Exercise

Why exercise matters#

Physical activity lowers the risk of nearly every major disease of aging Strong. In a 2023 meta-analysis of 94 cohorts with over 30 million people, reaching the equivalent of 150 minutes a week of moderate activity was associated with 31% lower all-cause mortality, 29% lower cardiovascular mortality, and 15% lower cancer mortality. These are observational data, backed by trials showing exercise improves blood pressure and insulin sensitivity.

The pattern holds disease by disease. In 59,325 UK adults wearing activity trackers, those doing more than about 68 minutes a day of moderate-to-vigorous activity had 74% lower risk of type 2 diabetes than the least active. Across 58 studies, active people had 20% lower risk of dementia, even over 20 or more years of follow-up. In 1.44 million adults, high leisure-time activity was linked to lower risk of 13 cancers, including 26% lower lung and 16% lower colon cancer.

Fitness and strength as vital signs#

Cardiorespiratory fitness, usually measured as VO2 max (the most oxygen your body can use per minute), is one of the strongest predictors of lifespan. In the Cleveland Clinic study, the least fit quarter for age and sex had an adjusted hazard ratio of 5.04 compared with the top 2.3%, larger than the risk from smoking (1.41) or diabetes (1.40). In 750,302 US veterans, the least fit had four times the mortality of the extremely fit, across ages, races, and both sexes, with no sign of harm at the top. The American Heart Association has argued that fitness should be measured as a clinical vital sign.

Strength tells a similar story. Muscle-strengthening activity is associated with 10–17% lower mortality, with the largest benefit at roughly 30–60 minutes a week. In 139,691 adults across 17 countries, every 5 kg of lower grip strength predicted 16% higher mortality, a better predictor than systolic blood pressure. Grip partly marks overall health, so whole-body strength is the real goal.

Both capacities fade by default. Peak oxygen uptake falls about 3–6% per decade in your 20s and 30s and more than 20% per decade in your 70s, and muscle mass falls about 3–8% per decade after 30. Physician Peter Attia's "centenarian decathlon" frames the response: list what you want to be able to do in your last decade (carry groceries upstairs, get up off the floor) and train well above that level now.

The four capacities#

How much exercise you need#

The optimal dose is well above the official minimum. In 661,137 adults, meeting the minimum was linked to 31% lower mortality, and three to five times the minimum to 39% lower, with no sign of harm even at ten times. In 116,221 US health professionals followed for 30 years, the benefit had nearly leveled off at 300–600 minutes a week of moderate activity or 150–300 of vigorous Strong. Both are observational, but the dose-response pattern is consistent across cohorts. Moderate intensity means breathing harder but still able to talk (brisk walking); vigorous means only a few words at a time (running, hard cycling).

Attia updated his advice in 2026: with about six hours for cardio, spend roughly four in zone 2 and two in harder work; with only two hours, skip dedicated zone 2 and do two interval sessions. Strength targets follow the 2026 ACSM position stand. The exact split is expert practice Weak; the total dose is the better-supported part.

Guideline minimum: the WHO 2020 guidelines and the US Physical Activity Guidelines set 150–300 minutes of moderate or 75–150 of vigorous activity a week, plus strength work on 2 or more days. Both say that going beyond 300 minutes adds benefit.

Timing and dose are flexible#

You do not have to spread activity evenly. In 89,573 UK adults wearing accelerometers, "weekend warriors" who got at least 150 minutes in one or two days had similarly lower risk across more than 200 diseases as people who spread it out, and in 350,978 US adults the two patterns had similar mortality.

Short bursts count too. Among 25,241 UK adults who did no formal exercise, three bouts a day of 1–2 minutes of vigorous incidental activity (fast stairs, hurrying uphill, carrying heavy shopping) were associated with 38–40% lower all-cause mortality and 48–49% lower cardiovascular mortality.

