Concepts & Frameworks

Absolute risk (AR) is the probability that an individual experiences an event — such as a heart attack, cancer diagnosis, or death — within a defined time period. Relative risk (RR) expresses that probability as a ratio compared with a reference group; relative risk reduction (RRR) is the proportional fall in event rate between treated and untreated groups. Because RRR is independent of baseline risk, a drug reducing cardiovascular events from 2% to 1% and one reducing them from 40% to 20% both yield a 50% RRR, yet their absolute risk reductions (ARR) are 1 and 20 percentage points respectively — giving numbers needed to treat (NNT, the inverse of ARR) of 100 versus 5. In longevity and preventive medicine, where baseline event rates are low in healthy middle-aged adults, interventions framed as cutting risk by 30–40% frequently reduce absolute risk by fewer than 2 percentage points over five years. A structured review of 222 articles in six leading journals — including BMJ, NEJM, and JAMA — found that 68% omitted underlying absolute risks alongside ratio measures (Schwartz et al. 2006), inflating perceived treatment benefit. Sound appraisal of any longevity intervention requires ARR and NNT alongside RRR; when baseline risk is low, even a large RRR may reflect a small absolute gain.

Allostatic load is the cumulative physiological cost of adapting to chronic stressors, representing wear and tear on regulatory systems that maintain homeostasis through continuous adjustment—a process termed allostasis. The concept was developed by Bruce McEwen and Eliot Stellar in 1993 and operationalised in epidemiology as a composite score across neuroendocrine (e.g. cortisol, DHEA-S), cardiovascular, metabolic, and immune biomarkers. Higher allostatic load predicts all-cause mortality, cardiovascular disease, cognitive decline, and accelerated biological ageing, and mediates socioeconomic disparities in health. In longevity research, it provides a framework for understanding how cumulative psychosocial and environmental stress translates into measurable organ-system dysregulation.

Antagonistic pleiotropy, formulated by evolutionary biologist George C. Williams in 1957, holds that genes selected for benefits early in life can cause harm later, after reproduction has occurred. Because selection pressure weakens with age, such alleles persist despite late-life costs. The hypothesis is a foundational explanation for why aging evolved and remains a leading evolutionary framework alongside mutation accumulation and disposable soma theory.

Biological age is an estimate of how old a person's body appears to be based on physiological and molecular markers, rather than the calendar. It can be derived from blood biomarkers (e.g. PhenoAge), DNA methylation patterns (epigenetic clocks), grip strength, gait speed or organ-specific proteomic signatures. Although widely used in longevity research, no single biological-age measure is yet endorsed by regulators as a clinical endpoint, and validation varies strongly between methods.

A caloric restriction mimetic (CR mimetic) is a compound that reproduces some molecular and physiological effects of caloric restriction — including AMPK activation, mTORC1 inhibition, sirtuin activation, reduced insulin/IGF-1 signalling, enhanced autophagy and favourable shifts in metabolic biomarkers — without requiring a sustained reduction in food intake. The concept was formalised by Ingram and colleagues in 1998. Leading candidates include rapamycin (mTOR inhibitor), metformin (AMPK activator), resveratrol (putative SIRT1 activator), NAD+ precursors (NMN, NR) and acarbose. Evidence for lifespan extension in mice is established for rapamycin and acarbose under ITP conditions; translation to humans remains an open research question, and no CR mimetic has demonstrated robust healthspan extension in a powered randomised human trial.

A centenarian is a person who has reached the age of 100 years or more. Centenarians are a key research population in longevity science because they typically delay or escape major age-related diseases. Studies such as the New England Centenarian Study and Japan's Okinawa Centenarian Study examine genetic, lifestyle, and environmental factors associated with exceptional human lifespan and healthspan.

Chronological age is the time elapsed since a person's birth, usually measured in years. It is the standard reference variable in demography, medicine and epidemiology and remains one of the strongest single predictors of mortality and many age-associated conditions. Unlike biological age, chronological age does not capture variation in physiological decline between individuals; two people of the same chronological age may differ markedly in functional capacity, disease risk and remaining healthspan, and depending on cohort and endpoint other measures can match or exceed its predictive value.

