Index
Every entry, A–Z
50 entries.
A
- Afternoon naps and sleep debt: what the experiments timedgoodBodyTen minutes of sleep in the nap, thirty minutes of sleep in the nap. The difference is a measured period of worse performance, and self-ratings do not notice it.
- Alcohol and longevity: what the cohorts say and what the genetics saymixedBodyTwo literatures, two shapes. The one that measures what people say they drink is J-shaped. The one that measures what their genes predispose them to is a straight line upwards. The gap between them is not a rounding error, it is the whole argument.
B
- Blood pressuresettledBodyThe most consequential modifiable number in ageing. Usually silent until it isn't.
- Blood pressure and dementia: what the pooled cohorts can and cannot separatemixedCardiovascularThe association is large, consistent and observational. The intervention evidence for dementia prevention is not the same study, and almost never quoted.
- Blue ZonesmixedMeasurementRegions reported to have exceptional longevity, and the data-quality problem underneath them.
- The boring basicssettledBehaviourThe short list that accounts for most of the modifiable risk. Nothing on it is interesting.
C
- Calorie restriction and ageing: what the human trials measured, and what they were built to avoidgoodBodyThe best human trial of calorie restriction exists, ran for two years, and did not measure lifespan. It measured the predictors — and the costs are in the same table.
- Chair-stand testgoodMeasurementFive chair rises and a stopwatch. Legs as a longevity readout.
- Circadian disruption and ageing: what the human studies supportmixedBrainThe mechanism is settled and the exposure definitions are not. A 2017 review of 26 studies says so in its own limitations section.
- Cognitive decline and the modifiable risksgoodBrainAround 45% of dementia risk is attributable to fourteen factors, most of them ordinary.
- Creatine and the aging brainmixedBrainA real but small memory effect, concentrated in the group that has been studied least.
D
- Daily step count and mortalitygoodBehaviourMore steps track with lower mortality. The round-number target is arbitrary; the gains start far below it.
E
- Exercise and agingsettledBehaviourThe intervention with the widest evidence base in this reference, by a distance.
F
- Falls and bone densitysettledBodyFracture risk is bone strength times falling. Most attention goes to the first half.
- FrailtygoodMeasurementA measurable loss of physiological reserve. Not a synonym for age.
G
- Gait speedgoodMeasurementThe cheapest survival predictor in geriatrics. Walk more, but do not train the test.
- Grip strengthgoodMeasurementA five-second test that predicts a lot. Train the body, not the grip.
- Grip strength and mortality: 38 cohorts, 1.9 million people, one number to carrygoodMeasurementA hand that squeezes a dynamometer hard is one of the few measures that predicts mortality as well as blood pressure does. Nobody has run the trial that would explain it.
- Growth hormone after fifty: what two randomised trials actually measuredmixedBodyTwo six-month trials, eighty men and twenty-one men, and a result that splits cleanly along the lines of what was combined with what.
H
- Hearing lossgoodBrainA midlife risk factor with a device-shaped solution that most people delay for years.
- Hearing loss and cognitive decline: what 50 cohorts and one randomised trial showmixedBrainThe association is among the largest in the sensory determinants of brain ageing. The intervention trial is the more important paper, and it is the one that came out flat.
I
- InflammagingthinBodyChronic low-grade inflammation with age. Well observed; poorly actionable so far.
L
- Lifespan vs. healthspangoodMeasurementYears alive versus years in good health. The gap between them is the thing worth shrinking.
M
- Meal timing and circadian rhythm: what the trials measured on a clockmixedBehaviourThe early eating window moves glucose and the clock. It does not move the microbiome, and the effect sizes are small.
- Menopause and longevity: what the cohorts actually showmixedBodyEarly menopause tracks with cardiovascular events. Whether it tracks with dying earlier is the part the cohorts disagree on.
N
- Nuts and mortalitygoodBehaviourA handful of nuts a day tracks with living longer. The nut probably gets too much credit.
O
- Omega-3 and ageing: what the trials measured when they measured telomeresmixedBodyCardiovascular outcomes move a little. Telomere outcomes do not move in the biggest trial, and the field's supporting meta-analyses cover 337 people.
