The short version
- The problem is not the length of life, it is the state of it. A man born in the UK today can expect about 79 years, and about 61 of them in good health. The gap is roughly two decades, and it has just got wider.
- Almost everything that closes that gap is boring and already known. Not smoking, blood pressure, moving, sleep, alcohol, and what is on the plate. Nothing on this page outranks those.
- Of the household exposures, one is worth acting on tonight: radon. It is the second leading cause of lung cancer in the UK, a test costs about the price of a takeaway, and most of the harm comes from homes nobody would think to test.
- Two others earn their place: a warm home and clean indoor air. Cold housing is the biggest exposure here and the least discussed — colder countries than ours have fewer excess winter deaths, because their homes are built for it.
- The supplements are last for a reason. They are the part of this topic with the weakest evidence and the loudest marketing, and putting them first would contradict everything above.
The number that reframes the question
Life expectancy is the figure everyone quotes. It is not the interesting one. The Office for National Statistics also publishes healthy life expectancy — the years you can expect to spend in self-reported good health — and the distance between the two is the part of your life that longevity advice should actually be aimed at.
| At birth, UK | Life expectancy | In good health | The gap |
|---|---|---|---|
| Men | 79.1 years | 60.7 years | 18.4 years |
| Women | 83.0 years | 60.9 years | 22.1 years |
Published February 2026, covering 2022–24. Healthy life expectancy fell by 1.8 years for men and 2.5 years for women against the previous period, and is now the lowest since the series began in 2011–13. Life expectancy itself edged up slightly over the same period. So the years are being added at the wrong end.
Two caveats about that table. “Good health” is how people describe their own health in a survey, not a clinical assessment — it is a real measure of something, but a subjective one. And these are averages across the whole country: the gap between the most and least deprived areas of England is around twenty years of healthy life, which is larger than anything an individual can do about with better habits. Longevity advice aimed at individuals cannot touch that, and should not pretend to.
Where this topic sits in the order of things
Longevity writing has a habit of putting everything on one list, so that a supplement with no human outcome data appears alongside not smoking as though they were comparable moves. They are not within two orders of magnitude of each other. Here is roughly how the things on this site actually rank, so that the rest of the page can be read in proportion.
| What | Roughly how much it matters | Where it is covered |
|---|---|---|
| Not smoking | The largest single item by a wide margin — on the order of a decade of life expectancy in lifelong smokers, and most of the excess risk is avoided by stopping before 40. | Vaping, for switching. Otherwise not a topic here. |
| Blood pressure | Years. Common, largely symptomless, and cheaply measured. | Blood pressure — the other number. |
| Atherogenic particle count | Years. The best-evidenced modifiable target in cardiovascular medicine. | Heart & cholesterol — the largest topic here. |
| Physical activity | Years, with most of the gain in the step from doing nothing to doing something. | Exercise & body composition. |
| Sleep | Probably years, though harder to quantify than the items above and easier to get wrong in both directions. | Sleep. |
| Alcohol | Months to years, strongly dose-dependent. | Alcohol. |
| Overall diet pattern | Years at the extremes of the range; less between two reasonable diets than the argument about them suggests. | Eating for the heart. |
| Cold housing | Large in the aggregate and concentrated in the old and the unwell. Close to zero if your home is already warm. | This topic. |
| Air pollution | Months on a population average, and mostly not under individual control outdoors. | This topic. |
| Radon | Small on average, substantial for the minority in affected homes — which is why testing beats estimating. | This topic. |
| The rest of the exposures | Individually small. Worth doing where they are nearly free, which most of them are. | This topic. |
| Longevity supplements | No demonstrated effect on human lifespan for any of them. | This topic, graded honestly. |
Four limits on that table. It is ordered by population impact, so it is not your personal order — if you already do not smoke, the second row is your first. The magnitudes do not add up and are not independent: cold housing acts partly through blood pressure, diet acts partly through particle count, and counting both twice would be double-counting. The bands are orders of magnitude drawn from the general literature, not effect estimates from a single study, and anyone quoting a precise number of days gained for these things is inventing precision. And the one remaining not a topic here row — smoking — is a gap in this site rather than a comment on its importance; it is the largest item on the table and is currently covered only for switching, under vaping. Sleep, blood pressure and alcohol were the other three gaps this table exposed, and all three now have chapters, which is what the table was for.
The useful conclusion is not that this topic does not matter. It is that this topic is the part you reach for after the rows above it, and that anything here which costs real money, real time or real worry has to justify itself against a table where it sits near the bottom.
