Key Takeaways
  • American women live to 81.4 on average; men to 76.5. The lead is real, and the levers that protect it are not identical to men's — the heart timeline, the bones, the screening trade-offs and the tip of the pyramid all differ
  • The base is the same five things — not smoking, fitness, strength, sleep, connection — and the payoff is bigger: women at 50 with five low-risk habits could expect 43.1 more years; with none, 29.0. Fourteen years. Women who smoke carry a mortality hazard ratio of 3.0; women below 5 METs on a treadmill, 3.1
  • The middle has a women's clock in it: blood pressure rises steeper from the twenties, and menopause before 45 carries 50% higher coronary risk. For the bones, heavy supervised lifting raised spine density 2.9% in eight months; calcium and vitamin D pills did not reduce hip fracture
  • Alcohol is the women's item: mortality risk rises from about two drinks a day for women where it doesn't for men, and each 10 g a day adds 7.1% to breast cancer risk
  • Screening is a trade-off with numbers: per 10,000 women screened from 50 for 20 years, 43 deaths prevented, 129 overdiagnosed. A calcium score finds plaque in 36% of "low-risk" women
  • The tip: HRT is safer than 2002 said and is not a longevity treatment — 18-year mortality hazard ratio 0.99. "Hormone balancing," collagen, biological-age tests and peptides have no mortality data at all

The men's version of this article opened with the gap: 4.9 years, most of it on the bottom three levels of the pyramid. The women's version has to open with something less flattering to the premise, which is that women already win. In the most recent US mortality data, life expectancy at birth was 81.4 years for women and 76.5 for men. So the question isn't how to catch up. It's what protects the lead, what erodes it, and where the evidence for women is different from the evidence for men — because it is, in ways that the generic longevity list never mentions.

Four differences shape this pyramid. The cardiovascular timeline: women's risk arrives later, then accelerates through the menopause transition, so the middle of the pyramid has a clock in it. The bones, which get their own level of attention and their own bad advice. The screening trade-offs, which for women are the most carefully quantified in medicine and the least often explained. And the tip, where hormone therapy sits with twenty years of contradictory headlines behind it. Same six levels as before, ranked by how much each thing moves mortality in the studies that measured it, with the study named every time.

The Longevity Pyramid for Women Six levels, widest at the base. Base: the foundation — don't smoke, fitness, strength, sleep, connection. Then body, blood and the menopause clock. Then eating and drinking. Then screening. Then supplements, mostly what doesn't work. Tip: HRT and the things sold hardest. 1 · The Foundation — don't smoke · fitness · strength · sleep · connection 2 · Body, blood & the menopause clock — waist · pressure · lipids · bones 3 · What you eat and drink 4 · Screening, with the numbers 5 · Supplements — mostly what doesn't 6 · HRT & the tip 1 · Foundation 2 · Body & blood 3 · Eat & drink 4 · Screening 5 · Supplements 6 · Tip

Width is the size and certainty of the mortality evidence. For women the top two levels are, if anything, sold harder and evidenced less than they are for men.

Foundation. Decades. Cohort effects in the hundreds of thousands; a 14-year spread at age 50.
Body, blood & bones. Years. Tape measure, cuff, a DEXA at the right age — and a clock men don't have.
Eating & drinking. Years. Alcohol is the women's item, twice over.
Screening. Real benefits, real overdiagnosis, and the numbers exist for both.
Supplements. Mostly large null trials. The bone pill women are told to take doesn't prevent fractures.
The tip. HRT, honestly framed, and the things with no data at all.
Level 1 · The base

The Five Things That Decide Decades

Evidence: Strong — large cohorts, replicated, consistent direction, women-specific numbers

The foundation is the same five habits as the men's pyramid, and it's the foundation for the same reason: each has been measured against death itself, in cohorts big enough that the answer isn't in doubt. The difference is the size of the payoff. Li and colleagues in Circulation (2018) followed 78,865 women in the Nurses' Health Study for up to 34 years. A 50-year-old woman with none of 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, diet quality in the top 40% — could expect to live another 29.0 years. With all five: 43.1. That's 14.0 years, against 12.2 for men in the same analysis. The lifestyle spread is wider for women, not narrower.