Compression of morbidity is a concept introduced by James Fries in 1980 describing a scenario in which the onset of chronic disease and disability is postponed faster than the increase in lifespan, so that severe illness is concentrated into a shorter period at the end of life. It is a guiding goal of geroscience and healthspan-oriented medicine. Empirical evidence is mixed: in some populations morbidity has compressed, in others it has expanded as lifespan rose.

Confounding is a distortion of the estimated association between an exposure and an outcome caused by a third variable — the confounder — that is independently associated with both. Classic criteria require the confounder to be a cause (or proxy) of the outcome and to be unequally distributed between exposed and unexposed groups, without lying on the causal pathway. In longevity research, the healthy-user bias exemplifies this: people who adopt a preventive intervention (e.g., statin therapy, caloric restriction, exercise) tend to have healthier baseline behaviors and socioeconomic profiles, so apparent survival benefits in observational cohorts may reflect those unmeasured advantages rather than the intervention itself. Adjustment methods — multivariable regression, propensity-score matching, inverse-probability weighting, and Mendelian randomization — can reduce but rarely eliminate confounding, because unmeasured or poorly measured variables leave residual confounding; Shrank et al. (2011) showed that even extensive covariate adjustment failed to fully remove healthy-user bias in Medicare pharmacoepidemiology studies. The E-value, introduced by VanderWeele and Ding (2017), quantifies how strong an unmeasured confounder would need to be — on the risk-ratio scale — to fully explain an observed association, giving readers a practical yardstick for the robustness of any observational longevity finding.

The Disability-Adjusted Life Year (DALY) is the cornerstone metric of the Global Burden of Disease framework, expressing population health loss as the sum of Years of Life Lost from premature mortality (YLL) plus Years Lived with Disability (YLD). One DALY equals one year of healthy life lost. YLL is calculated as the number of deaths multiplied by the standard remaining life expectancy at the age of death; YLD multiplies prevalence of a condition by a disability weight (0 to 1) reflecting severity. Originally developed by Murray and Lopez at Harvard in 1996 in collaboration with WHO and the World Bank, DALYs now drive global health prioritisation. GBD 2019 estimated approximately 2.5 billion DALYs worldwide, with cardiovascular disease, neonatal disorders, and cancers leading the burden.

The disposable soma theory, proposed by Thomas Kirkwood in 1977, posits that organisms allocate finite metabolic resources between somatic maintenance and reproduction. Because natural selection favors reproductive success, the body invests only enough in repair to survive likely environmental hazards, leaving residual damage that accumulates as aging. The theory remains influential in evolutionary biogerontology and underlies modern thinking on caloric restriction and trade-offs.

Epigenetic drift describes the progressive, largely stochastic divergence of DNA methylation patterns between cells, tissues, and individuals as age advances. The landmark demonstration was Fraga and colleagues' 2005 PNAS study showing that monozygotic twins start life nearly epigenetically identical but accumulate substantial differences in global and locus-specific 5-methylcytosine and histone acetylation with age and environmental exposure. Drift differs conceptually from the clock-like, deterministic methylation changes exploited by Horvath-type epigenetic age estimators: drift adds noise and reduces inter-cell coherence, whereas clock signal is directional. Issa (2014) framed drift as a vicious cycle in which environmental insults degrade the epigenome and predispose to age-related disease and cancer. Drift is now a recognised hallmark of aging.

Frailty is a clinical state of increased vulnerability to stressors resulting from accumulated deficits across multiple physiological systems, leading to diminished reserve and resilience. Two complementary operationalisations dominate the literature: the phenotypic frailty model of Fried and colleagues (2001, Cardiovascular Health Study), which defines frailty by at least three of five criteria (unintentional weight loss, exhaustion, low grip strength, slow walking speed, low physical activity); and the Frailty Index of Mitnitski and Rockwood, which counts the proportion of health deficits present across 30–70 items (symptoms, signs, diagnoses, laboratory values). Both predict adverse outcomes—falls, hospitalisation, disability, and mortality—independently of chronological age, and frailty prevalence rises sharply after age 80. In geroscience, it is a key functional outcome for evaluating senolytic, senostatic, and other geroprotective interventions.

