Longevity curriculum

Many biologists define aging as harmful processes that build up and make an organism steadily more likely to die; senescence names the related fall in survival and reproduction with age. The hallmarks-of-aging framework calls aging a progressive loss of physiological integrity but does not class it as a disease, while the SENS program treats it as one, so the line is still argued.

Healthspan is often estimated as health-adjusted life expectancy, which weights each year lived by health status. A 2024 analysis of WHO data for 183 countries found life expectancy rose faster than this measure from 2000 to 2019, so the years lived with disease or disability grew. A reanalysis of mouse diet studies also found that diets that extended life did not compress late-life frailty.

Average life expectancy rose sharply in the twentieth century, at first largely because deaths in infancy and childhood fell. A classic comparison of India, Mexico, Sweden and the US between 1900 and 1950 found that, despite very different early-life death rates, death risk after age 40 climbed at a similar pace in all of them. Whether human lifespan has a fixed ceiling is disputed; some analyses report the record age at death still rising.

James Fries proposed it in 1980; it needs illness rates to fall faster than death rates, not a fixed lifespan. Among University of Pennsylvania alumni, low-risk people reached minimal disability about 8 years later but died only about 4 years later than high-risk peers. Population data conflict: US surveys found disability falling faster than death rates in 1982 to 2004, but healthy life expectancy lagged lifespan globally in 2000 to 2019.

People of the same age differ widely in health and in how long they live, so researchers proposed biological age to reflect a person's physiological and functional state rather than elapsed time. It is estimated from functional tests, blood markers or DNA methylation clocks. The measures are not interchangeable: clocks trained on age, on mortality or on counts of cell division each encode different information.

In Holloszy's split, primary changes are seen as inevitable, while secondary ones are not and can be sped up by inactivity. Skin research draws a similar line between intrinsic aging on sun-protected skin and photoaging from ultraviolet light. One small study used osteoarthritis patients to model long-term inactivity and inflammation, and linked most age-related muscle gene changes to those factors, with about 200 genes tied to primary aging.

Rowe and Kahn's 1997 model of successful aging had three parts: low risk of disease and disability, high physical and cognitive function, and active engagement with life. In a review of 50 epidemiological studies, 19 applied all three, and many also measured psychological well-being, social participation or self-rated health. Critics argue that flourishing and frailty often occur together, which the successful-aging criteria cannot account for.

In more than 3,000 records of captive naked mole-rats, the hazard of death did not rise with age, even far past maturity. Hydra show no apparent senescence, and a study of zoo-kept turtles and tortoises found about three-quarters of 52 species aging slowly or negligibly. Some modeling suggests such reports underestimate the rise in mortality, so specific cases remain debated.

Genes, sex, infections, diet, activity, stress, socioeconomic status and access to care all shape how fast and in what form aging appears. Some changes are not one-way: in a Japanese cohort aged 65 to 70, about one in ten moved to a less frail status over five years, though about one in five got worse. Still, a marker moving back does not by itself show that aging reversed.

The actuary Benjamin Gompertz showed in 1825 that over much of adult life, age-specific death rates climb exponentially. Doubling time lets species be compared: one classic analysis put it near 220 days for house mice and 3,100 days for one 1969 sample of US women. The pattern does not fit every species, and at the oldest human ages apparent plateaus may partly reflect errors in age records.

Life expectancy is the average number of years remaining at a given age under current death rates. In the US, falling deaths from infection in the twentieth century cut infant and child mortality and raised life expectancy at birth; in developed countries, gains after the 1950s were tied to falling death rates in old age. Because the figure at 65 ignores deaths before 65, it tracks survival in later life.

A survival curve plots the share of a group still alive at each age. As premature deaths declined, more people lived to old age and deaths bunched into a narrower age range, giving the curve a squarer shape. Fries expected deaths to cluster around a fixed mean near 85, but a Swiss analysis from 1876 to 2006 found rectangularization and a rising mean age at death each explained about half of the rise in life expectancy.

Sustained low fertility shrinks the young base of the age pyramid, while rising longevity swells the older groups. Both raise the old-age dependency ratio, the people above an old-age threshold divided by the working-age population. One demographic analysis argues that a fixed threshold ignores longer lives; measuring old age by remaining life expectancy makes population aging look less burdensome.

Across 183 WHO member states, women lived longer than men on average but also had a larger gap between total and healthy life expectancy. A study of 22 European countries split women's extra unhealthy years into a survival effect and a health effect, and in most cases more than half came from women living longer. The health effect varied by measure and country.

A 2024 analysis split the US population into ten groups using race and ethnicity, county, metropolitan status, income and residential segregation. The life expectancy gap between the highest and lowest groups was about 13 years in 2000 and widened in the following two decades. Income did not map neatly onto the ranking: the group with the highest income in most years ranked only fourth or fifth before 2020.

Researchers checked US Social Security death records listing ages of 110 or more against a list of validated supercentenarians. The share of invalid claims rose with claimed age, from about 65 percent at 110 to 111 to every claim of 120 or more. The authors also sorted false claims into types, including clerical entry errors, pension fraud, national pride and family notoriety. Bad records above age 105 can also create false mortality plateaus.

In 1900 pneumonia, tuberculosis and diarrhea with enteritis led US causes of death, and about 30 percent of all deaths were in children under 5. By 1997 heart disease and cancers caused more than half of all deaths. Most people now live far beyond the ages at which they once would have died, and advancing age is the major risk factor for chronic diseases such as cancer, cardiovascular disease and dementia.

Predators, disease and accidents thin out every cohort, so a gene whose effects begin only late in life acts on fewer survivors and is weakly selected. Fisher and Haldane first saw this, Medawar and Williams developed it, and Hamilton formalized it mathematically in 1966. The force of selection often, though not always, declines with age, and that decline underlies the classic theories of aging.

If selection weakens with age, mutations whose harm appears only late can accumulate to higher frequency. Haldane pointed to Huntington's disease, with an average onset near age 35, as an example. Experiments mostly in fruit flies support the idea, but some findings fit poorly: yeast gathers few mutations with age, and mice engineered for high mutation rates did not always lose lifespan.

Because selection is strongest early in life, an allele's early benefit can outweigh its later cost. In one study of Tsimane forager-horticulturalist women, those with two copies of APOE e4, a variant tied to Alzheimer's disease and cardiovascular disease, had more children. Cellular senescence may be another case: it guards against cancer early but harms tissues as senescent cells pile up.

If predators or accidents make a long life unlikely, investing in repair that would pay off only later is a poor strategy, so selection favors enough upkeep to reach reproductive age and little more. One study linking reproductive investment to shorter life in dogs fits this logic. But mice given heavy free-radical damage did not always lose much lifespan, which challenges the role of damage.

Birds tend to age more slowly than mammals of similar size, a pattern linked to flight lowering predation, and protective traits such as shells, burrowing and tree living are tied to long life in land vertebrates, notably bats, birds and naked mole-rats. Fruit fly experiments with high or low adult mortality bore this out. Yet in guppies and one nematode species, higher mortality that favored better body condition led to longer life.

The old rate of living theory holds that lifetime energy use per gram of tissue is roughly constant, and scaling laws across mammal species fit it. Dogs break the pattern: large breeds live shorter lives than small ones, which one modeling study linked to the energy cost of growth. Single species can also sit far off the trend: the naked mole-rat's maximum lifespan is about five times that predicted for a 40-gram rodent.

August Weismann proposed that aging evolved to remove old individuals and free resources for the young, a benefit to the group. Selection is now understood to act on genes rather than on populations, and most theories treat aging as a non-adaptive by-product. Genes still shape longevity, apparently through upkeep and repair rather than a strict clock, though some researchers still argue for adaptive aging.

