- American men live to 76.5 on average; women to 81.4. Most of that 4.9-year gap is behaviour and biology you can act on, not fate
- The base of the pyramid is five things: not smoking, cardiorespiratory fitness, strength, sleep, and not being isolated. Each carries mortality effects measured in hundreds of thousands of people. Low fitness alone carried five times the mortality of elite fitness in a 122,007-patient cohort — a bigger effect than smoking in the same data
- Men at 50 with five low-risk habits could expect 37.6 more years; with none, 25.5. Twelve years, from things that cost nothing
- The middle of the pyramid is your waist, your blood pressure, your lipids, what you drink — and for men specifically, alcohol accounts for 6.8% of all male deaths, three times the female figure
- Screening has a men's section: a coronary calcium score, a colonoscopy, and a PSA decision that is genuinely a trade-off, not a box to tick
- The tip — testosterone for normal-T men, NAD, peptides, biological-age tests — is what gets sold hardest. None of it has a mortality trial behind it. TRAVERSE showed TRT is safe for men who need it; it showed nothing about men who don't
Every January the same magazine cover runs: the ten biohacks that will add a decade. The order is always wrong. The peptide stack goes above the treadmill, the NAD infusion above sleep, the $400 biological-age test above the free blood pressure cuff at the pharmacy. So here is the same list with the order fixed — ranked, level by level, by how much each thing actually moves mortality in the studies that measured it, and with the study named every time. It's built for men because the numbers for men are different, and mostly worse. (The women's pyramid is a different shape.)
Start with the gap itself. In the most recent US mortality data, life expectancy at birth was 76.5 years for men and 81.4 for women — a difference of 4.9 years. Some of that is biology. Most of it is the leading causes of death on that same report — heart disease, cancer, unintentional injury, and, tenth on the list, suicide — arriving earlier in men, and the behaviours that bring them. Which means most of it is on the pyramid.
Width is the size and certainty of the mortality evidence, not the size of the marketing budget. The two are close to inversely related.
The Five Things That Decide Decades
Evidence: Strong — large cohorts, replicated, consistent directionThe foundation is five habits, and the reason they're the foundation is that each has been measured against death itself, in cohorts large enough that the answer isn't in doubt. Everything higher on the pyramid is measured against a proxy — a biomarker, a scan, a test score. These are measured against the thing that matters.
The cleanest single summary comes from Li and colleagues in Circulation (2018), who followed 44,354 men in the Health Professionals Follow-up Study and 78,865 women in the Nurses' Health Study for up to 34 years, logging 42,167 deaths. They scored five low-risk factors — never smoking, a BMI of 18.5 to 24.9, at least 30 minutes a day of moderate-to-vigorous activity, moderate alcohol, and a diet-quality score in the top 40%. A 50-year-old man with none of the five could expect to live another 25.5 years. With all five: 37.6 years. Twelve years, and the hazard ratio for all-cause mortality between the two groups was 0.26. Cardiovascular mortality: 0.18.
Those five overlap with, but aren't quite, the five I'd put at the base. Here is the base, with the number that earns each one its place.
Don't smoke — and it's not too late
The largest modifiable cause of death, and the one with the most precise numbers. Jha and colleagues in the New England Journal of Medicine (2013) followed 88,496 men and 113,752 women: smokers lost more than ten years of life, with a mortality hazard ratio of 2.8 in men. The part men need to hear is the second half. Quitting between 25 and 34 regained about ten of those years — essentially all of them. Quitting between 35 and 44 regained about nine. Between 45 and 54, about six. The window doesn't close; it narrows.
Cardiorespiratory fitness — the biggest lever most men never measure
In 2018 the Cleveland Clinic published treadmill data on 122,007 patients, 59% of them men, followed for a median of 8.4 years with 13,637 deaths (Mandsager, JAMA Network Open). The lowest-fitness group had a mortality hazard ratio of 5.04 against the fittest — five times the risk. Set beside the same cohort's other risk factors, low fitness outranked them all: smoking carried a hazard ratio of 1.41, diabetes 1.40, coronary artery disease 1.29. Being below average in fitness, versus above, was worth 1.41 on its own. And the benefit didn't plateau: the "elite" group still beat the "high" group, hazard ratio 0.77. There was no level of fitness at which more stopped helping.
This is the lever that most separates the men who age well from the men who don't, and it's the one the magazine cover puts lowest. The Norwegian 4x4 is the best-studied way to raise it; Zone 2 is the base underneath.