The free radical theory of aging, proposed by Denham Harman in 1956, originally attributed aging to cumulative cellular damage from oxygen-derived free radicals, drawing on rate-of-living and oxygen-toxicity reasoning. Harman's 1972 update, the mitochondrial free radical theory of aging (MFRTA), specifically implicated mitochondrial ROS and mtDNA as the central drivers. While oxidative damage is undeniably involved, large antioxidant trials largely failed, and the theory is now considered partial. Modern frameworks integrate it with mitochondrial dysfunction and redox signaling.

Gerontology is the scientific study of aging across biological, psychological, and social dimensions. Established as a formal discipline in the early 20th century, with Ilya Mechnikov coining the term in 1903, it encompasses biogerontology, social gerontology, and geriatric medicine. It remains the broader umbrella field within which geroscience focuses specifically on molecular and cellular mechanisms relevant to disease prevention.

A geroprotector is any drug, supplement, or lifestyle intervention targeting fundamental aging mechanisms to extend healthspan. Unlike disease-specific treatments, geroprotectors act on upstream processes shared across pathologies: mTOR signaling, senescent cell accumulation, DNA damage responses, and mitochondrial dysfunction. Moskalev, Kennedy, and colleagues (Aging Cell, 2016) proposed four criteria: lifespan extension at the population level, measurable shift of aging biomarkers toward a younger state, acceptable toxicity with a wide therapeutic margin, and minimal side effects. The Geroprotectors.org database (Moskalev et al., Aging, 2015) catalogs over 200 candidates — rapamycin, metformin, resveratrol, NAD⁺ precursors — with mechanism, model-organism data, and FDA status profiles. Evidence is heavily weighted toward animal models: rapamycin reliably extends lifespan in mice and shows immune-rejuvenating signals in early human trials, yet no randomized controlled trial in healthy adults has established mortality or healthspan benefit (Konopka & Lamming, GeroScience, 2023). The field is largely investigational; the TAME trial (Targeting Aging with Metformin) is among the first prospective studies testing a geroprotector against a composite aging endpoint.

Geroscience is an interdisciplinary field that investigates the biological mechanisms of aging and their causal links to chronic disease. Coined around 2007 by researchers at the Buck Institute and formalized by the NIH-led Geroscience Interest Group, it rests on the premise that targeting aging itself can simultaneously delay multiple age-related conditions. It now underpins translational efforts like the TAME trial.

Gompertz law, formulated by the British actuary Benjamin Gompertz in 1825, describes the empirical observation that human mortality risk increases exponentially with adult age: specifically, the force of mortality (hazard rate) approximately doubles every 8 years in most high-income populations. Mathematically, the instantaneous mortality rate is expressed as μ(t) = a·e^(bt), where a is the baseline mortality rate and b is the age-dependent acceleration. The law holds across most of adult life in humans and many other species, but mortality deceleration or plateaus observed at very old ages suggest it is not universal beyond the oldest cohorts. Gompertz dynamics are central to actuarial science, epidemiology, and the theoretical biology of ageing.

HALE (Healthy Life Expectancy) is a WHO summary metric defined as the average number of years a person can expect to live in full health, adjusting total life expectancy downward by the time-equivalent lived in states of less than perfect health. It is computed via the Sullivan method using age-specific mortality from life tables and age-specific prevalence of disability/health states from the Global Burden of Disease (GBD) framework, with disability weights anchored between 0 (perfect health) and 1 (death). The HALE gap (life expectancy minus HALE) quantifies the morbidity burden over a lifetime. According to WHO Global Health Observatory data (most recent update 2019), Germany's HALE at birth is approximately 70 years versus a total life expectancy of about 81 years, implying roughly 11 years lived in compromised health. HALE underpins cross-country health system comparisons and the Sustainable Development Goal 3 targets.

