The individual question and the aggregate question have different answers, and the answers are not in tension. They are the same mechanism observed at two scales, and the mechanism is habit.
Ask whether a given sixty-five-year-old can be strong, aerobically fit, metabolically clean and sexually functional. The answer is yes, with a discipline that is well understood, cheap, and documented in randomised trials. Ask whether a population of sixty-five-year-olds will be those things. The answer is no, and the reason is not that the population lacks the information. It is that the population has to make the decision every day, for forty years, and the environment is engineered against it.
The individual is near-certain. The aggregate is near-impossible. Both statements are true, and the gap between them is where the money, the suffering, and the policy failure all live.
I. What Actually Declines
The decline is real and it is measurable, and the first thing to understand is that it is not one process. It is several, running at different rates, and they respond to different inputs.
Aerobic capacity. VO₂max falls at roughly 1 per cent per year after thirty in sedentary adults, or about 10 per cent per decade. That is the headline number. The quieter number is that the fall is not a property of age. It is a property of detraining. A lifelong endurance athlete loses roughly half that rate, and the difference between the two curves by seventy is the difference between climbing a flight of stairs and needing a chair at the top of it.
Muscle mass and strength. Sarcopenia runs at roughly 3–8 per cent per decade after thirty, accelerating after sixty. Strength falls faster than mass, because the motor units are lost before the fibres are. The consequence is not cosmetic. It is the loss of the reserve that determines whether a fall is a bruise or a fracture, and whether a fracture is a recovery or a death.
Metabolic capacity. Insulin sensitivity falls, visceral fat rises, and the glucose response to the same meal worsens decade by decade. The trajectory is not fixed. It is the most responsive of the three to training, and the least responsive to any drug currently licensed.
Vascular and endocrine function. Endothelial function declines, nitric oxide availability falls, and the penile circulation is the first place it shows. That is not a joke and it is not a separate problem. It is the same problem, in the smallest vessel in the body, and it announces itself years before the coronary event.
The twin literature puts the heritability of lifespan at roughly 20–30 per cent. The rest is environment and behaviour. The genome sets the range. The behaviour determines where in the range you live.
II. What Responds
Now the other half, which is the half the doomer literature omits.
Resistance training works at any age. The landmark demonstration is Fiatarone and colleagues in JAMA in 1990: eight weeks of high-intensity resistance training in nonagenarians produced strength gains of roughly 170 per cent and measurable muscle hypertrophy. The subjects were ninety. The adaptation machinery was intact. It had simply not been asked to work.
Aerobic training attenuates the VO₂max slope. The trained curve is not flat, but it is roughly half the gradient of the sedentary one. The practical translation is that a trained sixty-five-year-old has the aerobic capacity of an untrained forty-five-year-old, and the reserve to absorb a decade of illness without losing independence.
Strength predicts survival. The push-up study — Yang and colleagues, JAMA Network Open, 2019, 1,104 middle-aged men — found that men able to complete more than 40 push-ups had a roughly 96 per cent lower risk of a cardiovascular event over the follow-up than men who could complete fewer than ten. Grip strength, in the PURE study, is inversely associated with all-cause mortality at roughly 16 per cent per 5 kg decrement. These are not supplements. They are load-bearing capacities.
The dose is small. The guideline is 150 minutes of moderate activity per week, or 75 vigorous, plus two resistance sessions. That is roughly 2 per cent of waking hours. The mortality reduction attached to meeting it is on the order of 20–30 per cent all-cause. There is no pharmaceutical on the market with that effect size and that side-effect profile.
The metabolic response is documented in Finland. The Finnish Diabetes Prevention Study — Tuomilehto and colleagues, NEJM, 2001 — took people with impaired glucose tolerance and gave them a lifestyle intervention: weight loss, dietary change, and moderate exercise. The incidence of type 2 diabetes fell by 58 per cent over the trial period. Fifty-eight per cent, from behaviour, with no drug.
III. The Individual Arithmetic
Put the numbers together and the individual case is not a hope. It is a projection with a wide margin of error and a clear sign.
A man of fifty who begins resistance training twice a week and walks briskly for thirty minutes on five days is not guaranteed to live to ninety. He is near-certain to be stronger at sixty than he is at fifty, to have a VO₂max closer to his forty-year-old self than to his sedentary peer, to have better glucose control, and to have a lower ten-year cardiovascular risk. Those are capacities, and capacities are near-certainly improvable. Lifespan is probabilistic. Capacity is not.