- Oral health and longevity: what four million people in a meta-analysis do and do not showmixedBodyThe association is the largest pooled one in preventive medicine. The causal claim, tested with instruments that do not lie about confounding, is the weaker half.
P
- Protein in later lifegoodBehaviourEFSA's 0.83 g/kg is a floor for adults generally; older adults are usually advised higher.
- Purpose in life and mortality: what the cohorts actually showmixedBehaviourPurpose tracks mortality in every cohort that has measured it. Nobody has assigned it, so nobody knows whether it does anything.
R
- Resistance training and muscle aging: what the trials showgoodBodyStrength responds at any age. Mass and daily function respond less, and the safety reporting is thin.
S
- Sarcopenia (muscle loss with age)settledBodyThe default loss of muscle and strength with age. The most reversible thing on this list.
- Sauna bathing and the Finnish cohorts: what two decades of follow-up showmixedBodyFrequent sauna use tracks with lower mortality in Finnish cohorts. Observational only; no trial has tested prescribing it.
- Sedentary timemixedBehaviourAn exposure distinct from inactivity. Real, and less catastrophic than the headlines.
- Short sleep and mortality: what the cohorts actually associategoodBodyBelow seven hours, 1.14. Above nine hours, 1.34. The per-hour figure is 1.06. None of it is a randomised trial, and the authors of both analyses say so.
- Sleep and aginggoodBehaviourSeven to nine hours. Sleep changes shape with age; the requirement does not fall much.
- Sleep and glucose metabolism: what the trials actually measuredgoodBodyInsulin sensitivity falls when sleep is cut in controlled trials. The number is 11 to 20 percent over a week, or an HOMA-IR rise of 0.30 over six weeks.
- Sleep and immunity: what the inoculation experiments measuredgoodBodyThe cold studies are experiments with an infection delivered on purpose. The inflammatory studies are experiments with sleep cut on purpose. Both point the same way and neither measures disease.
- Sleep fragmentation vs sleep duration: which one correlates with whatmixedBehaviourFragmentation tracks with frailty, death and midlife cognition where duration does not. Nobody has run the experiment that would settle it.
- Smoking and alcoholsettledBehaviourStopping smoking is the single largest return available. Alcohol's safe level keeps shrinking.
- Social connection and isolationmixedBehaviourA well-replicated association with an unusually stubborn reverse-causation problem.
- Social connection and mortality: what 90 cohorts and 2.2 million people showmixedBehaviourThe association is the largest in the social determinants literature. The causal claim has never been tested in an experiment, and the weakest component — living alone — is the one everybody quotes.
- Strength training after sixtysettledBehaviourTwo sessions a week, progressive load. Nothing else does what this does.
T
- Telomere length and mortality: what the cohorts showgoodMeasurementShort telomeres track higher mortality in almost every cohort that has measured them, and the effect shrinks in the oldest old. Nobody has shown that changing them changes anything.
- Telomeres and longevity: the numbers from 121,749 people, and what they do not measuregoodBodyA 2018 meta-analysis of 25 studies, 121,749 individuals and 21,763 deaths: one standard-deviation shorter leukocyte telomere length carried a hazard ratio of 1.09 for all-cause mortality, and the shortest quarter 26 per cent higher than the longest. A 2017 individual-participant analysis of 12,199 adults found 1.23 for the shortest quintile. The association weakens above 80 years.
- The ten-second balance testgoodMeasurementTen seconds on one leg, no equipment — and a surprisingly strong signal about survival.
- Training load and fatigue: what the studies could and could not measurethinMeasurementThere is no blood test for overtraining. The best-reviewed candidate is a battery of measurements, and the review that mapped it found 39 studies out of 5,561.
U
- Ultra-processed food and mortality: what the pooled evidence actually gradesmixedBodyThe largest pooled look at industrialised food and death, printed with its own certainty ratings.
V
- Vitamin D in older adultsgoodBody15 µg a day; the case is bone and deficiency correction, not the longer list.
- VO2 max and cardiorespiratory fitnessgoodMeasurementAerobic capacity. Strongly predictive, genuinely trainable, rarely measured.