Lopez-Otin and colleagues set these three tests, so each hallmark needs interventions that can push aging in both directions. A 2026 review calls the tests sound but not enough unless aging and the outcome are clearly defined: a change that rises with age may drive aging, offset it or simply result from it, and worsening a process can cause disease without reproducing aging.

Primary hallmarks such as genomic instability and telomere attrition were described as clearly harmful. Antagonistic ones, such as deregulated nutrient sensing, senescence and mitochondrial dysfunction, were seen as beneficial at low levels but damaging at high levels. Stem cell exhaustion and altered intercellular communication were integrative. One 2026 review calls the split a heuristic, not a fixed causal order.

DNA damage can trigger senescence and inflammation, and inflammation can in turn impair stem cells, mitochondria and protein upkeep; damaged mitochondria can rouse immune responses by releasing their DNA. Such loops predict nonlinear results. The same review adds that aging varies so much between people that one pathway is unlikely to give uniform benefits, while indiscriminate drug combinations can add toxicity.

Age is the strongest shared risk factor for cardiovascular disease, type 2 diabetes, cancer and neurodegenerative disorders, so slowing the processes behind aging might postpone several conditions together rather than one at a time. That idea has moved some research toward shared mechanisms. So far the evidence supports effects on specific diseases and functions, not a demonstrated slowing of human aging.

Damage-centered views see aging as molecular damage that accumulates because repair and upkeep have limited capacity. The hyperfunction theory instead proposes that growth signaling continues past development into adulthood, which its proponents call a quasi-program. Against damage as the main driver, some mice tolerated heavy oxidative damage or high mutation rates without much shorter lives.

Agencies such as the FDA and EMA define treatment indications as diseases or their manifestations, and aging is treated as a major risk factor rather than a disease. The TAME trial of metformin, whose design the FDA approved, uses time to any of several age-related conditions or death as its main endpoint. Composite endpoints carry a risk: a drug could meet one by working on a single condition.

Macroautophagy is the cell's route for delivering its own contents to lysosomes to be broken down and recycled; dysbiosis refers to disturbed gut microbes, such as shifts in species and lost diversity, though no single aged microbiome has been found. A 2022 meeting summary had proposed other candidates, such as altered mechanical properties and splicing problems. A 2026 review notes that no finite list captures every age-related process.

Critics say the list offers no hierarchy for deciding which process to target first, and some doubt that molecular damage is the main driver. A 2026 review concludes the framework should guide testable hypotheses rather than serve as a checklist of drug targets. A better reading on one marker, such as an inflammatory measure or an epigenetic clock, does not show that disability or death was delayed.

Inside sources include replication mistakes, spontaneous chemical breakdown, reactive metabolites and oxidants released during inflammation; outside sources include ultraviolet and ionizing radiation and chemical agents. Ultraviolet light, for example, mainly links neighboring pyrimidine bases on the same DNA strand. Repair is extensive but never error-free, and somatic mutations build up with age.

Base excision repair fixes single damaged bases, such as oxidized ones; nucleotide excision repair removes bulky lesions spanning a stretch of DNA; mismatch repair corrects pairing errors left by replication. A double-strand break is either joined directly by non-homologous end joining or rebuilt by homologous recombination, which copies the intact sister chromatid.

Single-cell sequencing shows the number of mutations rising with age across many human tissues, and expanding mutant clones in blood and epithelial tissue show selection acting on that variation. Effects depend more on where a mutation lands and how its cell behaves than on the count. Many mutations are neutral, and the signaling that damage sets off may matter as much as the changed sequence.

Werner syndrome, from mutations in a gene for a DNA maintenance and repair protein, can bring graying, cataracts, diabetes, hardened arteries and cancers from early adulthood. Hutchinson-Gilford progeria, from a lamin A mutation, causes slowed growth in childhood and severe artery disease. Both are partial copies of aging; children with progeria do not get cataracts or osteoarthritis, for example.

Lamins form a mesh under the inner nuclear membrane that supports nuclear shape, chromatin organization, DNA repair and replication. The progeria mutation yields progerin, which leads to misshapen nuclei, loss of heterochromatin at the nuclear edge, DNA damage and early cell senescence. Progerin is also made in some cells of healthy older people, but its link to normal aging is a hypothesis.

Sequencing finds it in more than a fifth of people over 80, most often through mutations in genes such as DNMT3A and TET2. Carriers have a higher risk of blood cancer and, in epidemiological studies, of cardiovascular events and death. Definitions and detection thresholds vary between studies, some mutations show inconsistent links, and screening remains limited to research.

In humans the repeat is TTAGGG, and with a protein complex called shelterin the end folds into a loop. Without that protection, repair machinery would read the chromosome end as a broken strand and try to join it to other ends or trigger cell death. A telomere worn very short probably can no longer form its loop, and the exposed end then sets off a lasting damage response.

Copying of the lagging strand starts from RNA primers. When the last primer at the chromosome tip is removed, DNA polymerase cannot fill the gap, so a little sequence is lost each round. Erosion also comes from processing of the ends and oxidative damage, and the enzyme telomerase can add repeats back in cells that make it.

Human fibroblasts typically divide about 40 to 60 times, and Hayflick reported about 50, before entering a permanent arrest called replicative senescence. The limit reflects telomere shortening, because most human body cells make little telomerase. Telomeres also act as stress sensors: damage to them can halt division well before cells reach their replication limit.

It carries its own RNA template and a catalytic protein that writes new repeats onto chromosome ends. Most somatic cells have little or no telomerase, so their telomeres shorten as they divide. More than 85 percent of human cancers maintain their telomeres through telomerase, and the rest use a separate mechanism called alternative lengthening of telomeres.

In large groups, blood-cell telomere length is linked to several age-related outcomes, yet inherited differences are large: length at birth varies by more than average shortening over a whole adult life. Readings shift with measurement method and blood-cell mix, and repeat tests add error. One review concluded that telomere length does not meet the basic criteria for an aging biomarker.

About half of people with dyskeratosis congenita carry mutations in TERT, TERC, DKC1 or TINF2, genes that maintain telomeres. Fast-dividing cells suffer most, giving nail, skin and mouth changes, bone marrow failure, lung scarring and higher cancer risk. One review calls such disorders strong evidence that short telomeres can cause disease, though severity varies widely.

Short telomeres restrain proliferation and can suppress early tumor growth, yet telomeres that become dysfunctional can destabilize chromosomes and promote cancer, as in dyskeratosis congenita. Restoring telomerase might renew tissues but could also help malignant clones. A 2026 review says proposed telomerase activators should not be presented as established anti-aging treatments.

DNA methylation adds methyl groups directly to DNA bases, while histone modifications tag the spool proteins DNA wraps around and signal whether a region should be used. Dividing cells pass on much of this pattern, which helps them stay specialized. Marks shift during development and tissue specialization, and they can also change over a lifetime and with exposures or disease.

Aging tissue tends to show an overall loss of methylation alongside gains at particular sites. In one study of mouse spleen, weakly methylated regions such as gene promoters tended to gain and heavily methylated ones to lose, a shift toward the middle, and individual mice grew more different from one another. The authors read these changes as largely random failures of upkeep, which is an interpretation.

Horvath's 2013 clock used 353 sites and tracked chronological age closely across many tissues, and studies link a clock age above real age to illness and death. But readings shift with cell mix, ancestry and acute stress, and change after pregnancy or surgery, so a moved number does not show that aging moved. Consumer tests now sell such readings, at least one without fully disclosed performance data.

