Wang, Nature Human Behaviour, 2023). That is still in the range of well-known risk factors. The World Health Organization estimated in 2025 that one person in six is lonely.

The Harvard Study of Adult Development has followed 724 men since 1938, and later their families. Its director, Robert Waldinger, sums up the findings this way: the people most satisfied in their relationships at 50 were the healthiest at 80.

Extra mortality risk, social factors

Living alone
32 %
Social isolation
29 %
Loneliness
26 %
Higher odds of death in follow-up from Holt-Lunstad et al., 2015, a meta-analysis of 70 studies. Observational: poor health also isolates people.

Why it works, probably

People close to you notice when something is wrong and push you to see a doctor. They get you out of the house. They make stress smaller. Some of the association also runs backwards, since illness isolates people. Both directions argue for the same action.

Treat it like training

Friendships fade through neglect, not conflict. Contact needs a schedule the same way exercise does:

  • A standing weekly call with a parent or an old friend.
  • One shared activity on the calendar: a walk, a club, a team, a choir.
  • A two-line message to someone you have not spoken to in months.

Combine it with other chapters. A weekly walk with a friend covers steps, daylight and connection in one hour. More ideas in how to stay in touch with friends.

Habit to track

Call or see a friend

Strong relationships went with 50% better survival across 148 studies. Contact fades unless someone schedules it.

2 times a week

Everyday protection: skin, teeth, air and heat

Skin
Sunscreen on exposed skin
Teeth
Brush twice, clean between once
Air
No smoke indoors, filter if needed

A few small routines protect you from slow damage. None is dramatic. All are cheap.

Skin

Sunscreen has one of the few randomized trials in everyday prevention. In Nambour, Australia, 1,621 adults were assigned to daily sunscreen or to using it as they pleased. About fifteen years after the trial began, the daily group had half as many melanomas, 11 against 22, a difference just short of statistical significance (Green, Journal of Clinical Oncology, 2011). Among the 903 participants under 55, skin aging across four and a half years was 24% less in the daily group (Hughes, Annals of Internal Medicine, 2013). Shade, clothing and a hat do the same job.

Teeth

Gum disease is tied to heart disease in observational data, and preliminary results presented in 2025 linked regular flossing to fewer strokes. That link is not proven. The sure benefit is enough: keeping your own teeth means eating well and comfortably at 80. See how to floss every day.

Air

Fine particle pollution is the second largest risk factor for lost healthy years worldwide (Global Burden of Disease 2023). The Air Quality Life Index estimates that it cuts global average life expectancy by 1.9 years compared with meeting the WHO guideline. You cannot fix a city alone, but you can avoid smoke indoors, ventilate while cooking, check the air quality before hard outdoor exercise and run a filter where the air is bad.

Heat and cold

Sauna has the most intriguing data. Among 2,315 Finnish men followed for about 20 years, those using a sauna 4 to 7 times a week had 40% lower mortality than those going once a week (Laukkanen, JAMA Internal Medicine, 2015). It is one cohort of men from one region, and frequent sauna users may simply have more time and money. Enjoy it if you have one.

Deliberate cold exposure is one of the signature tools in Protocols. People feel alert afterwards, and that part is real. The health claims rest on little: one widely cited study compared eight winter swimmers with eight controls (Soberg, Cell Reports Medicine, 2021), and no trial has shown that cold plunges lead to less disease or a longer life. One practical note: cold water straight after lifting blunted gains in muscle and strength in a 12 week trial (Roberts, Journal of Physiology, 2015), so keep the two apart. Check with a doctor first if you have a heart condition.

Habit to track

Clean between my teeth

A two minute habit with a sure payoff: your own teeth at 80. It also stacks well onto brushing, which you already do.

Every day

Supplements and drugs: mostly not yet

Extends life
Nothing so far, in humans
Best studied
Vitamin D, omega-3, multivitamins
Evidence
Weak to early

This is the chapter the longevity industry would write differently. As of late 2026, no supplement and no drug taken to slow aging has been shown in a human trial to extend the life of healthy people.

The well-studied supplements

Vitamin D and omega-3. The VITAL trial gave 25,871 US adults vitamin D, omega-3, both or placebo for about five years. Neither reduced the main targets, major cardiovascular events and invasive cancer, although a secondary analysis found fewer heart attacks with omega-3 (Manson, New England Journal of Medicine, 2019). In a later analysis of 777 older Swiss adults, omega-3 slowed three epigenetic clocks by about three to four months across three years (Bischoff-Ferrari, Nature Aging, 2025). That is a small shift in a marker, not fewer illnesses.