Steps add to structured training rather than replace it. In 15 pooled cohorts, mortality kept falling until about 6,000–8,000 steps a day for adults 60 and older and 8,000–10,000 for younger adults. A 2025 meta-analysis found 7,000 steps a day, compared with 2,000, was linked to 47% lower all-cause mortality, 38% lower dementia risk, and 28% fewer falls.

Sitting less, and standing desks#

In a 2024 study of 83,013 adults wearing devices, sitting more than 10 hours a day was linked to higher cardiovascular risk, while standing more than 2 hours a day was linked to an 11% higher risk of circulatory problems such as varicose veins for each extra 30 minutes.

Exercising safely#

Who should see a doctor first#

Most people can start moderate exercise without a checkup. The American College of Sports Medicine's screening model no longer screens on risk factors alone, which sent too many healthy people to doctors. It asks three questions: are you already active (at least 30 minutes of planned moderate exercise three days a week for 3 months), do you have known heart, metabolic (diabetes), or kidney disease, and do you have symptoms?

The real cardiac risk#

Hard exertion briefly raises cardiac risk, but regular exercise lowers it overall. In a classic study of cardiac arrest, the risk during vigorous exercise was 56 times the risk at other times in sedentary men but 5 times in the most active, whose overall risk was 40% of the sedentary men's. For heart attacks, the relative risk after heavy exertion was 107 in people exercising less than weekly and 2.4 in those exercising five or more times a week. The absolute risk is small: among 21,481 male physicians, about one sudden death per 1.5 million episodes of vigorous exertion. Build intensity over weeks.

Endurance exercise and atrial fibrillation#

Atrial fibrillation (AF) is an irregular heart rhythm that raises stroke risk. The American Heart Association describes a U-shaped or reverse-J-shaped relationship between exercise dose and several cardiac changes at the extremes, including AF, accelerated coronary calcium, and heart-muscle scarring. A meta-analysis found athletes had about 2.5 times the odds of AF of non-athletes, with higher risk under age 55.

Keep this in proportion: fitness shows no upper limit of mortality benefit, and the AF signal comes from people training many hours a week for years. The mortality benefit of extra activity levels off around three to five times the guideline, so volume beyond the optimal targets in this chapter is for performance, not longevity. If you train that much and you notice palpitations or an irregular pulse, get an ECG. AF itself is covered in Heart and blood vessels.

Common injuries and how to avoid them#

Most injuries come from doing too much too soon. In 5,205 runners tracked with GPS watches, a single run 10–30% longer than the longest in the previous 30 days was linked to a 64% higher injury rate, and one more than twice as long to more than double the rate, while week-to-week mileage changes showed no clear link. Strength training is the best-supported prevention: across 25 trials with 26,610 people, it cut sports injuries to less than a third, while stretching had no effect.

Sharp pain, pain that worsens as you continue, or pain that lingers into the next day means back off, and see a clinician or physical therapist if it persists.

Aerobic base training#

The two thresholds, simply#

As you speed up, your body passes two landmarks. At the first (the aerobic threshold, also called LT1 or VT1), blood lactate begins to rise above resting levels, around 2 mmol/L, breathing deepens, and slow-twitch fibers are burning fat about as fast as they can. At the second, lactate climbs steeply (often around 4 mmol/L) and carbohydrate takes over as fuel. "Zone 2" means steady work just below the first threshold.

In exercise scientist Stephen Seiler's widely used scale, a lactate of about 2 mmol/L falls in a band of roughly 75–85% of true maximum heart rate, and elite endurance athletes do about 80% of sessions below the first threshold. Those are averages. Lactate values shift with diet, and a watch's "zone 2" is usually a percentage of an estimated maximum heart rate that may not match your physiology. Because lactate suppresses fat release and fat burning, err slightly easy rather than slightly hard.

Finding your aerobic threshold#

Triangulate with two or three of these. If you do only one, do the heart rate drift test.

