Factors
What affects longevity?
Longevity research draws on very different kinds of evidence: randomised trials, long-running cohort studies, and modelling. An association found in a population does not by itself establish cause, and no figure here can be converted into a personal number of years.
Each factor below sets out what the evidence shows, what is reasonably well established, and what is still uncertain.
- Physical activityOne of the most consistently studied influences on mortality risk, with a steep benefit at the low end of activity.What the research consistently findsThe steepest difference is between almost no activity and a modest amount, across many countries and measurement methods.
- FitnessMeasured fitness is one of the strongest single predictors of mortality risk in the research literature.Why it stands outThe gradient between the least and most fit groups is steeper than for most single risk factors — but fitness is partly inherited.
- SmokingThe clearest and best-quantified longevity factor in the literature, with strong evidence that stopping reverses much of the risk.What is firmly establishedRisk rises with the amount smoked and falls after stopping, reproduced across countries and decades with understood mechanisms.
- SleepBoth short and long sleep durations are associated with higher mortality, but the direction of cause is genuinely unclear.What remains uncertainMost of the evidence rests on sleep duration reported once, and long sleep may be a symptom of illness rather than a cause of death.
- StrengthGrip strength and muscle-strengthening activity are both associated with lower mortality, with a smaller evidence base than aerobic activity.What the research findsLower handgrip strength predicts mortality independently of muscle mass; the evidence base is smaller than the aerobic literature.
- DietDietary patterns are consistently associated with mortality, but individual nutrient claims are far weaker than they sound.Where claims get weakOverall dietary patterns show the most consistent associations; published 'years gained' figures come from modelling, not measurement.
What these factors have in common
Risk is not destiny
These studies describe how groups differ, not what will happen to one person. The lowest-risk point in a population curve describes a group, not an individual.
The biggest claims need the strongest evidence
An association with mortality is not the same as proof that changing a behaviour extends life. Where trials exist, they measure disease events rather than lifespan.
These factors overlap
Activity, fitness and strength are measured separately but move together, and each is partly a marker of underlying health rather than an isolated switch.
How we know what we know
Different questions are answered by different study designs. Most of what is known here comes from following large populations over time, which shows who tends to live longer rather than what happens if one person changes. The table below describes the designs and findings behind each factor, taken from the sources listed on its page.
Swipe the table sideways to see all columns.
| Factor | Kind of evidence |
|---|---|
| Physical activity | Observational cohort studies · Both self-reported and device-measured activity · Association with mortality · Causation not established |
| Fitness | Observational cohort studies · Directly measured exposure (graded treadmill testing) · Association with mortality · Causation not established |
| Smoking | Long-running observational cohort studies · Dose–response observed · Association with mortality reproduced across countries and decades · Understood biological mechanisms |
| Sleep | Observational cohort studies · Self-reported sleep duration · U-shaped association with mortality · Direction of cause unresolved |
| Strength | Observational cohort studies · Association with mortality, independent of muscle mass · Evidence for lifespan extension is not established |
| Diet | Observational cohort studies with self-reported intake · Randomised trial evidence for a dietary pattern, for cardiovascular events · Association with mortality for overall patterns · Single-nutrient and supplement claims not supported by trials · 'Years gained' figures come from modelling, not measurement |
Descriptors summarise the study designs and findings in the sources cited on each factor page. They are descriptions, not grades or rankings.
Continue exploring
- How SnapLongevity judges evidenceThe standards behind every claim on these pages.
- Can humans actually extend lifespan?What is established, and what is still speculation.
- Healthspan by countryHow many years are expected to be lived in good health.
- Years ahead explorerPublished life-table figures for an age, sex and country.