Multivitamins. In the COSMOS trial a daily multivitamin did not reduce heart disease or cancer. Sub-studies found slightly better memory scores, and in 2026 a slight slowing on two of five epigenetic clocks. The US Preventive Services Task Force says the evidence is insufficient to recommend multivitamins for preventing heart disease or cancer, and it recommends against beta-carotene and vitamin E.

Creatine. Combined with resistance training, creatine added about 1.4 kg of lean mass in older adults across 22 trials (Chilibeck, 2017). Claims for memory and thinking are not established, and the European Food Safety Authority rejected one in 2024.

The longevity candidates

Rapamycin. It was the first drug to extend life in both male and female mice in the rigorous US mouse testing program. In people, the 48 week PEARL trial of weekly low doses missed its main endpoint in 2025. Bryan Johnson stopped taking it. Attia still rates it as promising but unproven.

Metformin. TAME, the trial designed to test this cheap diabetes drug against aging in about 3,000 older adults, was still not fully funded in 2026 and had not enrolled anyone. Metformin alone did not extend life in the mouse program.

NMN and NR. These raise NAD levels in the blood. A 2025 meta-analysis of eight NMN trials found no significant effect on blood sugar, insulin or lipids, and no trial has looked at disease or death.

Taurine. It extended mouse lifespan by 10 to 12% in 2023. In 2025, US National Institute on Aging researchers reported that taurine does not reliably fall with age in humans, undercutting the original idea. There are no human outcome trials.

Senolytics. Drugs that clear worn-out cells work well in mice. Human trials have been small and mixed, and the best known company in the field shut down in 2025.

Resveratrol. It failed in the same rigorous mouse tests that rapamycin passed.

Longevity molecules, by evidence

Creatine with strength trainingGoodAdds lean mass in older adults. Not tested against lifespan.
Omega-3EarlyNo effect on the main heart endpoint in VITAL. Slightly slower epigenetic clocks in one trial.
Daily multivitaminEarlyNo effect on heart disease or cancer. Small signals on memory.
Vitamin DEarlyNo effect on cancer or heart events in a large general-population trial.
RapamycinSpeculativeExtends life in mice. The main human trial missed its endpoint.
MetforminSpeculativeThe trial that would test it has not enrolled anyone.
NMN and NRSpeculativeRaise NAD in blood. No outcome benefit shown.
Taurine, resveratrolSpeculativeTaurine: mouse data, with setbacks since. Resveratrol: failed the mouse tests.
Our reading as of late 2026, for healthy adults and for the goal of a longer life. A drug can rank low here and still be the right treatment for a disease.

How to think about it

Three rules cover most cases.

First, correct a real deficiency if a blood test shows one. That is ordinary medicine, not longevity hacking.

Second, anything strong enough to slow aging is strong enough to have side effects. Johnson's experience with rapamycin is the cautionary tale. Prescription drugs belong in a conversation with a doctor.

Third, compare every pill with the alternatives in this guide. If a supplement worked as well as being fit, you would not need an advertisement to hear about it. The pill regimens of Johnson and Kurzweil are personal experiments, not a result.

Measuring: what is worth tracking

Worth watching
Fitness, blood pressure, sleep timing
Handle with care
Biological age scores
Rule
Track actions before outcomes

Johnson measures hundreds of markers. You need very few, and the most useful ones are free.

Biological age tests

Epigenetic clocks estimate age from chemical marks on DNA. The best known early one, built by Steve Horvath in 2013, predicts calendar age. Newer ones such as GrimAge and DunedinPACE are tuned to mortality risk or to the pace of aging, and Johnson publishes his DunedinPACE score as a headline number.

They are real research tools and poor personal ones. Across large groups, the newer clocks do predict who gets sick and dies sooner. For one person, the same blood sample analysed twice can come back years apart on some clocks, and nobody has shown that lowering your score lowers your risk of disease or death. A 2026 analysis of 51 intervention studies found that many interventions did not move the clocks at all and that different clocks often disagreed. Use them for curiosity, not for decisions.

Wearables

A watch or ring is good at steps, resting heart rate, trends in estimated fitness and when you sleep. It is weak at sleep stages, and its "readiness" scores are proprietary guesses. Trends across weeks mean something. A single morning's number does not.

Continuous glucose monitors have no shown benefit for people without diabetes. Whole-body scans were covered above.

A sensible personal dashboard

Daily, automatic
  • Steps
  • Sleep and wake times
  • Resting heart rate
Weekly
  • Training sessions
  • Blood pressure
  • Alcohol-free days
Yearly, with a doctor
  • Cholesterol
  • Blood sugar
  • Waist
Skip for now
  • Biological age score
  • Glucose sensor
  • Whole-body MRI
What an ordinary person can track without a clinic. The left groups cost nothing or very little.