Clocks from Horvath and Hannum predict chronological age very accurately but capture health outcomes less well. Later models such as PhenoAge, GrimAge and DunedinPACE folded in clinical measures, mortality-linked proteins, smoking or change over time, and DunedinPACE reports a pace of aging rather than an age. Because they were trained on different targets, the clocks carry different information.

It exposes cells for a short time to the Yamanaka factors, OCT4, SOX2 and KLF4 with or without MYC, used to make induced pluripotent stem cells, aiming to stop short of erasing cell identity. In mice it has restored some functions, and in one study of old male mice it lengthened remaining lifespan. Risks include lost identity, dysplasia and teratomas, and no clinical study has shown systemic use safe or effective in people.

Heterochromatin is DNA packed tightly enough to keep it silent, and it holds many transposable elements, sequences that can copy themselves into new spots. Aging cells lose some of it, which can derepress those repeats and trigger immune signaling. In aging mouse tissues, several retrotransposon families began to be transcribed and in advanced age some moved, though their role in aging is a hypothesis.

Proteostasis is the upkeep of a cell's proteins. Chaperones help new or damaged proteins fold, while the ubiquitin-proteasome system and autophagy break down unneeded or damaged proteins and organelles. One review suggests the three arms can partly stand in for each other, which may explain why levels of their individual components vary so much in aging cells.

Stress activates the transcription factor HSF1, which switches on heat shock proteins such as Hsp70 that refold damaged proteins or route them to disposal. Cells from older people and many aged tissues show weaker Hsp70 induction, and in worms HSF1 supports both stress tolerance and lifespan. The decline is not universal: one mouse study found no age difference in brain chaperones, though heart levels fell.

In autophagy a cell encloses worn-out proteins or organelles and sends them to the lysosome to be broken down and reused. Disabling autophagy genes shortens lifespan in worms and flies, and chaperone-mediated autophagy declines in old rodents. The evidence is uneven: a review of muscle aging notes contradictory findings and that many studies did not measure flux, the actual rate of breakdown.

A relay of enzymes called E1, E2 and E3 attaches chains of the protein ubiquitin to a target, which the proteasome then cuts apart. Proteasome function declines with age in many mammalian tissues studied. Protein aggregates can also inactivate proteasomes, so a slowdown may feed itself: misfolded proteins pile up, which further blocks degradation.

In Alzheimer's disease, beta-amyloid clumps into plaques between neurons and tau forms tangles inside them; in Parkinson's disease, many brain cells contain Lewy bodies, clumps of alpha-synuclein. These diseases usually appear late in life, which one review ties partly to weakening protein-degradation systems. NIA notes that researchers are still sorting out which brain changes cause Alzheimer's and which result from it.

Rapamycin blocks the protein-building kinase mTORC1 and induces autophagy; calorie restriction and exercise lower mTORC1 activity and raise autophagy in several tissues. One review proposes the latter two may owe their longevity effects to this protein upkeep, though how much of rapamycin's effect runs through autophagy is unclear. Human data are thin: 15 long-term calorie restrictors showed higher autophagy gene activity than controls.

When unfolded proteins pile up in the endoplasmic reticulum, three sensors called PERK, IRE1 and ATF6 slow protein production, raise chaperone levels and speed removal of misfolded proteins. If the stress lasts, a branch running from PERK to the factor CHOP pushes the cell toward apoptosis. In worms, this response and several other stress responses have been shown to decline with age.

Insulin and IGF-1 bind receptors that activate a chain of kinases, PI3K then AKT, which supports growth and holds back FOXO transcription factors. Mutations that reduce this signaling lengthen life in C. elegans, fruit flies and mice. In people the link is less tidy: some studies found lower IGF-1 in centenarians than in other older adults, others measured higher levels, and both high and low levels are tied to insulin resistance.

mTOR responds to nutrients, growth factors and cellular energy, increasing protein synthesis and suppressing autophagy. In 2009, rapamycin fed to mice starting at 19 months of age extended lifespan in both sexes, and many later studies in different labs and mouse strains repeated the result. In healthy older adults, a 2025 review found the data have yet to show that rapamycin or similar drugs delay aging.

When ATP runs low and AMP rises, AMPK switches on: it slows protein and fat synthesis, boosts fuel-burning pathways, inhibits mTOR and promotes autophagy. Its activity appears to decline with age, and raising it has extended life in several animal models. More is not always better: persistent overactivation can be harmful, and metformin activated AMPK in fruit flies without extending their lives.

Sirtuins remove chemical tags from proteins and consume NAD+ as they work, so a fall in NAD+ with age was proposed to weaken them. A 2021 review of studies across species found evidence for a general NAD+ decline very limited and often based on one tissue. The precursor NMN raises NAD+ in cell and animal studies and appears generally safe in human trials so far, but a 2023 review noted few outcomes measured and excessive market hype.

Animals with chronically lower IGF-1 signaling may live longer, perhaps because their cells grow more slowly and accrue less damage. Larger dog breeds tend to have higher growth hormone and IGF-1 and generally shorter lives than small breeds, which a 2024 review calls a potential correlation. The trade-off cuts both ways: IGF-1 also drives muscle growth, and low levels have been linked, inconsistently, to low muscle mass.

Worms are cheap to grow, live about three weeks and carry counterparts of about two-thirds of human disease genes, which suits fast lifespan screens. The insulin/IGF-1 receptor gene behind long-lived worm mutants is conserved from yeast to mammals. Results do not always carry over: metformin extended life in worms but not in fruit flies, and one review blames the wider animal-to-human gap on species differences and redundant pathways.

Electrons from NADH and succinate pass along protein complexes in the inner mitochondrial membrane to oxygen, which is reduced to water. Complexes I, III and IV move protons across that membrane as this happens, and the resulting gradient drives ATP synthase. Mitochondria also carry their own small DNA, and 13 of its 37 genes encode enzymes for this process.

A small share of electrons leaks from the transport chain and partly reduces oxygen to superoxide and related molecules, which makes mitochondria the main source of these reactive species in cells. At physiological levels they act as signals in redox balance and stress responses, and raised levels switch on the energy sensor AMPK. Damage to proteins, lipids and DNA begins when production outruns antioxidant defenses.

Denham Harman proposed that reactive oxygen species from normal metabolism damage molecules and drive aging, implying antioxidants should help. NCCIH reports that large trials of vitamin E and beta-carotene supplements found no protection against cancer or heart disease, and beta-carotene raised lung cancer risk, most clearly in high-risk groups such as smokers. A 2022 review adds that antioxidants have not slowed aging in all animal models.

In hormesis, a small dose of a stressor triggers defenses that leave cells sturdier, while a large dose causes harm; the mitochondrial version is called mitohormesis. In a 2009 trial, high-dose vitamins C and E given to untrained young men blunted exercise-induced signals for new mitochondria. Most studies of these vitamins found no effect on performance and a few found a blunted training response.

Each cell holds many copies of mitochondrial DNA, so mutant and normal copies can coexist, a mix called heteroplasmy, and respiration suffers only once the mutant share passes a threshold. In colorectal lining from 207 people, low-level mutation frequency did not change significantly with age, but clonally expanded mutations rose sharply. In short-lived worms, deletions of mitochondrial DNA were very rare and showed little clonal expansion.

Mitochondria that lose their membrane potential recruit PINK1, which activates parkin to tag them for disposal by autophagy, a process called mitophagy. New mitochondria are made under the direction of the coactivator PGC-1alpha. In aging muscle, endurance training raises PGC-1alpha and mitochondrial content, though less than in young muscle, and some studies of very old rats saw no rescue.