Strength — measured in your handshake
Two cohorts, half a million people each. In the PURE study across 17 countries (Leong, Lancet 2015; 139,691 people), every 5 kg less grip strength meant a 16% higher risk of death from any cause, 17% higher cardiovascular death — and grip predicted death better than systolic blood pressure did. In UK Biobank (Celis-Morales, BMJ 2018; 502,293 people, 13,322 deaths), the figure for men specifically was 1.16 per 5 kg for all-cause mortality and 1.22 for cardiovascular mortality, with the weakness threshold set at under 26 kg. Grip is a proxy for total muscle, and total muscle is what men lose from their forties onward unless they lift. Two strength sessions a week is the dose; this piece has the test and the fix.
Sleep — seven to eight, and not much more
The pooled evidence is a U. Cappuccio's meta-analysis in Sleep (2010) — 16 studies, 27 cohorts, 1,382,999 people, 112,566 deaths — found short sleep carried a relative risk of 1.12 and long sleep 1.30. The long-sleep number is mostly reverse causation (sick people sleep more), which is why the message isn't "sleep as much as possible." It's that chronically short sleep is a measurable mortality risk, and men are the ones who wear it as a badge. The site's sleep system is the practical version.
Not being alone — the lever men drop first
Holt-Lunstad's meta-analysis in Perspectives on Psychological Science (2015) pooled studies from 1980 to 2014 and found social isolation raised the odds of death by 29%, loneliness by 26%, and living alone by 32% — effects in the same range as the physical risk factors above, and larger in people under 65. This is the base-level lever men are worst at: the friendships that thin out through the forties, the reluctance to see a doctor, the tenth leading cause of death on the CDC's list. It isn't soft. It's a mortality risk factor with an odds ratio, and it belongs in the foundation.
The Numbers a Tape Measure and a Cuff Can Read
Evidence: Strong — large meta-analyses; the numbers are cheap to getOne level up are the measurements. They're a level up rather than at the base because most of them are downstream of the base — fitness, strength and not smoking move all of them — but they earn their own tier because they tell you whether the base is working, and because two of them are men's problems specifically.
Your waist, not your weight. Men store fat viscerally — around the organs — and visceral fat is the metabolically active kind. Jayedi's meta-analysis in the BMJ (2020) pooled 72 cohorts and 2,528,297 people: every 10 cm of waist circumference carried an 11% higher all-cause mortality, every 0.1 of waist-to-hip ratio 20%, every 0.1 of waist-to-height ratio 24% — and the associations held after adjusting for BMI. In other words, two men with the same weight and height, one carrying it at the waist, are not at the same risk. Larger hips and thighs, in the same analysis, went the other way (hazard ratios 0.90 and 0.82), which is the muscle showing up again. This is the plan for the waist.
Blood pressure, lipids, glucose. The three numbers behind the first, fourth and seventh leading causes of death. None of them needs a subscription. Blood pressure is a cuff at any pharmacy; the lipid panel — and ideally ApoB, which counts the particles rather than the cholesterol inside them — and a fasting glucose or HbA1c are one draw a year. Men are worse than women at getting the draw done, which is the whole reason it's on the list.
The Pattern, and the Men's Problem
Evidence: Strong for the pattern; Strong and unwelcome for alcoholDiet sits one level above the foundation for a reason that surprises people: the cohort effect of the overall eating pattern is real and large — diet quality is one of Li's five factors — but it's hard to separate from the base, because people who eat well also tend to move, sleep and not smoke. The pattern that keeps showing up is the one every Blue Zone eats: mostly plants, legumes daily, fish more than red meat, very little of anything that comes in a wrapper. A 2026 meta-analysis in BMJ Nutrition, Prevention & Health — 10 prospective cohorts, 86,098 cases of hypertension across 309,853 people — put the legume effect at a relative risk of 0.84 for high versus low intake, roughly linear up to about 170 g a day. The site's recipe section is that pattern, dish by dish.
Protein deserves its own line for men over 40, because the muscle on the base level has to be built from something; the protein guide has the numbers.
Then there's alcohol, which is where the men's version of this article stops being the general version.