A hazard ratio is the ratio of the instantaneous event rate in one group to that in a reference group at any given moment during follow-up, derived from a Cox proportional-hazards regression model. An HR of 0.75 means the treated group experiences the event at 75% of the rate of controls throughout follow-up, not that overall risk is reduced by 25% at a fixed time point — a common misreading. The proportional-hazards assumption requires that this ratio remain constant over time; violations (e.g., time-varying drug effects) must be tested, and when present, time-restricted or parametric models are more appropriate. In longevity and survival studies the HR is the dominant effect measure, but its magnitude depends on baseline hazard and follow-up length, limiting direct comparisons across trials.

Healthspan is the period of life spent in good health, free from serious chronic disease and major functional impairment. It is conceptually distinct from lifespan, which counts total years lived. In longevity research healthspan is increasingly preferred as an outcome because the goal is to compress the years of frailty and disease at the end of life. Operational definitions vary and may use disease-free survival, disability indices or composite biomarker scores.

Heritability of lifespan is the proportion of variance in age at death attributable to additive genetic differences among individuals in a defined population. Classic twin-study estimates placed narrow-sense heritability at roughly 20–30% (Herskind et al. 1996); large-scale genomic analyses and a 2018 study in Genetics (Ruby et al.) using genealogical databases suggested even lower heritability once marital assortment is properly accounted for, with some estimates falling below 10% for lifespan itself. More recent analyses (Shenhar et al. 2026, Science) that better account for confounding factors place the intrinsic heritability of human life span at approximately 50%, suggesting earlier estimates may have been deflated by failure to separate intrinsic from extrinsic mortality. Even at the higher end, modifiable factors — lifestyle, environment, stochastic events — account for a substantial portion of variance. Specific genetic variants such as APOE ε4 and FOXO3A show replicated associations with mortality risk and exceptional longevity respectively, even if their individual effect sizes are modest.

The hyperfunction theory of aging, proposed by Mikhail Blagosklonny in 2006, holds that aging is driven by the continued overactivity of nutrient- and mitogen-sensing growth pathways — chiefly mTOR (mechanistic target of rapamycin) — rather than by passive accumulation of molecular damage. During development these pathways are essential for growth and reproduction; after developmental completion they are never switched off, creating a "quasi-program" — a purposeless extension of the growth program never selected against because post-reproductive harm confers no fitness penalty. The resulting cellular hyperfunction drives pathological processes such as cellular senescence, hypertrophy, fibrosis, and sterile inflammation, which produce canonical age-related diseases. The theory does not deny that molecular damage accumulates, but argues that mTOR-driven hyperfunction is life-limiting before damage alone would be. Experimental support comes from rodent studies in which rapamycin extended lifespan even when administered late in life (Harrison et al., 2009), and from a randomized trial (Mannick et al., 2014) in which low-dose everolimus improved influenza vaccine response in elderly humans. Whether mTOR inhibition translates to lifespan extension in healthy humans remains open; the PEARL trial provided preliminary safety and healthspan data in 2024 but no definitive outcome evidence as of 2025.

The Kaplan-Meier estimator is a nonparametric method for estimating the survival function S(t) — the probability of surviving beyond a given time t — from censored time-to-event data. At each event time, the estimate is updated as the ratio of subjects remaining at risk minus those who experienced the event, multiplied forward as a product-limit. The resulting step function graphically displays survival over follow-up and allows group comparisons via the log-rank test. Key assumptions include that censoring is non-informative (i.e., subjects who leave the study do not systematically differ in prognosis) and that survival probability is independent across individuals. The median survival — where the curve crosses 50% — is the standard summary statistic; mean survival is rarely used because it requires the curve to reach zero.