Track the action, not the outcome

Here is the idea that links this guide to habit tracking. Outcomes such as blood pressure, fitness or sleep quality move slowly and you cannot act on them directly today. Behaviors are yours to do or not do, today.

From outcome to daily action

Outcome you hope forAction you can tick

Lower blood pressureWalked 30 minutes

Higher aerobic fitnessDid cardio, 3 times this week

Better sleepGot up at my set time

Stronger at 70Strength session done

Each slow outcome on the left is driven by an action on the right that you can tick off today.

Check outcomes a few times a year to see whether the actions are working. Track the actions every day. More in habits to track and systems vs goals.

The frontier: what may come, and when

Proven in humans
No anti-aging treatment so far
Newest step
Cell reprogramming, first human trial
Evidence
Speculative

Kurzweil's bet is that bridge two is being built right now. Here is an honest status report.

The theory

Biologists describe aging through "hallmarks": shared processes of decline such as DNA damage, worn-out cells that refuse to die, failing mitochondria and chronic inflammation. A landmark paper listed nine in 2013 and its 2023 update listed twelve (Lopez-Otin, Cell). The frontier is the attempt to treat those processes directly, instead of one disease at a time.

Where things stand in late 2026

Five frontier bets and their status

1Cell reprogrammingFirst dosing announced in June 2026, in a safety trial for eye disease
2Plasma exchangeLowered biological age scores in a small trial in 2025
3Dog longevity drugsA daily pill is partway through US approval, tested in 1,300 dogs
4AI drug designAn AI-discovered drug showed early promise in a small trial in 2025
5Prizes and public moneyA 101 million dollar XPRIZE named its finalists in August 2026
Status as of late 2026. None has shown a longer life in people.

Reprogramming. Resetting the chemical marks of old cells with a subset of the Yamanaka factors rejuvenates tissue in mice. In January 2026 the US regulator cleared what the company calls the first human trial of this approach, a gene therapy from Life Biosciences injected into the eye for two diseases of the optic nerve, and in June the company announced that the first participant had been dosed. It is a small safety study. It is also a real first.

Plasma exchange. A small sham-controlled trial reported that plasma exchange combined with antibody infusions lowered biological age scores by about 2.6 years, against about 1.3 years for plasma exchange alone (Buck Institute, Aging Cell, 2025). It was funded by a company in the field and measured with clocks. Johnson's own experiment with plasma from his son found no benefit.

Dogs first. The company Loyal has passed two of three steps toward conditional US approval of a daily pill meant to extend healthy life in older dogs, with a trial in about 1,300 animals. As of late 2026 no drug is approved to extend lifespan in any species.

AI. The 2024 Nobel Prize in Chemistry went partly to the creators of AlphaFold, which predicts protein structures. In 2025 a drug whose target and molecule were both found with AI showed promising results in an early 71 patient trial for lung fibrosis (Nature Medicine). Kurzweil's claim that simulations will replace clinical trials remains a prediction.

Money. The 101 million dollar XPRIZE Healthspan named its finalists in August 2026 and will award its grand prize in 2030. A US government agency has committed up to 144 million dollars to finding early markers of aging and running trials of treatments that preserve function with age.

The sober view

Against all this stands the demographic record. Life expectancy gains are slowing, not accelerating, and as of late 2026 no human trial has shown that a treatment aimed at aging itself extends lifespan. Every positive human result so far is a marker or a small study. Longevity escape velocity is a hope with a date attached.

The good news is that you do not have to choose a side. If Kurzweil is right, the winning move is to reach the 2030s and 2040s in the best health you can. If the skeptics are right, the winning move is identical. Either way, it is bridge one.

The plan: turn this guide into three habits

Start with
Three habits, not thirty
Make them
Small, scheduled and visible
Review
Once a week, kindly

Nobody fails at longevity for lack of information. They fail in week three, when the plan was too big. The research on habits is as useful here as the research on health.

How habits actually form

In the best known study, habits took about 66 days to become automatic for the typical person, with a range from 18 to 254 days, and missing a single day did not derail the process (Lally, European Journal of Social Psychology, 2009). Three practical rules follow:

  • Start smaller than feels useful. Ten minutes of walking that happens beats an hour of training that does not. See the two minute rule.
  • Tie it to a time and place. "After I make coffee, I walk around the block" works better than "I will walk more". These if-then plans are called implementation intentions, and they pair well with habit stacking.
  • Never miss twice. One missed day is noise. Plan for it, and read what to do when you break a streak.

A sensible order

Do not start everything at once. Sleep and walking come first because they make everything else easier.