The obvious objection is the J-curve — the long-standing finding that light drinkers outlive abstainers. Zhao's 2023 meta-analysis in JAMA Network Open went looking for it in 107 cohorts, 4,838,825 people and 425,564 deaths, with the "abstainer" group cleaned of former drinkers who'd quit for health reasons. It wasn't there. Occasional drinkers had a relative risk of 0.96 against lifetime non-drinkers, low-volume drinkers 0.93 — neither significant. At 45 to 64 grams a day (roughly three to four drinks), the risk was 1.19; at 65 grams and up, 1.35. The honest read for a man: the protective glass was a statistical artefact of who gets counted as a non-drinker, and from about three drinks a day the risk is significant and climbs. Li's "moderate alcohol" factor is best read as "not heavy," not as a prescription.
The Tests Worth Doing — and the One That's a Judgement Call
Evidence: Moderate — each test is a real trade-off with numbers on both sidesScreening is the level where the pyramid stops being about prevention and starts being about catching what the base didn't stop. Three tests belong on a man's list. One of them is contested enough that the honest answer is a conversation.
Coronary artery calcium score — once, in your forties or fifties
A CT scan, no contrast, that counts the calcified plaque in your coronary arteries. In MESA, the multi-ethnic cohort published by Detrano in the New England Journal of Medicine (2008; 6,722 people), a score of 101 to 300 carried a 7.73-fold adjusted risk of a coronary event against a score of zero, and a score over 300 a 9.67-fold risk; each doubling of the score raised the risk of a major event by 15 to 35%. The reason it's on a men's list is that a man in his forties with a zero is a man whose risk calculator can be believed, and a man with a 300 is a man who needs to be on a statin now, whatever his cholesterol says. It costs about the price of a dinner out, and most men have never heard of it.
Colonoscopy — with the effect size read honestly
The only randomised trial of screening colonoscopy is NordICC (Bretthauer, NEJM 2022): 84,585 people aged 55 to 64 in Poland, Norway and Sweden, followed ten years. Inviting people to a colonoscopy reduced colorectal cancer incidence from 1.20% to 0.98% — a relative reduction of 18% — and the death reduction, 0.31% to 0.28%, wasn't statistically significant. Only 42% of those invited actually showed up, which drags the intention-to-screen numbers toward zero. The 13-year data, presented at Digestive Disease Week in May 2026 and not yet peer-reviewed, sharpened both halves: incidence reduction held at 19% overall and reached 45% among people who actually attended, while mortality still didn't separate. The honest summary: the procedure clearly prevents cancers in people who have it; whether that translates to lives saved is, on the trial evidence, less certain than the campaigns suggest. It's still on the list. It just isn't the guarantee it's sold as.
PSA — the decision, not the test
This is the one where the numbers cut both ways, and the US Preventive Services Task Force's 2018 statement is unusual for laying them out. For men 55 to 69 it's a Grade C — an individual decision — and for men 70 and over a Grade D, recommending against. Per 1,000 men screened over about 13 years: roughly 1.3 prostate cancer deaths prevented and 3 metastatic cases prevented, against about 240 false-positive results, 220 biopsies, 50 cases of sexual dysfunction and 15 of urinary incontinence caused by treatment, with 20 to 50% of screen-detected cancers estimated to be overdiagnoses that would never have caused harm. A man with a family history or of African descent sits further toward "yes." A man at average risk who values continence and function should understand exactly what he's trading before he says yes to a number he can't un-know.
The Short List With Trials Behind It
Evidence: Mixed — a few have real trials; most of the shelf has noneSupplements sit near the top not because they're bad but because their effects, where they exist, are measured in months and in specific people. This level is short on purpose.
Creatine has the best evidence of anything on the shelf for men who train — the muscle findings are solid, the brain findings genuinely contested — and it has its own claim-by-claim review here rather than a repeat.
Fish oil is the cautionary tale, because it's the supplement men most often take "for the heart," and the three big trials disagree in an instructive way. VITAL (Manson, NEJM 2019; 25,871 people, 1 g a day, 5.3 years) found no reduction in major cardiovascular events — hazard ratio 0.92, not significant — and no effect on cancer. STRENGTH (Nicholls, JAMA 2020; 13,078 high-risk patients, 4 g a day of mixed EPA and DHA) was stopped early for futility at a hazard ratio of 0.99, with more gut side effects. REDUCE-IT (Bhatt, NEJM 2019; 8,179 statin-treated patients with triglycerides of 150 to 499) used 4 g a day of purified EPA and cut events from 22.0% to 17.2% — hazard ratio 0.75 — but against a mineral-oil placebo that may itself have raised risk in the control group, and with atrial fibrillation rising from 3.9% to 5.3%. The lesson isn't "fish oil works" or "fish oil doesn't." It's that a specific prescription-strength formulation, in a specific high-triglyceride patient already on a statin, has one positive trial with an asterisk — and the capsule from the supermarket has two large null ones.