A senescent cell has stopped dividing, with the p53-p21 and p16-RB pathways engaged, yet it does not die. It stays metabolically active, resists the programmed cell death that would remove it, and releases inflammatory and protein-degrading factors. Leonard Hayflick and Paul Moorhead described the state in the early 1960s, when cultured human cells stopped dividing after a limited number of cycles.

Critically short telomeres are read as DNA damage, setting off a response that halts the cell cycle, and in senescent cells that damage response persists. Oxidative stress and mitochondrial dysfunction can trigger the same arrest, and an activated cancer gene such as RAS causes oncogene-induced senescence. The trigger matters, since the mix of factors a senescent cell secretes differs with the trigger and the cell type.

The senescence-associated secretory phenotype, or SASP, includes cytokines such as IL-6, chemokines, growth factors and protein-cutting enzymes, and its makeup depends on the cell type, the trigger and the time since arrest. These signals can push neighboring cells into senescence. In mice, transplanted senescent cells caused the recipients' own cells to senesce, even at a distance.

Transient, programmed senescence helps in embryonic development, supports wound repair and limits excess scarring, after which immune cells clear the arrested cells. Halting damaged cells also stops them from passing on DNA damage, an early barrier to cancer. Harm comes when senescent cells linger, and a researcher quoted by NIA says it is not yet known how to remove the harmful ones while sparing the useful ones.

Senescent cells can recruit immune cells such as macrophages that remove them; in one liver-injury study, cells unable to senesce were not targeted. With age, the number of senescent cells in the body rises. One review points to weaker immune clearance, continued production of new senescent cells, or both, and NIA describes senescent cells accumulating as the aging immune system becomes less efficient.

Researchers look for halted DNA replication, raised p16 or p21, DNA damage foci at telomeres and activity of the lysosomal enzyme SA-beta-galactosidase. Each marker misses some senescent cells or flags other cells, and not all senescent cells are inflammatory. A 2024 review found no single marker that accurately measures senescent cell burden, and reliable noninvasive measurement in people remains a major hurdle.

In mice, senolytics such as dasatinib plus quercetin delayed several age-related conditions, and other studies reported a 36% longer average lifespan in treated older mice. The first human trial, in 14 people with lung fibrosis, was open-label with no placebo group, and a later pilot in 9 people showed fewer senescent cells in fat than before treatment. NIA says any safe benefit for people may be years or decades away.

Intestinal lining cells are born at the base of crypts and die at the villus tip about 4 to 5 days later, and blood is a regenerating tissue supplied by stem cells in the bone marrow. The heart and central nervous system show little or no self-renewal: human heart muscle cell turnover is estimated at about 1% a year at age 25, falling to about 0.45% a year at 75.

Adult stem cells replace cells lost to wear and injury. With age, muscle stem cells, called satellite cells, decline in number and in their capacity to divide and form muscle, and blood stem cells skew their output. Some researchers argue much of this is imposed from outside, since old stem cells regain function when their signals are made youthful, while another review credits both intrinsic and environmental causes.

Stem cells live in a niche, a surrounding microenvironment that helps decide whether they stay dormant or divide. In aged muscle, fibers display less of the Notch ligand Delta, so satellite cells activate poorly, and forcing Notch signaling in old mice restored regeneration. Extra FGF2 from old fibers pulls stem cells out of rest until the pool runs down, though some propose this rise compensates for a defect inside the stem cells.

In heterochronic parabiosis, a young and an old mouse are joined so they share circulation. Old partners showed better muscle, liver and brain stem-cell activity, while young partners declined. But the old mouse also gains young organs and a livelier partner, the rescue was partial, and later studies contested key reports crediting the blood factor GDF11. FDA says young-donor plasma has no proven benefit for aging.

FDA requires regenerative medicine products to be approved before they are marketed, with testing under its oversight in a clinical trial first. It reports clinics selling unapproved stem cell, cord blood and exosome products, with harms including blindness, tumors and infections, and says none is approved for conditions such as osteoarthritis or Alzheimer's. A trial-registry listing or FDA registration does not make a product legal.

Blood stem cells give rise to myeloid cells, such as neutrophils and monocytes, and lymphoid cells, such as T, B and NK cells. In old mice these stem cells favor the myeloid line and regenerate less well after transplant. With age, lymphoid production falls while myeloid output holds or rises; reviews associate declining stem-cell function with weaker immune responses and tie the myeloid bias to more myeloid leukemia in older people.

Acute inflammation fights pathogens and repairs injury, then should subside. In most older people a persistent low-level inflammatory state appears instead, with higher circulating IL-6 and TNF-alpha even without infection; researchers call it inflammaging. It is a significant risk factor for illness and death in older adults, though its proposed causes remain hypotheses and whether it drives conditions such as frailty is unclear.

The SASP of senescent cells has been proposed as a main origin of inflammaging. Debris left by poor cell clearance activates innate immunity, and failing mitochondrial quality control releases free mitochondrial DNA, which, given its bacterial ancestry, engages receptors like those that sense microbes. The gut barrier also weakens with age, and in obesity, visceral fat gathers macrophages and releases IL-6 and TNF-alpha.

The liver makes C-reactive protein in response to inflammation, and levels can rise with infection, autoimmune disease, smoking, obesity, insomnia or depression. A CRP result shows the amount of inflammation but not its cause or site. In one sample of healthy people, IL-6 was higher at older ages and gave fairly stable repeat readings, yet researchers note that a few cytokines cannot capture the complexity of inflammaging.

Higher inflammatory markers in older people go with greater risk of dementia, disability and frailty, and low-grade inflammation is also associated with cardiovascular disease, type 2 diabetes and cancer. Association is not proof of cause, but the CANTOS trial found that blocking the cytokine IL-1beta lowered cardiovascular events compared with placebo. For neurodegeneration, a causal role for cellular senescence is not yet established.

Ending inflammation is an active process: macrophages engulf dying neutrophils, a step called efferocytosis, and lipid mediators such as resolvins and lipoxins help wind the response down. Old mice resolved an experimental inflammation more slowly and had lower levels of these mediators, and aging was linked to less lipoxin A4 in the urine of healthy volunteers. Studies in people are still needed.

Studies report that older people's gut microbiota tends to be less diverse, with fewer anti-inflammatory groups such as Bifidobacterium and more potentially harmful ones, while the gut barrier weakens. In animal models, age-related dysbiosis raises gut permeability and systemic inflammation. The pattern is not uniform: some centenarian groups show higher diversity, and diet, medication and long-term care are linked to differences too.

The thymus is the organ where T cells mature. It starts to involute, losing mass and structure, as early as birth and no later than puberty, and its output of new naive T cells falls with it; in older adults, measured production of fresh T cells is minimal. How much of the weaker immunity of old age this decline directly causes is still poorly understood.

Naive T cells carry the variety needed for new targets; research suggests they typically recognize tumor neoantigens. Reviews describe the smaller, less diverse naive pool of old age as tilting T cells toward memory types and leaving holes in the repertoire, which may weaken control of viruses such as influenza. These links are proposed, not settled, and some researchers suggest the shift also helps older hosts.

With age, the germinal center response and its output of plasma cells weaken, so older adults reach lower antibody levels after vaccination and those levels fade faster. For people 65 and older, CDC prefers flu vaccines with a higher dose or an adjuvant, an added ingredient that strengthens the response. CDC reports that flu vaccination still lowers the risk of medical visits and hospital stays in older people.