14.0 yrs
Extra life expectancy at age 50 for women with all five low-risk habits versus none — 43.1 years against 29.0 (Li, Circulation 2018; 78,865 women followed up to 34 years)
01

Don't smoke — the hazard is higher for women

Jha's analysis in the New England Journal of Medicine (2013) followed 113,752 women and 88,496 men. The mortality hazard ratio for women who smoked was 3.0, against 2.8 for men — and the survival curves are the plainer statement: 70% of never-smoking women were alive at 79, against 38% of smokers. More than a decade of life. The reversal numbers hold across both sexes: quitting at 25 to 34 regained about ten years, at 35 to 44 about nine, at 45 to 54 about six, and quitting before 40 removed roughly 90% of the excess risk. The window narrows; it doesn't close.

02

Cardiorespiratory fitness — three times the mortality at the bottom

The men's article used a mixed cohort. The women's has its own: the St James Women Take Heart Project, in which Gulati and colleagues put 5,721 asymptomatic women, mean age 52, on a treadmill in 1992 and tracked deaths to 2000 (Circulation, 2003). Every extra MET of exercise capacity — roughly the difference between a slow walk and a brisk one — meant 17% lower mortality after adjusting for the Framingham risk score. Women below 5 METs had a hazard ratio of 3.1 against women above 8; women at 5 to 8, 1.9. Three times the death rate, from a variable no blood test measures and no supplement moves. The Norwegian 4x4 is the best-studied way to raise it; Zone 2 is the base.

03

Strength — and it's a bone lever, not just a muscle one

In UK Biobank (Celis-Morales, BMJ 2018; 502,293 people, 54% women), every 5 kg less grip strength carried a hazard ratio of 1.20 for all-cause mortality in women — slightly higher than the men's 1.16 — 1.19 for cardiovascular death, and, unexpectedly, 1.24 for breast cancer. Weakness was defined at under 16 kg for women. Grip is a proxy for total muscle, and muscle for women is also the thing that loads the skeleton.

Which is where the bone evidence belongs — at the base, not the supplement level. The LIFTMOR trial (Watson, Journal of Bone and Mineral Research 2018) took 101 postmenopausal women, mean age 65, with low bone mass, and randomised half to eight months of twice-weekly, thirty-minute, supervised heavy lifting — five sets of five at over 85% of one-rep max, deadlifts, squats, overhead press — against a home programme. Lumbar spine bone density rose 2.9% in the lifting group and fell 1.2% in the controls; at the femoral neck the lifters held (+0.3%) while controls lost 1.9%. They also got taller, by 0.2 cm against a 0.2 cm loss. One minor back spasm in 70 sessions. That is a larger and better-evidenced bone result than the calcium pill on level five, and it comes with the mortality benefit of the strength itself. The grip piece has the test; creatine for women covers the one supplement with a real case here.

04

Sleep — seven hours, at the same time

The mortality curve for sleep is a U in both sexes: Cappuccio's meta-analysis in Sleep (1,382,999 people) found short sleep carried a relative risk of 1.12 and long sleep 1.30, the latter mostly illness recorded as habit. The newer finding, from 60,977 people wearing accelerometers (Windred, Sleep 2024), is that regularity predicts death better than duration — the most regular fifth had 30% lower all-cause mortality than the least regular. For women the sleep story has an extra chapter, because vasomotor symptoms through the menopause transition fragment sleep for years, and that's one of the things the tip of this pyramid genuinely treats. The sleep and longevity piece has the whole ladder.

05

Not being alone — same size effect, and it's often the lever women hold for everyone else

Holt-Lunstad's meta-analysis in Perspectives on Psychological Science (2015) found social isolation raised the odds of death by 29%, loneliness by 26%, living alone by 32% — larger under 65, and, the authors note, consistent across gender. Women are more often the ones maintaining everyone else's connections, and the data on widowhood and on caregiving are a reminder that the lever runs both ways. It's on the base level because the effect size is base-level.