Late-life mortality deceleration — the mortality plateau — is the observed phenomenon in which the age-specific hazard of death ceases its exponential Gompertz-law acceleration and flattens into an approximately constant rate, typically from around 105 years onward. Barbi et al. (2018, Science), using administrative records of all 3,836 Italians aged 105 or older between 2009 and 2015, reported an essentially flat hazard curve with an annual death probability of approximately 47–48% (baseline hazard ≈ 0.645), concluding that biological limits to lifespan may be more elastic than previously assumed. A leading mechanistic explanation invokes population heterogeneity: cohorts at extreme ages are disproportionately composed of biologically resilient individuals, so selective survival — rather than genuine slowing of cellular aging — produces the apparent plateau in aggregate hazard (the frailty-selection effect formalized by Vaupel and colleagues). The finding is contested: Gavrilov and Gavrilova (2019, PLOS Biology) showed by simulation that age-reporting errors above 105 can generate spurious deceleration even in perfectly Gompertzian data, implicating inadequate birth-record validation. Dang et al. (2023, Demographic Research), using French cohort data followed from age 105, found no plateau: the Gompertz slope parameter remained significantly positive beyond 105, indicating death rates continue to rise. These studies illustrate that extreme-age mortality dynamics are highly sensitive to data quality, national context, and cohort structure.

Lifespan is the total length of time an organism lives, from birth to death, typically expressed in years for humans. In population terms it is summarised by life expectancy at birth or at a given age. Maximum lifespan refers to the longest documented age reached within a species; for humans this is around 122 years. Lifespan is influenced by genetics, environment, behaviour and access to medical care, and is a classic outcome in longevity research.

Longevity escape velocity describes a hypothetical threshold at which medical advances extend remaining life expectancy by more than one year per calendar year, effectively outrunning aging. Popularized by biogerontologist Aubrey de Grey in the early 2000s, it remains a speculative concept rather than an empirically validated milestone. Mainstream geroscience treats it as an aspirational framing rather than a near-term forecast.

Maximum lifespan is the longest documented or theoretically possible age that a member of a species can reach under optimal conditions, distinct from average life expectancy across a population. In humans, the empirical record stands at 122 years and 164 days, set by the French supercentenarian Jeanne Calment (1875–1997); no independently validated case has come close since. Dong, Milholland & Vijg (Nature, 2016) analysed global demographic data from the International Database on Longevity and found that the annual age at death of the world's oldest person plateaued around 115 years after 1995, concluding that a soft ceiling near 115–125 years is biologically imposed — likely by the accumulating molecular and cellular damage that characterises aging and that no medical intervention has yet reversed at scale. This interpretation is actively contested: Lenart & Vaupel (Nature, 2017) argued that the same data are statistically consistent with continued, if slow, increases in maximum age, and that Calment's record is an extreme-value outlier rather than evidence of a hard limit. Olshansky et al. (Nature Aging, 2024), analysing survival trends in the longest-lived national populations, found life expectancy gains decelerating sharply and concluded that radical extension beyond the current empirical maximum remains implausible within the twenty-first century absent breakthroughs not yet demonstrated in humans. Whether maximum lifespan is a fixed biological constant or a plastic ceiling that geroscience interventions could progressively raise remains one of the central, unresolved questions in aging research.

Mendelian randomization (MR) uses germline genetic variants — typically single-nucleotide polymorphisms associated with an exposure in a genome-wide association study — as instrumental variables to estimate the causal effect of that exposure on an outcome, exploiting the random allocation of alleles at conception as a natural experiment. The method relies on three core assumptions: the instrument is robustly associated with the exposure (relevance), affects the outcome only through that exposure (exclusion restriction), and is independent of confounders (independence). Violations — through pleiotropy, population stratification, or weak instruments — are major pitfalls; sensitivity analyses including MR-Egger, weighted-median, and CAUSE help detect and partially correct for horizontal pleiotropy. In longevity research, MR has been widely used to test causal links between biomarkers such as LDL-C, CRP, IGF-1, or BMI and lifespan outcomes without requiring decades-long randomized trials.