Innate immune cells, including neutrophils, monocytes and macrophages, change with age. Reviews report that aging macrophages and monocytes engulf debris less efficiently, letting it pile up and feed chronic inflammation, while a larger share of monocytes take on an inflammatory profile. Aged neutrophils live longer and home in on targets less well; in old mice with flu, too many reached the lungs and worsened inflammation.

Cytomegalovirus (CMV), a common herpesvirus, is usually caught in childhood or adolescence and then stays latent, often for life, reactivating from time to time. The immune response to it shows memory inflation, a long-term expansion of CMV-specific T cells estimated to fill half or more of the CD8 memory pool. Whether CMV promotes heart disease, cancer or shorter life is still debated, and its effect on health span is unclear.

The immune system normally removes damaged or abnormal cells, and research suggests tumors often display new antigens that naive T cells recognize. Reviews propose that the shrinking naive T-cell pool may partly explain the higher cancer rates of old age. NCI explains that cancer-causing genetic changes most often build up slowly with age, and that the body clears cells with damaged DNA less well as people age.

Elastic fibers let the aorta stretch with each heartbeat and release that stored energy between beats, smoothing blood flow. With age, elastin fragments while collagen accumulates and becomes cross-linked, so the wall stiffens. Pressure waves then travel faster and reflect back during the beat itself, raising systolic pressure and pulse pressure, systolic minus diastolic. The aorta stiffens far more than peripheral arteries do.

The endothelium is a single layer of cells lining blood vessels, and one of the substances it releases, nitric oxide, widens them. When nitric oxide falls, as reviews describe with aging and with risk factors such as hypertension, diabetes, smoking and high cholesterol, vessels relax less and shift toward an inflamed, clot-prone state that favors atherosclerosis. One review places this among the first steps in arterial stiffening.

Atherosclerosis is the buildup of plaque, a mix of fat, cholesterol, calcium and other substances, in artery walls. Plaque often starts forming in childhood and worsens with age, and many people have no symptoms until a heart attack or similar event. NIA stresses that although atherosclerosis is common, it is not a normal part of aging, and NHLBI reports that it can sometimes be reversed with lifestyle changes and medicines.

Between beats, the heart's left side must relax to refill. With age its wall can thicken, and in heart failure with preserved ejection fraction it becomes too stiff to relax fully, so it fills with too little blood even though the muscle is not weak. NHLBI names high blood pressure as a main cause, with obesity and diabetes also causes, and NIA notes this form affects older adults with long-term hypertension and obesity.

Cardiac output, the blood the heart pumps per minute, is heart rate times stroke volume. NIA notes that the aging heart cannot beat as fast during exertion or stress, though resting rate changes little. A review of women masters athletes found maximal heart rate fell with age regardless of fitness, likely from pacemaker changes, and called this a significant factor in falling aerobic capacity; stroke-volume findings were mixed.

A stiff aorta cannot store pressure during each beat and release it between beats, so systolic pressure, the first number, climbs while diastolic pressure, the second, can fall. NIA calls the result, isolated systolic hypertension, the most common form of high blood pressure in older adults and traces it to stiffening of the major arteries. NIA adds that this stiffening can raise pressure even in people with heart-healthy habits.

NHLBI calls high blood pressure, unhealthy cholesterol and smoking key risk factors that can start plaque buildup, and says high blood pressure accounts for more US atherosclerosis deaths than any other changeable factor. Diabetes, inactivity and excess weight add risk, which climbs as factors stack up. In SPRINT, among adults 50 and older, a systolic target below 120 cut the risk of cardiovascular disease and death.

In insulin resistance, muscle, fat and liver cells respond poorly to insulin, so blood glucose can rise. NIDDK lists age 35 or older as a risk factor alongside excess weight and inactivity. One review calls the causes in older adults likely multifaceted, spanning age itself, less activity, inflammation and more body fat, and notes that losing muscle shrinks the main tissue that takes up glucose.

Subcutaneous fat lies under the skin; visceral fat lies inside the abdomen. One review describes aging as shifting fat toward visceral depots while lower-body subcutaneous fat declines, and links this redistribution to insulin resistance, diabetes and heart disease. Visceral fat may do harm partly by sending fatty acids and inflammatory signals to the liver. The review calls the link a correlation, which alone does not prove cause.

Ectopic fat is fat stored outside its normal sites, such as in the liver and inside muscle cells. Reviews describe aging fat tissue as storing lipid less well, so lipid spills into these organs and causes lipotoxicity. In muscle, poorly balanced handling of stored fat produces lipids such as ceramides and diacylglycerols that promote insulin resistance, though how much muscle fat drives the changes of aging muscle is not yet clear.

Advanced glycation end-products, or AGEs, form when sugars react with proteins, fats or DNA without an enzyme. They collect mainly on long-lived proteins such as collagen, where cross-links make it less flexible and harder to break down. The stiffness link is not simple: in 16 human Achilles tendons, collagen disorganization tracked mechanical behavior while AGE content showed no significant link.

Reviews describe total fat mass rising from middle age, body water falling, and muscle mass declining from early adulthood, with fat also moving into the abdomen and between muscles. Body weight can hide these shifts: in the Health ABC study of people in their 70s, weakness and fat infiltration of the thigh muscles progressed regardless of changes in body weight.

Metabolic flexibility describes how well the body shifts between fuels such as fat and glucose, and between storing and using them, as eating, fasting and exercise change demand. Reviews report inflexibility in obesity, insulin resistance and type 2 diabetes, and credit regular activity with improving flexibility. A 2025 review describes a decline with age, but a 2018 review noted that few studies had measured it in older people.

Sarcopenia means a progressive loss of muscle mass, strength and function with age. Definitions disagree: a 2019 European one pairs weak grip with low lean mass, while another consortium uses weak grip and slow walking with no mass test. In about 9,000 older white men, the two definitions flagged few people and agreed poorly, yet each was linked to higher mortality.

Muscle size reflects a balance between building and breaking down protein. In anabolic resistance, older muscle shows a smaller building response to exercise and dietary protein, and some studies suggest older adults need larger protein doses for a response like younger adults. Separating age itself is hard, since inactivity, low protein intake and insulin resistance also impair the response, while evidence for inflammation is mixed.

A motor unit is one motor neuron plus the muscle fibers it controls. With age, motor units are lost, and surviving neurons can sprout to reinnervate orphaned fibers, enlarging their units. One study of men found fewer motor units in every older group, but only men with sarcopenia lacked enlarged units, suggesting failed rescue marks the condition. Fiber shrinkage mainly affects type II fibers.

Bone is living tissue that the body continually breaks down and replaces, and osteoporosis develops when more is broken down than replaced. With age, bone loss speeds up and new bone forms more slowly. Women, who reach a lower peak bone mass than men, may lose bone quickly for several years at menopause, and low estrogen afterward is a listed risk factor. By 65 or 70, men and women lose bone at the same rate.

For people with weakened bones, a fall is more likely to cause a break, and CDC attributes most hip-fracture hospital visits to falls. Outcomes are poor: one review puts one-year mortality after hip fracture at roughly 15 to 30 percent, with fewer than half of survivors regaining prior function. Researchers increasingly see the fracture as reflecting frailty that was already there, as well as speeding it.

Joint cartilage is about 70 to 80 percent water, which keeps it resilient. MRI studies show knee cartilage thinning with age, and changes in aggrecan, a water-attracting matrix molecule, plus collagen cross-linking leave it less resilient and more brittle. Osteoarthritis is common in old age but not inevitable. Tendon studies conflict: some find aged tendons stiffer, but most report unchanged or reduced stiffness.