How to train your base#

Optimal: three or four sessions of 45–60 minutes a week, about 3–4 hours in total; two sessions maintain a base you already have. This is coaching practice Weak, and the adaptations take months. Do not let zone 2 crowd out intensity: a 2025 review found the evidence does not support zone 2 as the optimal intensity for mitochondria or fitness in the general public, especially when training time is limited. With only 2 or 3 hours a week, keep the strength sessions and make most of the cardio hard.

Choose what you will do for years. Track progress as power or pace at the same heart rate, checked every 4–8 weeks.

Polarized or pyramidal#

"Polarized" means mostly easy, a little very hard, and little in between; "pyramidal" adds a moderate middle tier. A small 2019 meta-analysis in trained athletes favored polarized training, but a 2025 analysis of 13 studies with 348 athletes found no overall difference, with recreational athletes possibly doing better on pyramidal plans Moderate. Keep most sessions easy, make a few genuinely hard, and do not live in the middle.

VO2 max training#

VO2 max, in milliliters of oxygen per kilogram of body weight per minute (ml/kg/min), is the ceiling of your aerobic engine. In US reference data it falls about 13.5% per decade, but it responds to intervals within weeks.

The Norwegian 4x4#

In a 2007 trial of 40 moderately trained men, four 4-minute intervals at 90–95% of maximum heart rate, each followed by 3 minutes of easy recovery, three times a week for 8 weeks, raised VO2 max by 7.2%, significantly more than long slow running or threshold running of equal total work Moderate. In practice: warm up 10–15 minutes, then run, ride, row, or climb at the hardest pace you can hold evenly across all four repeats. You should be breathing hard and able to say only a word or two by the end of each interval. Start with two or three repeats if you are new to this.

How often and when#

Optimal: one interval session a week once you have a base, and two during a block when you are trying to raise VO2 max. This is expert practice Weak, not a tested longevity dose. More is not better: beyond two or three hard sessions a week, fatigue usually builds faster than fitness, which is why elite endurance athletes keep about 80% of sessions easy. Do intervals fresh. Combining cardio and strength training does not reduce muscle growth or maximal strength, but it can blunt explosive power when done in the same session rather than at least 3 hours apart.

Intervals suit older adults too. In the Generation 100 trial, 1,567 Norwegians aged 70–77 did twice-weekly intervals, moderate training, or the national guidelines for 5 years; mortality did not differ significantly, though the interval group trended lower. With heart disease, get clearance first.

VO2 max benchmarks#

Reference values from the US FRIEND registry for treadmill-measured VO2 max in healthy adults (Kaminsky 2022), in ml/kg/min:

Optimal goal: at or above the 90th percentile for your age and sex (the right-hand columns), and the top 2.5% ("elite") if you can get there Moderate. Mortality was lowest in the elite group, and even moving from the top quarter ("high") to elite was linked to 23% lower mortality, including in people 70 and older. Attia's case for aiming this high is to plan backward from what you want to do late in life: VO2 max falls about 13.5% per decade, so a high number at 50 is what keeps hiking and stairs easy at 85. Guideline minimum: no guideline sets a fitness target; the American Heart Association asks clinicians to assess fitness routinely.

Watch estimates suit trends better than absolute values: in a small 2024 study, Apple Watch estimates were off by about 16% on average. A lab exercise test is the reference (see how often to test).

Strength training#

Why strength matters#

Muscle absorbs much of the glucose from a meal, lifts you out of a chair, and catches you when you trip. It is never too late: in a 1990 study of 10 frail nursing-home residents averaging 90 years old, 8 weeks of heavy training raised strength by 174% and tandem walking speed by 48%. Sarcopenia, the age-related loss of muscle, is covered in Muscle, bone, and mobility.

Principles#

The 2026 American College of Sports Medicine position stand pooled 137 systematic reviews covering more than 30,000 people. Resistance training improved strength, muscle size, power, walking speed, and balance Strong, and few details mattered much. Strength gains were largest with heavy loads (at least 80% of your one-repetition maximum), full range of motion, two to three sets, lifts done early in the session, and at least two sessions a week. Muscle growth was larger with at least 10 sets per muscle group per week. Training to complete failure, machines versus free weights, and periodization schemes made no consistent difference. Lighter loads build muscle about as well when sets are taken close to failure, but heavier loads build more maximal strength.