A meta-analysis is a statistical procedure that quantitatively synthesises results from multiple independent studies addressing the same research question, yielding a pooled effect estimate with narrower confidence intervals than any individual study alone. Gene V. Glass coined the term in 1976: effect sizes are extracted from each study, weighted (typically by inverse variance), and combined into a weighted average — visualised in a forest plot, where each horizontal line represents one study and the diamond at the bottom represents the pooled result. Heterogeneity — the degree to which true effects vary across studies beyond chance — is quantified by I²: below 25% indicates low heterogeneity, 50–75% moderate, and above 75% high, though I² is sensitive to the number of included studies. Publication bias, assessed via funnel plots, Egger's test, or trim-and-fill, can skew pooled estimates. In longevity research, meta-analyses aggregate observational cohorts or trials to detect modest effect sizes — such as the survival benefit of physical activity or the association between telomere length and mortality — that individual studies lack power to resolve. Their conclusions are only as valid as the studies they pool: shared systematic biases are amplified, not corrected.

Mortality doubling time (MDT) is the number of years it takes for age-specific mortality risk to double, derived directly from the Gompertz exponent b as MDT = ln(2)/b. In contemporary high-income populations, the MDT for all-cause mortality is approximately 7–8 years in mid-adulthood, meaning a 50-year-old's annual risk of dying is roughly twice that of a 42–43-year-old. MDT is a compact summary of the rate of actuarial ageing and is used comparatively across species (where it varies from months in short-lived organisms to ~8 years in humans, while species such as the naked mole-rat appear to defy Gompertz dynamics entirely with no detectable increase in hazard rate over decades of life — Ruby et al., eLife 2018) and across population subgroups, enabling detection of interventions that alter ageing rate rather than merely shifting baseline mortality.

Multimorbidity is defined as the co-occurrence of two or more chronic conditions within the same person, without designating a primary or index disease — a distinction from the related but person-centred concept of comorbidity. Its prevalence rises sharply with age: roughly 50% of adults over 65 in high-income countries live with three or more chronic conditions. Multimorbidity is strongly associated with polypharmacy, functional decline, reduced quality of life, greater healthcare utilisation and higher mortality, and it challenges single-disease clinical guidelines that were developed in trial populations that often excluded it. In geroscience, multimorbidity is both a key outcome of biological aging and a prime motivation for targeting upstream aging processes rather than individual diseases sequentially.

Mutation accumulation theory is an evolutionary explanation for senescence, first proposed by Peter Medawar in his 1952 lecture "An Unsolved Problem of Biology." It holds that late-acting deleterious mutations — those harmful only after peak reproductive age — escape natural selection and accumulate over generations. Selection pressure declines with age: mutations lethal before reproduction are purged, while those reducing fitness only post-reproduction persist in Medawar's "selection shadow"; genetic drift then raises them to high frequency, eroding physiological function in older individuals. Charlesworth (2001) formalised this in a quantitative genetic model predicting that mean and additive genetic variance of age-specific mortality rates rise exponentially with age, consistent with human and Drosophila data. Turan et al. (2019) provided molecular evidence, detecting an age-related decrease in transcriptome conservation (ADICT) across 16 tissue types in five mammalian species, with late-expressed genes under weaker purifying selection and enriched in apoptosis and inflammatory pathways. The theory differs from antagonistic pleiotropy in requiring no early-life benefit, only late-life harm; which mechanism predominates for any given trait remains empirically unresolved.

Negligible senescence describes organisms that show no measurable functional decline, increase in mortality risk, or loss of reproductive capacity with chronological age. The term was coined by biogerontologist Caleb Finch in his 1990 book 'Longevity, Senescence, and the Genome' to characterize species such as certain rockfish, certain tortoises, and hydra. Naked mole-rats, often cited in this context, exhibit extremely slow but not strictly negligible senescence and became associated with this discussion through later work (e.g., Buffenstein 2008 onward). Negligible senescence is studied as a comparative biology benchmark for understanding why most mammals, including humans, do age.