In untrained people, even short spells of cutting daily steps cause anabolic resistance, insulin resistance and losses of muscle, strength and fitness. Younger people regain what they lost when activity resumes, older adults less so, and bed-rest studies show faster muscle loss and more variable recovery in older people. One proposed model sees sarcopenia as drops after illness or injury with incomplete recovery, not a steady slide.

NIA lists slower recall of words and names, more trouble multitasking and mildly reduced attention as common in later life, while many studies find older adults have larger vocabularies than younger ones. Older adults may lag on complex memory or learning tests, yet given enough time to learn a task they usually do as well. These changes are not dementia, which disrupts daily life.

Raymond Cattell and John Horn split intelligence into fluid ability, solving novel problems, and crystallized knowledge, a person's store of general knowledge. Cross-sectional studies consistently find processing speed and fluid reasoning declining from early adulthood while verbal knowledge holds much longer. Yet crystallized knowledge also falls after about 70 to 75, and one analysis calls the two-way split too simple.

Cognitive reserve was proposed to explain why the same amount of brain disease causes mild symptoms in one person and severe ones in another. Education, occupation and literacy are used as proxies for it. A 2019 expert consensus warns that educated people may just test better and live healthier, which muddies the proxies, and that the biology behind reserve remains poorly understood.

In mild cognitive impairment, or MCI, people have more memory problems than others their age but can still care for themselves and do their usual activities. MCI may be an early sign of Alzheimer's disease, yet it does not always progress: in a Japanese community study of people 65 and older, about 31 percent with MCI had normal cognition five years later, though some who reverted later returned to MCI.

Dementia is a loss of thinking, memory and reasoning severe enough to disrupt daily life. Alzheimer's involves amyloid plaques and tau tangles; Lewy body dementia, deposits of the protein alpha-synuclein; frontotemporal dementia, abnormal tau or TDP-43, often before age 60; and vascular dementia, damage to the brain's blood vessels or blood flow. Autopsy studies of older adults with dementia found many had mixed changes.

In Alzheimer's, beta-amyloid clumps into plaques between neurons and tau forms tangles inside them. The amyloid cascade hypothesis, proposed in 1992, makes amyloid the initial trigger, and supporters cite genetic evidence for it. Two plaque-clearing antibodies, lecanemab and donanemab, are FDA-approved and slowed decline in some people, but critics question how meaningful the benefit is and whether one linear cause fits.

Observational studies link high blood pressure in middle age, diabetes and smoking to a higher later risk of dementia, though NIA cautions that such studies cannot prove cause. A narrower result comes from a study related to the SPRINT trial, where lowering systolic pressure below 120 reduced the risk of mild cognitive impairment. Vascular dementia itself arises when the brain's blood vessels or blood flow are damaged.

NIA reports that older adults can still learn new skills, form new memories and build vocabulary, and that evidence is growing that the brain keeps its ability to adapt. What practice transfers to is another matter: in a three-month randomized trial of 103 healthy older adults, a commercial brain-training app improved the trained tasks but not untrained tests of memory, speed, attention or language.

The lens, an inner part of the eye, helps focus light on the retina. With age it becomes harder and less flexible, so near objects blur. NEI calls this presbyopia a normal part of aging that everyone gets, usually after 45, and says it usually stops worsening after 65. It is not the same as farsightedness, a separate focusing error that also blurs near vision.

A cataract is a cloudy area in the lens that forms as lens proteins break down and clump, a process that starts around age 40; by 80, most Americans have one or have had it removed, and surgery corrects the vision problems it causes. Age-related macular degeneration damages the macula, the part of the retina for sharp central vision, and is a leading cause of vision loss in older adults, though it does not cause total blindness.

Presbycusis develops gradually and usually affects both ears equally, so people may not notice it. A review of cochlear aging describes damage to the outer hair cells that amplify sound, to the tissue that keeps the cochlea electrically charged, and to auditory neurons, which produces high-frequency loss and likely trouble following speech in noise. Noise exposure is a major cause, yet aged cells degenerate even without it.

NIA reports that older adults with hearing loss have a higher risk of dementia and faster declines in memory and concentration than peers who hear normally, and pooled studies link hearing aids and implants to lower risk of long-term decline. These are associations. In the ACHIEVE randomized trial, a hearing intervention showed no effect on cognition overall, though a prespecified analysis found slower decline in a higher-risk group.

Smell problems increase with age; in one study nearly a quarter of men aged 60 to 69 had one. Without smell, food tastes bland, impaired taste can change eating habits, and taste buds may start to decline after 50. Smell loss is also studied as an early marker of Alzheimer's and Parkinson's disease, but it has many other causes, from respiratory infections to medications.

With age the skin's outer layer thins and loses elasticity, and its glands may make less sweat and oil. CDC notes older adults adjust less well to sudden temperature changes and are more prone to heat illness, partly through conditions and medicines that alter heat responses or sweating. For cold, a review proposes less fat under the skin, a lower metabolic rate and weaker constriction of peripheral blood vessels as reasons.

Kidney filtration declines with normal aging, and one review reports lower liver mass, reduced liver blood flow and less activity of drug-metabolizing enzymes in older adults, so many medicines clear more slowly and can accumulate. Separating aging itself from damage by high blood pressure, diabetes and smoking is hard, and those diseases can use up the kidney's functional reserve.

NCI calls advancing age the most important risk factor for cancer overall and for many single cancer types, and overall incidence climbs steadily with age. Its genetics guidance explains that DNA changes from random mistakes or carcinogens build up over many years, and most cancers arise by chance from that accumulation. Still, cancer can occur at any age: bone cancer is diagnosed most often in children and teens.

NCI notes that scientists think a healthy cell needs more than one DNA change to turn cancerous. Armitage and Doll first linked the age pattern of certain cancers to the number of such cellular events, and later models still fit stage numbers to registry data. The fit has limits: for most cancers, incidence growth slows with age and levels off around 85 to 90, a late plateau or decline that some models fail to capture.

Tumor suppressor pathways such as p53 respond to DNA damage with repair, cell death or senescence, a permanent halt to division in cells at risk of turning cancerous. The authors of a 2008 study argued this fits antagonistic pleiotropy: stopping risky cells helps early in life, while senescent cells that accumulate may impair tissue repair and contribute to age-related disease. That cost rests on indirect evidence.

More cells and longer lives should mean more cancer, yet across species cancer risk does not track body size: Peto's paradox. A 2016 study found elephants carry 20 copies of TP53 and their cells readily self-destruct after DNA damage, though a 2022 reanalysis concluded most copies spread by drift, not selection. No evidence shows whales evolved extra tumor suppressor copies, so solutions likely differ by species.

NCI calls tobacco a leading cause of cancer and cancer death, ties obesity to at least 13 cancer types, reports strong evidence that alcohol can cause cancer, and says UV damage can lead to skin cancer. High-risk HPV and hepatitis B and C infections cause some cancers, and CDC notes HPV and hepatitis B vaccines can help prevent cancer. NCI cautions that population studies on their own cannot prove an exposure causes cancer.

Survival is counted from diagnosis, so finding a cancer earlier lengthens measured survival even if the person dies when they would have anyway; NCI calls this lead-time bias. Overdiagnosis, finding slow cancers that would never have caused harm, also inflates survival. So NCI says the only reliable way to know whether a screening test cuts cancer deaths is a randomized trial showing fewer deaths in the screened group.

Fried and colleagues defined frailty as having three or more of five criteria: low grip strength, slow walking, exhaustion, low activity and unintentional weight loss. One or two criteria mark prefrailty, a higher risk of becoming frail. A 2023 review calls this the recommended international standard. Its limit is that it scores outcomes rather than causes, because the biology driving frailty is still poorly understood.