Five things worth doing
These are not the top five of the eighteen exposures in the register, and the difference is worth stating. That register orders by how much each exposure costs the UK population. This list orders by something narrower and more useful to one person: how much you would gain, weighed against how cheap it is and how much of it is genuinely under your control.
The first four are the large levers, and they are large enough that the rest of this topic is a refinement on them. That is worth saying on a longevity page, because the genre spends most of its attention on the refinements. The fifth is here because it is real, measurable, and almost never listed alongside the others.
Overall diet pattern
Not a single food, and not a supplement — the pattern. A largely plant-forward, minimally processed diet with adequate protein and fibre is the version with the most consistent outcome evidence behind it, and every specific claim beneath that level is smaller than the pattern itself. The single most useful change for most people is reducing the ultra-processed share, which is also the top-ranked entry on the exposure register by population burden. Eating for the heart is the whole treatment; building the plate is the practical version.
Physical activity
The intervention with the broadest evidence base of anything on this site, and the one where the gap between doing nothing and doing a little is far larger than the gap between doing a little and doing a lot. Aerobic work for cardiovascular and metabolic risk, and resistance training for the thing that decides independence later — muscle and bone. See exercise and body composition, and bone density, where loading turns out to matter more than the supplement most people take instead.
Alcohol
The clearest case on this site of an exposure whose accepted position moved. The protective-J-curve reading has not survived better methods: analyses correcting for the sick-quitter effect and Mendelian randomisation both point to no safe protective dose, with risk rising from low intake for several cancers. That does not make one drink a catastrophe, and this page is not going to pretend it does — but the accurate statement is that less is better all the way down, which is not what most people were told. The alcohol chapter sets out the evidence and the caveats.
Sleep, and the two exposures nobody thinks of
Sleep is the fourth large lever and has its own chapter. Two things that act on it, and on much else, belong here rather than being lost in the register: a home kept above 18 °C in winter, which is probably the largest single exposure in this topic by total burden and is a housing problem rather than a weather one, and chronic night noise. Both are graded on the register, and both are addressable.
Stress — and it is not only a feeling
Chronic stress has measurable physiology, and the easiest place to see it is a glucose trace. Cortisol and adrenaline drive the liver to release glucose and make tissues less responsive to insulin, so a stressful meeting or a bad night raises blood glucose with no food involved at all — which is exactly the pattern people notice, and are surprised by, when they wear a continuous glucose monitor. The same axis raises blood pressure, disrupts sleep, drives central fat deposition, and changes eating behaviour.
What it is not is a licence to sell you anything. The interventions with actual evidence are unglamorous and overlap the four above: physical activity, sleep, and social connection, with structured psychological therapy where there is a treatable condition underneath. Where this matters most is interpretation — a raised glucose reading, a raised blood pressure, or a poor night during a hard month is often the stress showing, and treating the number without asking about the context is a common way to get the answer wrong.
The limit: stress is far harder to measure, define and randomise than the other four, so the evidence is weaker by nature rather than by neglect. It is on this list because leaving it off implies it does not act on the body, and it plainly does.
Four things the deep chapters turned up that change what is worth doing
- The largest single source of air pollution in the UK is domestic wood burning — about a fifth of PM2.5 emissions, more than twice road transport exhaust, and the only major source rising. If you have a stove or an open fire, that is a larger lever than anything else in this topic. Air pollution.
- Ventilation is now a job you have to do deliberately. Sealing and insulating homes removed the accidental air changes older houses had, and damp and mould are better evidenced as a respiratory hazard than most of the register. Seal-and-insulate and ventilate are one instruction, not two.
- "UK rules" is no longer a single answer for consumer chemistry. Northern Ireland follows EU law on fragrance allergen labelling (about 82 named substances against Great Britain's 26) and on the bisphenol ban in food contact. Great Britain has done neither yet.
- Dementia risk is largely the same list as heart risk. The 2024 Lancet Commission attributes around 45% of dementia to 14 modifiable factors, and the two largest are hearing loss and raised LDL in midlife. That is now a chapter in the heart topic, because that is where the factors already live.
A word about the longevity industry
- Nothing on this site is for sale, and nothing here carries an affiliate link. That is worth saying explicitly in this topic in particular, because it is the one where most of the writing you will find elsewhere is attached to a shop.
- Biological age tests, at present, mostly measure how well you did on the questionnaire. Treat a number that moves when you sleep well for a fortnight with suspicion.
- An intervention with a strong mechanism and no human outcome data is a hypothesis. Some of them will turn out to work. Most will not, and there is no way to know in advance which.
- If a claim depends on a study in mice, worms or cells, the plain translation is “this is why someone is running a human trial”, not “this works”.