Level 2 · Body, blood, and the clock

The Numbers — and the Timeline Men Don't Have

Evidence: Strong — large meta-analyses and cohorts; two of them women-only

One level up are the measurements. Two of them are the same for everyone. Two are women's.

Waist, not weight. Jayedi's meta-analysis of 72 cohorts and 2,528,297 people in the BMJ (2020) put every 10 cm of waist circumference at 11% higher all-cause mortality, every 0.1 of waist-to-hip ratio at 20%, independent of BMI, with the J-shape holding in both sexes. Larger hips and thighs went the other way (hazard ratios 0.90 and 0.82) — which is the muscle from level one showing up, and also why the gluteofemoral fat pattern more common in women is, on this measure, protective. This is the plan for the waist.

Blood pressure — earlier than you think. Ji and colleagues pooled 144,599 blood pressure readings from 32,833 people across four US cohorts spanning 43 years (JAMA Cardiology, 2020) and found that women's blood pressure rises more steeply than men's, relative to each sex's starting point, beginning in the third decade — the twenties — and continuing through life. The practical implication is that "women's heart risk starts at menopause" is a half-truth: the trajectory that produces the post-menopause event started thirty years earlier. A cuff at the pharmacy, once a year from your twenties, is the cheapest thing on this pyramid.

The menopause clock. Muka's meta-analysis of 32 studies and 310,329 women in JAMA Cardiology (2016) found that menopause before 45, compared with 45 or later, carried a relative risk of 1.50 for coronary heart disease, 1.19 for cardiovascular death and 1.12 for all-cause mortality. This isn't a lever — you don't choose the timing — but it's a risk marker that a doctor should know and that most intake forms don't ask. Early menopause moves a woman up a risk category the way a family history does, and it shifts the balance on both the screening below and the HRT question at the tip.

Lipids and glucose. The lipid panel — ideally with ApoB — and an HbA1c, one draw a year. Same as the men's list, with one note: the pre-menopausal lipid profile is often reassuring and changes after, so a number from 45 is not a number for 55. The site's cholesterol piece for women covers the specifics.

Level 3 · Eating and drinking

The Pattern, and the Women's Problem

Evidence: Strong for the pattern; strong and specific for alcohol

The eating pattern is the one every Blue Zone eats and the one this site's kitchen is built around: mostly plants, legumes daily, fish more than red meat, very little in a wrapper. The legume line has a fresh number — a 2026 meta-analysis in BMJ Nutrition, Prevention & Health (10 prospective cohorts, 309,853 people) put high versus low intake at a relative risk of 0.84 for hypertension, roughly linear up to about 170 g a day. Protein deserves a line for women over 50 specifically, because the muscle and bone on level one have to be built from something and intake tends to fall exactly when the need rises; the protein guide has the targets.

Then alcohol, which is where the women's pyramid diverges from the men's most sharply, for two reasons that stack.

The mortality threshold is lower. Zhao's 2023 meta-analysis of 107 cohorts and 4,838,825 people in JAMA Network Open found no protective effect of light drinking in either sex once the studies were corrected for who counts as an abstainer — and it found the risk rising earlier in women. At 25 to 44 g a day, roughly two to three drinks, women's all-cause mortality was 1.21 times that of lifetime non-drinkers; men's at the same intake was 1.01, not significant. Men's risk didn't become significant until 45 g. The authors' summary: the increased risk starts "at lower levels for women than men."
And there's a second endpoint men don't have. The Collaborative Group on Hormonal Factors in Breast Cancer pooled 53 studies — 58,515 women with invasive breast cancer and 95,067 without — in the British Journal of Cancer (2002). Each additional 10 g of alcohol a day, about one drink, raised the relative risk of breast cancer by 7.1%, linearly. At 35 to 44 g a day the relative risk was 1.32; at 45 g or more, 1.46. If the relationship is causal — the authors' qualifier — about 4% of breast cancers in developed countries are attributable to alcohol. Tobacco, in the same analysis, had "little or no independent effect." The wine is the one that matters here.

Put together: the level at which a woman's drinking starts costing her years is about two drinks a day, and every drink below that is buying a measurable slice of breast cancer risk. That's not a temperance argument. It's the same evidence ranking the rest of the pyramid uses, applied to the one habit where the numbers for women are worse than the numbers for men.