Rockwood and colleagues measure frailty by counting deficits, usually 40 or more symptoms, signs and diseases, rather than checking five fixed criteria. Both this index and the phenotype predict outcomes such as hospitalization and death. A 2024 review of 415 trials found the phenotype the most common tool for choosing participants and as an outcome, while the index led among trials testing whether frailty changed a treatment's effect.

Aging is associated with a reduced capacity to repair and regenerate tissue, and the resulting loss of reserve under stress is called homeostenosis. In frailty research, a minor illness or fall knocks function down, and recovery stalls below the old baseline. Repeated drops can cascade into lost independence and higher risk of disability and death. Reviewers present this as a conceptual model of how frailty develops.

NIA lists many contributors to falls in older adults: duller vision and reflexes, sarcopenia, balance and gait problems, blood pressure that drops on standing, painful feet, certain medicines and household hazards. It adds that the more medications a person takes, the more likely a fall. NIA also notes that a fall can alert a doctor to a new medical problem or to medication or eyesight issues that can be corrected.

NIA calls polypharmacy, using many drugs at once, a growing concern: people 65 and older tend to take more medicines than any other age group, and more drugs can raise the risk of side effects. A prescribing cascade starts when a new drug is given to treat an adverse reaction to another. Deprescribing, safely reducing or stopping medicines that are potentially inappropriate or unnecessary, is an active research area.

MedlinePlus describes delirium as confusion that starts over hours or days, comes and goes, and mainly disrupts attention and alertness. Infections, medicines, dehydration, surgery and hospital stays are common causes, and older age and existing dementia raise the risk. Dementia instead begins slowly, usually with memory loss, and almost never improves. The two can coexist, and treating the cause of delirium often brings full recovery.

Basic activities of daily living are self-care and mobility tasks, and they tend to decline more slowly than instrumental ones. In a nine-year Brazilian cohort of community-dwelling older adults, slow walking speed predicted later dependence in basic activities, but low grip strength did not. In UK Biobank, stronger grip went with lower mortality after accounting for fitness. Both are observational, showing prediction, not cause.

A 2023 meta-analysis of 94 cohorts with over 30 million people found the biggest risk differences between no activity and about the recommended 150 minutes a week of moderate-to-vigorous activity, smaller ones up to roughly double that, and small, uncertain ones beyond. The Physical Activity Guidelines likewise say any amount of activity brings some benefit. These are associations from cohort studies, not trial results.

In a retrospective study of 122,007 patients given treadmill tests, low fitness carried an adjusted mortality risk comparable to or greater than smoking or coronary disease, and the fittest group lived longest, with no upper limit of benefit seen. The authors stress the design cannot prove causation: fitness may partly mark people already headed for lower mortality, and patients referred for testing may not represent the public.

Among 70,913 UK Biobank adults aged 40 to 69, the strongest grip group had about a fifth lower risk of death than the weakest after adjusting for aerobic fitness, and fitness predicted risk after adjusting for grip. Those high in both had roughly half the mortality of those low in both. Median follow-up was 5.7 years and the cohort was observational, so this shows prediction, not cause.

The Physical Activity Guidelines tell adults to move more and sit less, citing a strong relationship between more sedentary time and heart disease, high blood pressure and all-cause mortality. In 73,729 UK Biobank adults wearing wrist monitors, activity of any intensity weakened the link between sitting and death, and high sitting time predicted cardiovascular death only in those with little moderate-to-vigorous activity.

The Physical Activity Guidelines state that activity lowers older adults' risk of falls and fall injuries, and CDC lists balance work alongside aerobic and strength activity for people 65 and up. A 2023 systematic review of trials in community-dwelling older adults found balance training cut injurious falls, and multicomponent training cut falls, fallers and fall-related emergency visits. It did not pool results.

A 2021 review argues one stimulus reaches many hallmarks: training increases muscle mitochondrial content, raises muscle autophagy markers in older women, lowers some pro-inflammatory cytokines and increases blood-forming stem cells. The authors propose the energy sensor AMPK as a shared hub. Several cited results come from mice, and the review notes human autophagy data are still scarce.

A Swedish cohort of 427 people tested from age 16 to 63 found estimated aerobic capacity and muscular endurance peaked between 26 and 36, then fell about 0.3 to 0.6 percent a year at first and 2.0 to 2.5 percent later. Exercisers tend to have higher capacity than non-exercisers of the same age. In masters endurance athletes, decline per decade ranged from 5 to 46 percent and closely tracked changes in training volume.

Reverse causation is the worry that illness already headed toward death lowers activity first, making inactivity look more harmful than it is. A 2023 meta-analysis found that studies keeping people with existing disease, or counting deaths in the first years of follow-up, tended to show stronger associations, which the authors say may partly reflect reverse causality. They excluded studies with under three years of follow-up to reduce it.

NIA defines calorie restriction as cutting habitual intake without depriving the body of essential nutrients. In many animal studies, rodents given 10 to 40 percent fewer calories lived longer and had fewer diseases, especially cancers, and a fasting diet lengthened worm lifespan by 40 percent. But some experiments found no benefit, some mouse strains lived shorter lives, and why restriction works in animals remains unknown.

Two rhesus monkey studies of calorie restriction began in the late 1980s: one at the University of Wisconsin reported better survival, while one at NIA found no significant survival effect. A 2017 joint analysis noted the NIA controls were not free-fed and the Wisconsin diet was semi-purified, with six times the sucrose and twice the fat. Both saw fewer age-related diseases, but longevity was not a primary outcome at either site.

In CALERIE, 218 young and middle-aged adults of normal or moderately high weight were randomized, and the restriction group aimed to cut calories by 25 percent for two years; it managed about 12 percent. It then had lower blood pressure, cholesterol and some inflammatory markers than controls. These are risk markers, not outcomes: NIA notes no human data link calorie restriction to longevity, and two years cannot show long-term effects.

NIA describes several fasting schedules: eating within a set daily window, alternating normal and near-zero days, the 5:2 pattern, and several restricted days in a row once a month. Fasting may cut calories just by leaving less time to eat. A 2024 scoping review of 30 trial reports judged intermittent fasting and calorie restriction about equally effective on cardiometabolic, cancer and cognitive outcomes, but called for more head-to-head work.

Restricting protein generally lowers age-related disease in animal studies; human data are limited. In one analysis of US survey data, high protein intake at ages 50 to 65 went with higher overall and cancer mortality, but over 65 with 28 percent lower mortality; reviewers speculate protein may help older adults keep weight and muscle. Across all adults over 50, it found no link between more protein and higher mortality.

A 2008 meta-analysis of prospective cohort studies found that each two-point rise in a Mediterranean diet adherence score went with about 9 percent lower overall and cardiovascular mortality and 6 percent lower cancer incidence or death; the overall figure came from eight cohorts of about 515,000 people. The authors note the diet is defined differently across studies, scores are partly subjective, and residual confounding is possible.

In the two-year CALERIE trial, restriction caused slight declines in bone density, lean mass and aerobic capacity, roughly in line with the weight lost, and some participants had brief anemia, so NIA advises clinical monitoring. Trials have typically enrolled adults 60 or younger, so results may not apply to older people, and a review warns that chronic restriction might impair immune function and wound healing.

NIA notes that older adults need about the same seven to nine hours as other adults but tend to go to bed and wake earlier, spend less time in each sleep stage, and wake more often. Slow-wave deep sleep declines with age, especially after about 30. Sleep disorders such as insomnia, sleep apnea and restless legs also become more common with age, and illness, pain and medicines can disturb sleep.