The number needed to treat (NNT) is the average number of patients who must receive an intervention for one additional patient to benefit over control. Introduced by Laupacis, Sackett, and Roberts in the New England Journal of Medicine (1988), NNT is the reciprocal of the absolute risk reduction (ARR): NNT = 1 ÷ ARR, where ARR equals the control-arm event rate minus the treatment-arm event rate. Because NNT anchors benefit to a concrete patient count, it makes absolute effect size transparent: an NNT of 50 means 50 people must be treated for one to avoid the target outcome. In aging and longevity medicine — where interventions target prevention of cardiovascular disease, cancer, or functional decline — NNT values are often far higher than intuition suggests. A re-analysis of the JUPITER trial (Ridker et al., 2009) found an NNT of 95 for the primary composite endpoint among the lowest-risk participants receiving 5-year rosuvastatin; statins in true primary prevention for healthy adults carry NNTs of 60-270 depending on endpoint and risk stratum. NNT depends critically on baseline risk, observation period, and comparator; a figure cited without this context can mislead, and confidence intervals derived from the ARR are essential for proper interpretation.

The oldest-old is a demographic and gerontological term for individuals aged 85 and above, the fastest-growing segment of most high-income country populations. This group shows heterogeneous functional and cognitive trajectories; a substantial minority maintains high functional capacity into the late eighties and beyond, a phenomenon sometimes termed successful aging. Compared with younger old adults (65–84), the oldest-old display distinct epidemiology: traditional cardiovascular risk factors such as hypertension and elevated LDL-C lose predictive power, while markers of physical function, nutritional status, and resilience become stronger mortality predictors. Studying this group is methodologically complicated by survival bias — those who reach 85 are a selected population — which may inflate associations between traits observed at that age and longevity.

Polypharmacy is conventionally defined as the concurrent use of five or more medications by one patient, though thresholds vary by definition (some use ≥4, hyperpolypharmacy typically ≥10); it must be distinguished from appropriate polypharmacy, in which multiple drugs are each evidence-based for the individual's conditions. Prevalence increases sharply with age and multimorbidity — over 40% of adults aged 65 and older in many high-income countries take five or more drugs. The clinical risks include drug-drug interactions, additive adverse effects, prescribing cascades (where a drug side effect is treated with another drug), impaired adherence, falls, cognitive impairment and hospitalisation. Deprescribing — the structured reduction of medications that lack net benefit — is an emerging discipline in geriatric and longevity medicine, supported by a growing evidence base for specific drug classes including proton pump inhibitors, benzodiazepines and anticholinergics in older adults.

A Quality-Adjusted Life Year (QALY) is one year of life weighted by health-related quality of life, where 1.0 represents one year in perfect health and 0 represents death (negative values for states worse than death are permissible). QALYs are computed by multiplying time spent in a health state by its utility weight, typically derived from instruments like EQ-5D or SF-6D. Originally formalised by Weinstein and colleagues, QALYs anchor cost-effectiveness analyses in health technology assessment: the UK's NICE applies a reference threshold of £25,000-£35,000 per QALY gained since April 2026 (previously £20,000-£30,000), while Germany's IQWiG uses an efficiency frontier approach rather than a fixed threshold. A treatment costing £15,000 that yields 2 QALYs (£7,500/QALY) would generally be deemed cost-effective by NICE.

A randomized controlled trial (RCT) is an experimental study design in which participants are allocated to an intervention or control condition through a chance-based process — the most reliable method for establishing whether a treatment causes its observed effect. Randomization distributes known and unknown confounders evenly across groups so that outcome differences are attributable to the intervention rather than pre-existing group differences. Blinding — masking participants, clinicians, or assessors to group assignment — reduces bias; double-blind designs, where neither party knows the allocation, are the strongest configuration. In longevity research, RCTs separate causal claims from association: several interventions appearing protective in cohort studies showed smaller or no benefits when rigorously tested. RCTs in aging science are difficult because meaningful endpoints such as delayed multimorbidity or extended lifespan require decades of follow-up. The TAME trial (Targeting Aging with Metformin), enrolling ~3,000 adults aged 65–79 with a composite age-related disease endpoint, shows the field's effort to test a geroscience intervention at scale. As of 2026, no anti-aging intervention has completed a powered RCT with all-cause mortality as a primary endpoint in humans.