Level 4 · Screening

The Tests, With Both Numbers Attached

Evidence: Moderate to strong — real benefits, quantified harms, and guidelines that agree on most of it

Women's screening is the most thoroughly quantified in medicine, and the quantification is almost never shared with the woman being screened. Here it is.

01

Mammography — one death prevented for every three overdiagnosed

The clearest statement of the trade-off is the UK's independent panel, chaired by Michael Marmot, in the Lancet (2012). Across 11 randomised trials, women invited to screening had a relative risk of breast cancer death of 0.80 — a 20% reduction. The other side: for every 10,000 women invited from age 50 for 20 years, 43 breast cancer deaths would be prevented, and 129 cancers — invasive and non-invasive — would be diagnosed and treated that would never have caused harm. One death prevented for about every three overdiagnosed. The panel's own caveat is that the trials are old and the estimates "should be regarded only as an approximate guide"; the direction of both numbers is not in doubt. Most guidelines recommend screening from 40 or 50, and this article does too — the point is to walk in knowing that a positive result has a real chance of being a cancer that would never have mattered, and to ask that question out loud when it happens.

02

Cervical screening — the one with the cleanest evidence

The US Preventive Services Task Force's 2018 statement (JAMA) is a Grade A: cytology every three years from 21 to 29, and from 30 to 65 either cytology every three years, high-risk HPV testing every five, or both together every five. Cervical cancer is the screening success story — a slow, detectable, treatable precursor and a test that finds it — and the harms are small. If you skip one screen on this list, don't make it this one.

03

Bone density — at 65, or earlier with a reason

The USPSTF's January 2025 update (JAMA) recommends DEXA screening for all women from 65, and for postmenopausal women under 65 at increased risk — early menopause, low body weight, a parent with a hip fracture, long steroid use. A Grade B in both cases. It's cheap, it's a five-minute scan, and it turns the bone conversation from guesswork into a number. What you do with the number is on levels one and five: lift, and be sceptical of the pill.

04

Coronary calcium — because "low risk" isn't

The calculators that estimate a woman's ten-year cardiovascular risk tend to call her low-risk well into her fifties. Kavousi and colleagues tested that in JAMA (2016): 6,739 women from five cohorts, all with a calculated ten-year risk under 7.5%, all given a coronary calcium CT. Thirty-six percent had calcified plaque. Those women had a hazard ratio of 2.04 for cardiovascular events compared with women whose score was zero — 4.33 events per 1,000 woman-years against 1.41. The absolute numbers are low and the authors call the improvement in prediction "modest," which is fair. But a third of women the calculator cleared had disease the scan could see. Once, in your late forties or fifties, especially with early menopause or a family history, it's the scan that can move a woman from "low risk, no treatment" to a statin she should have been on.

Colorectal screening is the same decision for women as for men — the NordICC trial (NEJM, 2022) is walked through in the men's article — and it starts at 45.

Level 5 · Supplements

The Bone Pill, and Other Null Results

Evidence: Strong — and mostly negative. The trials are large; the effects aren't there

The supplement level of the women's pyramid is unusual: the trials are big, women-focused, and clear. They are also mostly null, and the most confidently recommended supplement in women's health is the one with the least behind it.

Calcium and vitamin D for bones. The Women's Health Initiative randomised 36,282 postmenopausal women aged 50 to 79 to 1,000 mg of calcium plus 400 IU of vitamin D daily or placebo and followed them for seven years (Jackson, NEJM 2006). Hip fracture hazard ratio: 0.88, confidence interval 0.72 to 1.08 — not significant. Total fractures: 0.96. Hip bone density was 1.06% higher, which is a statistically real, clinically trivial number next to LIFTMOR's 2.9%. And kidney stones rose 17% (hazard ratio 1.17, significant). The authors' own summary was that supplementation "did not significantly reduce hip fracture, and increased the risk of kidney stones." Dietary calcium is a different question; the pill is this one.