Optimal: three full-body sessions a week, the easiest way to reach 10 or more hard sets per muscle group with time to recover; two is the guideline minimum. Cohort studies found the mortality benefit of strength work peaked around 30–60 minutes a week, but they measured light, self-reported activity; the extra sessions build the strength, muscle, and bone that keep you independent later.

The five movement patterns#

Add single-leg work (split squats, step-ups, single-leg deadlifts) for balance and to even out side-to-side differences.

How to progress#

Use double progression: add reps until you reach the top of your range on every set, then add load. Machines, dumbbells, and bodyweight are fine at first; a few sessions with a qualified coach are worth it before loading a barbell heavily. When progress stalls, take a lighter week or change the rep range. Extra protein adds modestly to training gains, with no further benefit beyond about 1.6 g per kg of body weight a day (see Nutrition), and creatine adds to gains in muscle and strength, including in older adults (see Supplements).

Older adults#

Strength matters more with age because the reserve is smaller. Power training, moderate loads lifted fast, improved physical function in the ACSM review: stand up quickly from a chair, step up briskly, push a sled. Start with machines or a supervised class if balance is shaky.

Women and bone#

Heavy lifting is one of the few things that builds bone after menopause: in the LIFTMOR trial, supervised heavy lifting and impact work raised spine bone density 2.9% while controls lost 1.2%. Details, screening, and drugs are in Muscle, bone, and mobility.

Training on GLP-1 drugs#

GLP-1 drugs such as semaglutide and tirzepatide produce large weight loss, part of it lean mass: across trials, lean mass made up anywhere from 15% or less to 40–60% of weight lost, though lean mass includes organs and water. A 2025 advisory from four obesity and nutrition societies recommends assessing strength and body composition at the start and preserving muscle and bone with resistance training and adequate protein. In a trial of 195 adults after diet-induced weight loss, exercise plus liraglutide produced about twice the body-fat reduction of either alone, and only the combination improved fitness and insulin sensitivity. On these drugs, lift two or three times a week, prioritize protein, and track strength, not just weight; the drugs are covered in Metabolic health.

Stability and mobility#

Stability is control of your body through a range of motion, on one leg, under load, and when something unexpected happens. It fades quietly and is the capacity most tied to falls.

Two tests to try#

One-leg stand. Stand on one leg, the free foot resting against the other calf, arms at your sides, for 10 seconds. In 1,702 adults aged 51–75, one in five could not, and they had 84% higher mortality over about 7 years after adjusting for age, sex, weight, and disease.

Sit-rise test. Sit down on the floor and stand back up using as little support as possible. Each half is scored out of 5, minus one point for each hand or knee you use for support. In 2,002 adults aged 51–80, the lowest scorers had more than five times the mortality of those scoring 8–10. These tests predict risk; it is not proven that improving your score changes it, but the training behind it is worthwhile.

Warm up dynamically, stretch later#

Static stretching right before training slightly reduces strength and power, mainly with long holds: a 4.6% drop with holds of 60 seconds or more versus 1.1% with shorter holds. Dynamic stretching produced small improvements, and static stretching had no clear effect on injury. Warm up by moving through range (leg swings, hip circles, walking lunges, light sets of your first lift) and stretch statically afterward if you want more flexibility.

Balance training#

Balance improves with specific practice. Optimal: 5–10 minutes most days, starting in midlife rather than after a fall; it fits easily into a warm-up. This is expert practice Weak, since the fall-prevention trials were done in older adults. Progress from feet together, to heel-to-toe, to one leg, then add eyes closed, a softer surface, head turns, or reaching. Single-leg exercises and loaded carries count; so does standing on one leg while brushing your teeth. Fall prevention is covered in Muscle, bone, and mobility.