Build the base in four stages

  1. 1Sleep and stepsFixed wake time, daily walk
  2. 2Add strengthTwo short sessions a week
  3. 3Numbers and foodCheckup, more plants, less alcohol
  4. 4Intensity and peopleOne hard session, regular contact
An example sequence across about six months. Move on when the current stage feels automatic, not when the calendar says so.

Priorities by decade

20s and 30s
  • Never start smoking
  • Build fitness
  • Regular sleep
  • Sunscreen
40s and 50s
  • Know your numbers
  • Protect muscle
  • Watch alcohol
  • Screenings
60s and beyond
  • Strength and balance
  • Protein
  • Hearing
  • Stay connected
Everything matters at every age. This is where extra attention pays most.

Pick your three

Here are the habits suggested in this guide. Choose up to three to start. If you already do some of them, pick from the others.

Pick your first three

Sleep
Move
Eat and drink
Check
Mind and people
Protect

0 / 3Tap the habits you want to start with.

Your picks are saved in this browser only. Come back and swap them when the first ones feel automatic.

What three months can look like

Twelve weeks, imperfect and working

10 days in a row, at bestDone: 65 · Missed: 8 · Planned skip: 11
  • Done
  • Missed
  • Planned skip
An example of one habit across 12 weeks: mostly done, a few planned rest days, a few misses, never two misses in a row.

Perfection is not the goal. The large studies in this guide followed people for years, and what they measured was a usual way of living, not a flawless month.

Review once a week

Take ten minutes on Sunday. What happened on the days that worked? What got in the way on the days that did not? Adjust the plan, not your opinion of yourself. A weekly review keeps a small system alive, and how to build a habit covers the full method.

This is the reason we built Moto the way we did. Strength training twice a week, a bedtime on work nights and a daily walk are three different rhythms, so Moto lets each habit have its own schedule, and a planned skip for illness or travel does not break a streak. It is free for three daily habits, which happens to be the right number to start with, and Pro adds the other schedules.

FAQ

What is the single most important habit for longevity?

If you smoke, quitting is first: it returns up to ten years of life. If you do not smoke, regular exercise has the strongest link with a long and healthy life, and the biggest gain comes from moving from inactive to a little active. Regular sleep is a close third because it makes the other habits easier to keep.

How many years can a healthy lifestyle add?

In a large Harvard analysis, people with five healthy habits at age 50 were projected to live 12 to 14 years longer than people with none, and people with four or five habits had about 8 to 11 more years free of major disease. These are observational figures comparing two extremes, so the real effect for one person is probably smaller and is uncertain. It is still far larger than anything shown for a supplement or drug.

Is Outlive by Peter Attia worth reading?

Yes, as a framework. Its core ideas are well supported: prevent slow diseases decades early, treat exercise as the main tool and train for the abilities you want late in life. Some of its specific targets, such as very low ApoB, high protein and early or extra screening, go beyond current guidelines and should be discussed with your own doctor.

Is Protocols by Andrew Huberman worth reading?

As a reference, yes, with a filter. Its free behavioral tools, such as morning daylight, regular sleep, exercise and walking after meals, are sound and agree with larger studies. Many of its more specific instructions rest on small or short studies, so treat them as experiments to try, and pick a few instead of all of them.

Does Bryan Johnson's Blueprint work?

Nobody knows, including him. It is an experiment on one person with no control group, many changes at once and results measured mainly with biological age tests that are unreliable for one person. The parts with real evidence behind them are the ordinary ones: consistent sleep, daily exercise and a diet built on plants.

What is longevity escape velocity?

It is the hypothetical point at which medicine adds more than one year to remaining life expectancy for every year that passes. Aubrey de Grey set out the idea in 2004, and Ray Kurzweil predicts it for the early 2030s. As of late 2026 no treatment aimed at aging itself has been shown to extend human lifespan, so it remains a prediction.

Which supplements are proven to extend life?

None. Large trials of vitamin D and omega-3 did not reduce their main targets, major heart events and cancer, in the general population. For multivitamins, one trial in male doctors found slightly fewer cancers and a later trial did not, and none showed fewer deaths. Compounds such as NMN, taurine and resveratrol have no human data on disease or death. Correcting a deficiency found by a blood test is sensible, and that is a conversation for your doctor.

Are biological age tests accurate?

They are useful in research on large groups and unreliable for one person. The same sample can give results years apart, and no test has been shown to predict whether a change in your score changes your risk of disease. Blood pressure, cholesterol, blood sugar and fitness tell you more for less money.

Is it too late to start at 60 or 70?

No. Modelling suggests that eating better from 60 could still add years, quitting smoking at around 50 returns about six years, and strength and balance respond to training at any age. In older adults the first goals are usually strength, balance, protein, hearing and staying socially connected.

Keep Reading