Vitamin D on its own. VITAL (Manson, NEJM 2019) gave 25,871 people 2,000 IU a day for a median of 5.3 years: invasive cancer hazard ratio 0.96, major cardiovascular events 0.97, neither significant. Its fracture analysis (LeBoff, NEJM 2022; 13,085 of the participants were women) found 1,991 fractures and no effect: total fractures 0.98, hip 1.01. The participants weren't selected for deficiency, and that's the correct reading — a measured deficiency is a reason to treat; a healthy woman taking vitamin D "for her bones" is taking a supplement with three large null trials behind it.

What's left. Creatine has the best evidence on the shelf for women who train, and it's cheap. Omega-3 for the heart has the same three-trial story as in the men's article — null in VITAL, null in STRENGTH, one positive prescription-dose trial in a specific population. Magnesium's sleep evidence is three small trials. Everything else — collagen, the "hormone balancing" blend, the greens powder — has no mortality data and mostly no outcome data at all, and it's on level six.

Level 6 · The tip

HRT, Honestly — and What Gets Sold Above It

Evidence: Strong for HRT's real effects; none for longevity from anything on this level

The tip of the women's pyramid is different from the men's because one thing on it has serious trial evidence — just not for the claim it's often sold on. It gets its own paragraph, and then the rest of the tip gets the short treatment it deserves.

Hormone therapy is safer than 2002 said, and it's not a longevity treatment. The Women's Health Initiative's oestrogen-plus-progestin trial (Rossouw, JAMA 2002; 16,608 women aged 50 to 79) was stopped early, and its hazard ratios made the headlines: coronary heart disease 1.29, breast cancer 1.26, stroke 1.41, pulmonary embolism 2.13. What the headlines left out were the absolute numbers, per 10,000 woman-years: 7 more heart events, 8 more strokes, 8 more clots, 8 more breast cancers — and 5 fewer hip fractures, 6 fewer colorectal cancers. Small numbers, both directions. Total mortality in that trial: 0.98.

The long follow-up settled the mortality question. Manson's 2017 analysis in JAMA followed all 27,347 women in both WHI hormone trials for a cumulative 18 years, logging 7,489 deaths: all-cause mortality hazard ratio 0.99, cardiovascular 1.00, cancer 1.03. Hormone therapy neither shortened nor extended life. Within that, the younger women — 50 to 59 at the start — did better than the older ones during the treatment years, an effect that faded over the full 18. The Menopause Society's 2022 position (Menopause) is the current consensus: for women under 60 or within ten years of menopause, with no contraindications, the benefit-risk balance is favourable for treating bothersome vasomotor symptoms and preventing bone loss; started later than that, the balance tips the other way because of coronary disease, stroke, clots and dementia. It is the most effective treatment for hot flushes and the genitourinary symptoms, it protects bone, and it's a reasonable choice for the woman it's indicated for. What it is not is a thing to take for longevity. The 18-year number is 0.99.

"Hormone balancing," collagen, and the biological-age test. The first is a category of supplement, not a treatment, and no product in it has a mortality trial or, mostly, an outcome trial. Collagen has skin-hydration studies of a few weeks and nothing longer; the skin article gives it its honest paragraph. Biological-age tests are addressed here — in the 2026 Aging Cell reliability study, most clocks were only moderately stable in the same person over short intervals — and the version marketed to women, with menopause framed as accelerated ageing to be reversed, is the same test with a different landing page. Peptides, NAD precursors and the geroscience pipeline: interesting, unproven in humans, no mortality data, and above the bone-density scan in every catalogue.

The Honest Scorecard

What decides decades: the base — not smoking (hazard ratio 3.0 for women who do), fitness (3.1 at the bottom of the treadmill), strength (1.20 per 5 kg, and the only thing that grew bone by 2.9%), regular sleep, connection. A 14-year spread at age 50. What decides years: the waist, the cuff from your twenties, knowing your menopause timing, the legumes, and treating two drinks a day as the threshold it is. What catches things: cervical screening without hesitation; mammography knowing the 43-to-129; DEXA at 65 or earlier with a reason; a calcium score once. What doesn't: the calcium-and-D pill for fractures, vitamin D for anything in a non-deficient woman, and the whole of level six for longevity — including HRT, which is a good treatment for something else.