This distinction matters because the wellness industry conflates them in both directions. The supplement seller promises lifespan and delivers nothing. The doomer says the decline is inevitable and delivers surrender. The correct statement is narrower and more useful: the capacity is near-certainly improvable, and the improvement is what determines the quality of the last two decades, which is the part you actually experience.
IV. The Virility Question
This deserves its own section because it is the one the culture cannot discuss without either euphemism or comedy, and because the physiology is unusually clean.
Erectile function is a vascular event. It requires endothelial nitric oxide signalling, adequate arterial inflow, and venous occlusion. It is therefore a direct readout of vascular health, in the smallest and most demanding vessel bed in the body. The clinical consequence is well established: erectile dysfunction is a sentinel for cardiovascular disease, typically preceding a cardiac event by three to five years. The penis is the body’s early-warning system, and the warning is being ignored because of the embarrassment attached to reading it.
The prevalence is not marginal. The Massachusetts Male Aging Study found that roughly half of men aged 40–70 report some degree of erectile dysfunction. That is not a fringe complaint. It is the median experience of the male population in the second half of life.
The response is the same response as everything else in this essay. The Esposito randomised trial, JAMA 2004, put obese men with erectile dysfunction on a lifestyle programme — weight loss, Mediterranean-style diet, increased activity. Roughly one in three regained normal function, and the intervention group’s scores improved substantially against controls. That is a behavioural intervention producing a clinical remission in a condition the culture treats as an inevitability or a pharmaceutical problem.
Testosterone declines at roughly 1–2 per cent per year after thirty, and a large share of that decline is not gonadal. It is adipose tissue, sleep apnoea, insulin resistance and inactivity. The endocrine system is reporting the metabolic state. Fix the metabolic state and the report changes.
So the virility question, stated without euphemism, is this: the function is vascular, the vasculature responds to training and to fat loss, the evidence is randomised, and the dose is the same 150 minutes and two sessions that everything else requires. The man who wants to be functional at sixty knows what to do. The gap between knowing and doing is the entire essay.
V. Why the Aggregate Fails
Now the turn. Everything above is true, cheap, and documented. And the population does not do it.
Adherence is the binding constraint, not knowledge. Roughly half of people who begin a structured exercise programme quit within six months. Gym membership data shows the median member attends a small handful of times per month; the membership is a donation, not a habit. Digital health applications retain a few per cent of users at ninety days. The knowledge is universal. The behaviour is rare.
Habit formation is slower than the motivation that starts it. Lally and colleagues, in the European Journal of Social Psychology in 2010, found that the median time for a behaviour to become automatic was 66 days, with a range from 18 to 254. The motivation that starts a programme is front-loaded and decays within weeks. The habit that sustains it is not formed until well after the motivation has gone. Most people quit in the gap.
The benefit is deferred and the cost is immediate. This is hyperbolic discounting, and it is not a character flaw, it is a documented feature of how people value the future. The training session costs an hour now. The avoided fracture is thirty years away. No amount of information about the fracture changes the discount rate, because the discount rate is not an information problem.
The prevention paradox. Geoffrey Rose’s observation is that a measure which brings large benefits to the population offers little to each participating individual, and a measure which offers large benefit to each individual offers little to the population. The individual who trains gets a large benefit. The public health campaign that tries to move the population gets a small average effect, because most of the population does not comply, and the aggregate effect is the average of a small compliant minority and a large non-compliant majority.
The environment is engineered against it. The default is the chair, the car, the elevator, the delivered meal, and the screen. The trained state is a deliberate departure from the default, sustained against a gradient. The untrained state requires no decisions at all, which is why it is the modal state.
Resolution data. Roughly one in five people who make a New Year’s resolution maintain it at two years. The rest cycle. The cycling is not failure in the moral sense. It is the base rate of behaviour change under a deferred reward.
VI. The Finnish Evidence, Which Is Unusually Good
Finland is a useful case because it ran the experiment at population scale and published the results.
The North Karelia project, beginning in 1972, is the canonical demonstration that population-level behaviour change is possible. Cardiovascular mortality in the region fell dramatically over the following decades. It worked, and it took a generation, and it required a state that was willing to intervene in diet, smoking, and physical activity continuously for thirty years.
The Finnish Diabetes Prevention Study is the canonical demonstration that the individual intervention works when it is delivered: 58 per cent reduction in diabetes incidence with lifestyle counselling.
The sauna literature — Laukkanen and colleagues — is the odd one, and it is worth citing because it is Finnish and because it is robust: frequent sauna use is associated with lower cardiovascular and all-cause mortality in large Finnish cohorts. The mechanism is plausibly the same as moderate exercise: heat stress, vascular conditioning, and the haemodynamic load of repeated exposure. It is not a substitute for training. It is a reminder that the vascular system responds to being loaded, by any means.