The glymphatic hypothesis holds that fluid moving along spaces around blood vessels flushes waste such as amyloid-beta from the brain. In mice, fluid inflow was about 90 percent lower in awake animals than in sedated ones, and sleeping mice showed high flow too. In people, amyloid-beta levels rise during waking and fall in sleep. Yet reviewers call the hypothesis debated, and which molecules it clears is contested.

Nearly every human tissue keeps its own daily rhythm, coordinated by a master clock in the brain that responds to light. Night shift workers face light and meals at the wrong biological time and have more metabolic and heart disorders, though they often eat differently too. In small, tightly controlled lab studies, simulated night work raised 24-hour blood pressure and inflammatory markers such as CRP and IL-6.

A 2025 meta-analysis of 79 cohort studies found that, compared with seven to eight hours, sleeping under seven hours was linked to about 14 percent higher mortality and nine hours or more to about 34 percent higher. The authors suggest long sleep may signal undiagnosed illness, and note that many studies relied on self-reported sleep and residual confounding is possible, so the curve does not show that sleep length causes death.

In acute stress, the HPA axis raises cortisol within about 15 to 20 minutes, and cortisol then damps the signals that triggered it. Under chronic or repeated stress, cortisol can become dysregulated, and the cumulative wear across body systems is called allostatic load. In one Scottish survey of 4,488 adults, a higher allostatic load score predicted death at 10 years but not at 5, and predicted no specific cause.

In obstructive sleep apnea the upper airway closes many times a night, cutting airflow and oxygen. NHLBI explains that low oxygen, poor sleep and the effort of restarting breathing can damage organs and blood vessels and may raise the risk of atrial fibrillation, high blood pressure, heart attack, stroke, type 2 diabetes and dementia in older adults. Risk rises with age, and with age fatty tissue can build up in the neck and tongue.

A 2010 meta-analysis of 148 studies and over 300,000 people found stronger social relationships tied to 50 percent higher odds of survival over an average of 7.5 years, and the authors judged the effect comparable to smoking. NIA likewise links isolation and loneliness to heart disease, depression and cognitive decline. Causality is hard to establish, because no one can be randomly assigned to be isolated.

NIA, drawing on John Cacioppo's research, defines social isolation as physical separation from other people and loneliness as the distressing feeling of being alone. Many older adults who live alone are neither lonely nor isolated, while some people feel lonely among family and friends. Twin data suggest each is an independent risk factor, but whether loneliness is the path by which isolation harms health remains an open question.

Among 6,985 Health and Retirement Study participants over 50, those with the lowest purpose scores had about 2.4 times the hazard of death of those with the highest, after adjusting for health and other psychological measures. The link held, slightly weaker, when first-year deaths were not counted, a check for reverse causation. Proposed routes such as lower inflammation are unproven, and raising purpose is untested.

NIA notes that some older adults with depression show emotional numbness or lost interest rather than obvious sadness, and may be less open about their feelings. Symptoms can include slowed movement or speech and trouble concentrating or remembering, and in some cultures depression may show up mainly as physical aches. Dementia can produce some of the same symptoms, and depression can be an early warning sign of possible dementia.

A 2020 systematic review of 422 studies with more than 7 million participants found ageism tied to significantly worse health in nearly all of them, across 45 countries. It counted structural harms, such as denied care and exclusion from clinical trials, and individual ones, such as older adults absorbing their culture's negative views of aging. Its headline figures are shares of studies with significant findings, not effect sizes.

The theory proposes that perceived time horizons, not age itself, steer priorities: when the future feels limited, people favor present emotional goals over gaining knowledge. It is often used to explain the positivity effect, older adults' greater attention to positive information. That effect is debated: one eye-tracking study found no overall age gap, though older adults looked longer at the more positive face late in each trial.

CDC reports that dementia caregivers face greater risk of anxiety, depression and poorer quality of life than other caregivers, and NIA notes caregivers are less likely to get preventive care and may face higher health risks, even premature death. Yet a 2022 review of population studies found informal caregivers had lower mortality than non-caregivers, possibly because healthier people take on caregiving.

Baker and Sprott defined a biomarker of aging as a measure that, alone or in a composite, predicts functional capacity late in life better than chronological age does. After 15 years of funded research, AFAR concluded none had been found. In one Scottish study, telomere length added predictive power to age but fell short of the strict criteria, while composites of several measures beat age alone. No consensus criteria exist.

First-generation epigenetic clocks such as Horvath's were trained to predict chronological age from DNA methylation. Second-generation clocks such as PhenoAge and GrimAge were trained on health phenotypes or time to death, and GrimAge partly uses methylation stand-ins for smoking history and plasma protein levels. A 2026 review says many such tools remain research measures rather than established instruments for routine clinical care.

FDA describes a surrogate endpoint as a stand-in, such as blood pressure, expected to predict how patients feel, function or survive, and requires extensive evidence before one replaces a clinical outcome. History shows why: after heart attacks, drugs that suppressed irregular beats raised sudden death in the CAST trial, and fluoride raised bone density while fractures rose. Raising HDL cholesterol did not reliably cut heart events.

The NIA and Wisconsin monkey studies kept animals on restricted diets for over 20 years, yet longevity was not a primary outcome at either site. NIA reports the longest human calorie restriction trial lasted two years and most relevant studies used lab animals. Regulators turn to surrogate markers when outcomes take very long to study, and FTC guidance says animal and lab studies alone cannot back a health claim without human trials.

A 2019 scoping review found reported animal-to-human translation rates from 0 to 100 percent, suggesting success may be unpredictable. In mice, calorie restriction shortened lifespan in some strains, and NIA's Interventions Testing Program uses genetically mixed mice at three sites, where several compounds helped only one sex. Studies also differ in whether control animals eat freely, one proposed reason two monkey studies disagreed.

In NIA's mouse testing program, rapamycin raised median and maximum lifespan in both sexes and acarbose also extended lifespan, while metformin alone did not at the dose tested. A 2021 critical review of metformin judged the evidence that it extends lifespan controversial and noted human trials such as TAME and MILES were designed to test anti-aging effects. These remain research questions, not established treatments for aging.

Under DSHEA, FDA cannot approve supplements before sale. A label may make structure or function claims, such as calcium building strong bones, if the maker has substantiation and notifies FDA, but it must state that FDA has not evaluated the claim and that the product is not meant to diagnose, treat, cure or prevent disease. FTC adds that the disclaimer cannot rescue a deceptive ad.

FDA lists tip-offs to health fraud: one product that cures many conditions, personal success stories, quick fixes, all-natural claims, miracle cures or secret ingredients, and conspiracy talk. It also warns that clinics often illegally market unapproved stem cell and exosome products, and it has received reports of blindness, tumors and infections. Separately, a 2026 review notes many biological age clocks remain research measures.

Disability-free life expectancy is the number of years from a given age spent free of disability, and healthy life expectancy counts years in good health. WHO's health-adjusted version converts a population's mix of health states into equivalent years of full health, using life tables, how common each state is, and how it is valued. Methods matter: a 2022 review pooled only comparable studies and set DALY results apart.

Most self-reported claims of extreme old age are false, so demographers validate ages by cross-checking independent records; a 2019 paper found records above age 105 often wrong in ways that may distort mortality curves. Critics have questioned Blue Zone ages, and defenders say they were validated this way. A 2026 scoping review found Okinawa, Ogliastra in Sardinia and Nicoya well characterized, Menorca inconclusive and Loma Linda unstudied.