The Protocol: What to Actually Do

01

This month: measure the base

A grip test (a $30 dynamometer; under 16 kg is the weakness line), a waist measurement at the navel, a blood pressure reading — whatever your age — and an honest weekly drink count against the two-a-day line. Write down the age your mother reached menopause, and yours if you're there.

02

This quarter: lift heavy, and get fit enough to be above 8 METs

Two strength sessions a week with real load — the LIFTMOR protocol was five sets of five, supervised, and that supervision is the safety mechanism, so start with a coach if you've never deadlifted. One hard interval session. Enough easy movement that a treadmill test in ten years puts you in the top group. If you smoke, this is the quarter you stop.

03

This year: the draw, the scans, and the screening decisions made on purpose

Lipids with ApoB and HbA1c. Cervical screening on schedule. Mammography per your guideline, having read the 43-to-129 once. DEXA at 65, or now if you have a reason. A coronary calcium score once, from the late forties, especially with early menopause or a family history. Drop the calcium-and-D pill unless a measured deficiency or a diagnosis says otherwise.

04

Ongoing: decide the tip on its own terms

If vasomotor symptoms are wrecking your sleep and you're under 60 or within ten years of menopause, HRT is a legitimate conversation to have with a doctor — as a symptom treatment with a favourable balance, not a longevity play. Spend the "hormone balancing" money on the coach. Spend the biological-age-test money on the DEXA.

Common Questions
What matters most for women's longevity?
The base: not smoking, cardiorespiratory fitness, strength, about seven regular hours of sleep, and not being isolated. Women at 50 with five low-risk habits could expect 43.1 more years versus 29.0 with none in the Nurses' Health Study — a 14-year spread, larger than the men's 12.2. In 5,721 women on a treadmill, each extra MET meant 17% lower mortality, and women under 5 METs had three times the death rate of women above 8.
Is hormone replacement therapy dangerous?
Less than the 2002 headlines said, and it's not a longevity treatment either. WHI's oestrogen-plus-progestin trial found small absolute changes per 10,000 woman-years — 7 more heart events, 8 more strokes, 8 more clots, 8 more breast cancers, 5 fewer hip fractures, 6 fewer colorectal cancers — and its 18-year follow-up of 27,347 women found no difference in all-cause mortality (hazard ratio 0.99). The Menopause Society's position: under 60 or within ten years of menopause, the balance favours treating symptoms and protecting bone. It treats symptoms. It doesn't extend life.
Should women take calcium and vitamin D for their bones?
Not for fracture prevention in women who aren't deficient — the large trials are null. WHI: 36,282 women, 1,000 mg calcium and 400 IU D for seven years, hip fracture hazard ratio 0.88 with a confidence interval crossing 1, and 17% more kidney stones. VITAL: 2,000 IU of D in 25,871 people, no reduction in cancer, cardiovascular events or fractures. What grew bone was heavy supervised lifting — 2.9% at the spine in eight months in LIFTMOR.
Should I have a mammogram?
Most guidelines say yes, and so does this article — knowing both numbers. The UK independent panel's estimate: per 10,000 women invited from 50 for 20 years, 43 breast cancer deaths prevented and 129 cancers overdiagnosed and treated that would never have caused harm, about one to three. Screening cut breast cancer mortality by about 20% across 11 trials. Go in informed, and ask the overdiagnosis question if a result comes back.
The Bottom Line

Women start ahead. The pyramid is how the lead gets kept.

The four-and-a-half years women hold over men are real and mostly biological. The fourteen years between the women who build the base and the women who don't are lifestyle, and they dwarf everything the wellness industry sells to women specifically. The bone pill doesn't work; the barbell does. The wine costs more than it costs a man. The screening numbers exist and nobody reads them out. And the thing at the tip with the scariest headlines turns out to be a decent treatment for symptoms and a non-event for mortality — which is more than can be said for anything sold above it.

Build the base until it's boring. Get the four numbers, and the fifth one that's a date. Screen on purpose. Then, if there's money left, spend it on a coach.