The pattern across all three is consistent. The intervention works. The population effect is slow and modest. The individual effect is large and fast. The difference is adherence and time.
VII. The Counterfactual
The comparison that matters is not to a population that does not exist. It is to the counterfactual in which the aggregate behaves as the individual does, and that counterfactual is instructive precisely because it is unavailable.
If everyone met the guideline, the savings in morbidity, disability and care would be enormous and would arrive within twenty years. That is what makes the public health case. It is also what makes the public health case hopeless as a policy proposition, because it requires a behaviour change that the evidence says will not happen at that scale without an environmental change that no government has been willing to make.
What does work at scale is the thing that does not require a decision each day. The seatbelt law worked because the default changed. The smoking ban worked because the default changed. The fluoridated water worked because the default changed. The interventions that moved populations were the ones that removed the daily choice, not the ones that informed it.
Applied to this problem, that means the interventions that will actually move the aggregate are the ones that build the training into the environment: the stairs that are more convenient than the lift, the streets that are safe to walk, the workplace that expects the session, the food environment that makes the default portion the correct one. None of those requires discipline. All of them are politically expensive.
The individual does not need any of that. The individual needs a barbell and a pair of shoes and a calendar. That is why the individual succeeds and the aggregate does not.
Habits is why you can’t bank on the Western workforce staying fit enough to outcompete the Africans, the Indians and the South-East Asians. One in five might, if we’re lucky, but inertia says maybe one in twenty. That’s maybe 10 million “our” guys against a billion “their” guys. A 100x difference in scale. An actuary can already see where the market is.
VIII. What It Actually Costs
The bill is small and the currency is time.
The guideline is roughly 150 minutes a week plus two resistance sessions, which is about 2 per cent of waking hours. The kit is a pair of shoes and, eventually, a barbell and a rack. The skill requirement is a coach for a few months to teach the four or five movements that matter, and then nothing.
The return is a capacity curve that is roughly half as steep as the sedentary one, a strength reserve that determines whether the fall is an inconvenience or an ending, a metabolic state that delays or prevents the diabetes diagnosis, and a vascular function that keeps the sentinel quiet.
The alternative is not a single dramatic loss. It is a series of instalments, paid decade by decade, in capacity, in independence, and in function. The untrained sixty-five-year-old has not made one bad decision. He has made the same small non-decision every day for thirty years, and the aggregate of those non-decisions is the difference between the two curves.
You can buy the supplement. The supplement does not do the squat.
The capacity is near-certainly there. The evidence is randomised, Finnish, and boring. The dose is 2 per cent of the week. The only missing input is the habit, and the habit takes 66 days to form, and most people quit in the gap between the motivation and the automaticity.
The individual who closes that gap is not disciplined in the moral sense. He is committed to a decision he made once and then stopped re-litigating. The aggregate will not make that decision, and no amount of information will make it, because the information was never the constraint.
The constraint is that the payment is due now and the benefit arrives later, and the body does not accept cheques.
References
Fiatarone, M.A., et al. (1990). High-intensity strength training in nonagenarians: effects on skeletal muscle. JAMA, 263(22), 3029–3034.
Yang, J., et al. (2019). Association between push-up exercise capacity and future cardiovascular events among active adult men. JAMA Network Open, 2(2), e188341.
Leong, D.P., et al. (2015). Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study. The Lancet, 386(9990), 266–273.
Tuomilehto, J., et al. (2001). Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. New England Journal of Medicine, 344(18), 1343–1350.
Esposito, K., et al. (2004). Effect of lifestyle changes on erectile dysfunction in obese men: a randomized controlled trial. JAMA, 291(24), 2978–2984.
Lally, P., et al. (2010). How are habits formed: modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009.
Rose, G. (1981). Strategy of prevention: lessons from cardiovascular disease. British Medical Journal, 282(6279), 1847–1851.
Laukkanen, T., et al. (2015). Association between sauna bathing and fatal cardiovascular and all-cause mortality events. JAMA Internal Medicine, 175(4), 542–548.
Feldman, H.A., et al. (1994). Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. Journal of Urology, 151(1), 54–61.
Norcross, J.C., et al. (2002). Auld lang syne: success predictors, change processes, and self-reported outcomes of New Year’s resolvers and nonresolvers. Journal of Clinical Psychology, 58(4), 397–405.
Pullinen, T., et al. / Finnish Diabetes Prevention Study follow-up publications on long-term lifestyle intervention outcomes.
Puska, P., et al. North Karelia Project: 25-year results on cardiovascular mortality decline in Finland.