Everything else — vitamin D without a measured deficiency, magnesium without a reason, the multivitamin, the greens powder — belongs to a category the site's supplement guide covers: things that are probably harmless and probably do nothing, purchased instead of the base.
What Gets Sold Hardest
Evidence: None for longevity — safety data at best, marketing at worstThe tip of the pyramid is where the money is, and it's narrow because the evidence is. Four things live here, and each deserves a precise sentence rather than a dismissal.
Testosterone for men whose testosterone is normal. The TRAVERSE trial (Lincoff, NEJM 2023) is the largest cardiovascular safety study of testosterone replacement ever run: 5,246 men aged 45 to 80, two fasting levels below 300 ng/dL, treated to a target of 350 to 750, followed a mean of 33 months. The primary cardiovascular endpoint came out at 7.0% on testosterone versus 7.3% on placebo, hazard ratio 0.96 — non-inferior, which settled a long argument about harm. It also settled nothing about benefit: no reduction in events, and higher rates of atrial fibrillation, acute kidney injury and pulmonary embolism in the treated group. For a man with two low readings and symptoms, replacement is a legitimate treatment with a now-quantified risk profile. For a man with normal testosterone who's been told it will make him live longer, there is no trial that says so, and TRAVERSE is the trial that would have.
NAD precursors, peptides, biological-age tests. Each has its own honest piece here — NMN and NR, peptides, and the reliability of epigenetic clocks — and the common thread is the same: real biology, real mechanism, no human outcome data, and a price that assumes the outcome data exists.
Rapamycin, metformin for the non-diabetic, and the rest of the geroscience pipeline. Genuinely interesting, genuinely unproven in people. There is no human trial in which any of them has reduced mortality or extended lifespan. That may change; the honest position until it does is that they belong in a trial, not in a protocol.
The Honest Scorecard
What decides decades: the base — fitness, strength, not smoking, sleep, connection — with cohort evidence in the hundreds of thousands and effects that dwarf everything above them. Low fitness alone outranked smoking as a mortality predictor in the same cohort. What decides years: your waist, your blood pressure and lipids, the eating pattern, and — for men specifically — alcohol, which accounts for 6.8% of male deaths and whose protective dose, on the cleanest analysis, doesn't exist. What catches what slipped through: a calcium score once, a colonoscopy with the effect size read honestly, and a PSA decision made with the trade-off in front of you. What's worth months: creatine if you train; prescription EPA if you're the specific patient REDUCE-IT enrolled. What has no mortality evidence at all: the entire tip, including testosterone for men who don't need it. The pattern: the width of each level and the volume at which it's sold are close to inversely related, and the gap between men's and women's life expectancy lives almost entirely in the bottom three.
The Protocol: What to Actually Do
This month: measure the base
A grip test (a $30 dynamometer, or a dead hang), a waist measurement at the navel, a blood pressure reading at a pharmacy, and an honest count of your drinks per week. Four numbers, no doctor required. Our calculator turns them into an estimate with the study behind every input.
This quarter: move the two biggest levers
Fitness and strength. Two strength sessions a week, one hard interval session, and enough easy movement to make the treadmill test in ten years say "above average." If you smoke, this is the quarter, and the Jha numbers say you get most of the decade back if you're under 45.
This year: the blood draw and the one scan
Lipids with ApoB, HbA1c, and if you're over 40 and have never had one, a coronary calcium score. Decide on colonoscopy timing with your doctor from 45. Read the PSA trade-off before you're 55 so the decision is yours and not the default.
Ongoing: spend the tip's money on the base
The peptide budget buys a lot of dynamometer. If a biological-age test motivates you to train, it's earned its price; if it replaces the training, it hasn't. And the lever men drop first — the friend you haven't called — costs nothing and carries a 29% odds ratio.
The pyramid is upside down in every ad you'll see this year. Turn it over.
Men lose 4.9 years to women, and almost all of it is on the bottom three levels: the fitness never built, the strength let go, the drinks not counted, the doctor not seen, the friend not called. None of those cost what the tip costs. All of them have mortality trials behind them.
So the order is the whole argument. Build the base until it's boring. Get the four numbers. Do the one scan. Then, if there's money left and the base is handled, the tip is yours to experiment with — as an experiment, with its honest label on. Just never in place of the treadmill.