The rate of living theory proposes that an organism's lifespan is inversely proportional to its mass-specific metabolic rate — the faster energy is consumed, the sooner the organism dies. Max Rubner articulated the idea in 1908 by comparing five domestic mammals (guinea pig, cat, dog, cow, and horse) and showing that lifetime energy expenditure per unit body mass is roughly constant across species; Raymond Pearl extended it in his 1928 book The Rate of Living, coining the term and demonstrating in Drosophila that lower temperatures, which slow metabolic rate, prolong survival. The theory gained mechanistic plausibility when Denham Harman proposed the free radical theory of aging in 1956, linking mitochondrial respiration to cumulative oxidative damage. Complicating evidence followed: birds and bats live substantially longer than non-flying mammals of similar body size despite comparable or higher metabolic rates, as reviewed by Munshi-South and Wilkinson (2010). Hulbert et al. (2007) attributed this partly to stronger mitochondrial antioxidant defenses and lower membrane peroxidizability in longer-lived taxa. Speakman (2005) showed that after controlling for body size, residual daily energy expenditure correlates negatively with maximum lifespan in mammals but not in birds, concluding that interspecific comparisons are confounded by differences in oxidative defense and repair capacity. The simple rate-of-living formulation has been superseded by more mechanistically specific theories, though it remains historically foundational.

The reliability theory of aging, advanced by Leonid and Natalia Gavrilov in the early 1990s, applies engineering reliability mathematics to biological systems. It models organisms as redundant networks of components that fail stochastically; aging arises as redundancy depletes, producing the observed Gompertz mortality curve. The theory elegantly explains late-life mortality plateaus and provides a quantitative bridge between molecular damage and population-level survival data.

In gerontology, physical resilience is operationally defined as the capacity to recover or maintain physical function after an acute health stressor - illness, surgery, fall, bereavement, or hospitalisation. Whitson and colleagues (2016, J Gerontol A) proposed it as a distinct construct from frailty: frailty captures pre-stressor vulnerability, whereas resilience captures the post-stressor recovery trajectory. Two people with identical baseline function can show very different recovery slopes after the same insult. Resilience is quantified by tracking gait speed, grip strength, ADL scores, or biomarkers over time after a defined stressor, typically with linear mixed models or area-under-recovery-curve approaches. High resilience predicts lower mortality and institutionalisation; low resilience identifies people who will benefit most from prehabilitation and targeted rehabilitation interventions.

Successful aging is a gerontological framework introduced by John Rowe and Robert Kahn (1987, Science) and elaborated in 1997 (The Gerontologist), distinguishing usual aging - where extrinsic factors compound intrinsic decline - from successful aging, where extrinsic factors are neutral or beneficial. The three pillars are: (1) low probability of disease and disease-related disability, (2) high cognitive and physical functional capacity, and (3) active engagement with life (productive activity plus interpersonal relations). The MacArthur Foundation Research Network on Successful Aging — which Rowe and Kahn directed — provided the empirical basis from which the model was developed. It has been criticised for setting a high disease-free bar that excludes most older adults living well with chronic illness, and for cultural narrowness; later frameworks (e.g., WHO healthy ageing) emphasise adaptation and functional ability over disease absence.

A supercentenarian is a person verified to have reached the age of 110 years or more. The 110+ threshold and term were popularized chiefly by L. Stephen Coles, founder of the Gerontology Research Group, with demographer James Vaupel contributing complementary validation work through MPIDR and the International Database on Longevity. The cohort numbers only a few hundred globally at any time and is studied for genetic resilience, late-life morbidity compression, and as a benchmark against unverified age claims.