Yoga, tai chi, and Pilates#

• Tai chi Strong: in 670 adults 70 and older at high risk of falling, a therapeutic tai chi program reduced falls by 58% compared with stretching and by 31% compared with conventional exercise.
• Yoga Moderate: recommended by the American College of Physicians, along with tai chi and exercise, as a first-line treatment for chronic low back pain.
• Pilates Weak: popular for trunk control, with less high-quality outcome research.

Recovery, heat, and cold#

Sleep comes first#

Adaptation happens between sessions. One night without sleep cut muscle protein synthesis by 18% and lowered testosterone by 24% in a small trial of 13 young adults; see Sleep. Evening exercise generally does not harm sleep, though vigorous sessions ending within an hour of bedtime may.

Deloads and soreness#

A deload is a planned stretch of lighter training designed to reduce fatigue and promote recovery. Many coaches schedule a lighter week every month or two or when performance stalls; this is expert practice, not trial evidence. Soreness after unfamiliar exercise is normal, fades as you adapt, and is a poor gauge of whether a session worked.

Sauna#

The best data come from the Kuopio Ischaemic Heart Disease study (KIHD) of 2,315 Finnish men followed for about 21 years. Compared with once-a-week users, men using a traditional sauna four to seven times a week had 63% lower risk of sudden cardiac death, 50% lower cardiovascular mortality, and 40% lower all-cause mortality, with sessions over 19 minutes doing best; the average sauna was about 174°F (79°C). The same cohort showed 66% lower dementia risk with frequent use.

These are observational data from Finnish men, and frequent users may differ in other ways. Small trials support plausible mechanisms: in an 8-week trial of 47 adults with cardiovascular risk factors, adding 15 minutes of sauna after exercise lowered systolic blood pressure by 8 mm Hg more than exercise alone. Call it a promising, low-risk habit Moderate, not a proven longevity treatment. Infrared cabins run cooler and are less studied.

Cold exposure#

Evidence for cold plunges is thin, and one effect works against lifting. In a meta-analysis of eight studies, cold-water immersion after resistance training probably blunted muscle growth slightly, and in one 12-week trial it reduced gains in strength and muscle mass. If muscle is a goal, plunge on non-lifting days; the minimum safe gap has not been tested.

A 2025 meta-analysis of 11 trials found cold immersion reduced reported stress 12 hours later and improved sleep quality in some studies, with no effect on mood Weak. The popular "11 minutes a week" comes from a small study of experienced winter-swimming men; it describes what they did, not a tested dose.

Putting it together#

These sample weeks cover all four capacities at three levels, with roughly 80% of cardio easy. The first builds toward the guideline minimum, the second covers every capacity, and the third hits the optimal targets above. Swap days and exercises freely. The author's own program is in My protocol.

Getting started: about 2–3 hours a week#

Solid: about 4–5 hours a week#

Optimal: about 6–8 hours a week#

Stay at a level for 2–3 months before moving up, and drop back during busy periods rather than stopping.

Timing and fasted training#

The best time is the one you will stick to. In 29,836 UK adults with obesity, those most active in the evening had the lowest mortality (hazard ratio 0.39, versus 0.67 for morning), but this is observational. Training before breakfast burns more fat during the session but makes no difference to body composition over time. Easy sessions are fine fasted; hard ones usually go better fed. If you take insulin or another drug that can cause low blood sugar, ask your doctor before training fasted.

Equipment and tracking#

Adjustable dumbbells or kettlebells, resistance bands, and a pull-up bar cover most needs; a gym adds heavier loads. A chest-strap heart rate monitor helps with intervals. Track weekly minutes by intensity, your main lifts, and the benchmarks below on the schedule in how often to test. Devices are compared in Wearables and home monitoring.

Numbers worth knowing#

How often to test#

Test often enough to see whether your training is working. Testing is cheap and low risk, and this schedule is expert practice Weak; no guideline recommends routine fitness testing in healthy adults.

If you have heart disease or symptoms, get clearance before any maximal test (see who should see a doctor first).

Action plan#