Why do we age?

DNA damage is not the same as a mutation.

DNA is continually exposed to chemical modification, replication errors and strand breaks. A lesion is damage to the molecule; a mutation is a lasting sequence change. Repair can remove a lesion, but an error copied into both strands may become a stable mutation. This distinction matters: measuring mutation burden does not capture every kind of DNA injury.

Base-excision repair handles many small lesions; nucleotide-excision repair removes bulky damage. Double-strand breaks can be rejoined by end joining or repaired using a homologous template when one is available. DNA-damage signalling can temporarily halt division, permitting repair; persistent damage can favour senescence or cell death. Different tissues use these responses differently.

Still unresolved: which lesions matter most in each tissue, and how much ordinary ageing is driven by mutations versus damage responses, altered gene expression or selection among cell populations.

DNA is a template; RNA carries an instruction.

A chromosome is one long DNA molecule packaged with proteins. A gene can be transcribed into RNA by RNA polymerase; for a protein-coding gene, processed messenger RNA is then translated by a ribosome into an amino-acid chain. Transcription produces an RNA copy of a DNA sequence; translation uses that message to assemble a protein. Some lesions block transcription; others distort a copied sequence or activate stress responses. Not every lesion prevents protein production, and most individual lesions are successfully repaired.

In base-excision repair, a DNA glycosylase removes a damaged base, leaving an abasic site. APE1 cuts the backbone near this site. Subsequent processing, DNA synthesis and ligation restore continuity. The intermediate is itself potentially hazardous, so the sequence has to be coordinated. Repair is a pathway, not one enzyme wiping damage away. Other lesion types require other pathways. NAD+ is consumed by PARP enzymes during some DNA-damage responses; it is not a universal fee paid for every repair reaction. Yang et al., neuronal DNA repair.

Telomeres protect an end that looks like a break.

Telomeres are repeated DNA sequences and associated proteins at chromosome ends. Their protective structure helps prevent normal chromosome ends from being treated as broken DNA. End-replication constraints, processing and oxidative injury can shorten or damage them. Critically dysfunctional telomeres activate a damage response.

Telomerase can extend telomeric DNA, but activity varies with cell type. A telomere measurement in blood reflects a mixture of cells and their histories; it is not a universal countdown for the brain, muscle or whole body.

Still unresolved: how to protect telomere function where it is limiting without enabling unwanted cell expansion. Longer is not automatically better.

Mitochondria convert energy and coordinate stress responses.

Electrons from fuel oxidation enter the respiratory chain through carriers including NADH. Electron transfer through the inner mitochondrial membrane helps pump protons from the matrix into the intermembrane space. Their return through ATP synthase drives ATP production. Oxygen accepts electrons at the end of the chain and is reduced to water. ATP then supports work throughout the cell.

Mitochondria also participate in calcium handling and cell-death signalling. Their function depends on proteins encoded in both nuclear and mitochondrial genomes. Fusion, fission and selective removal of damaged mitochondria— mitophagy —help maintain the network. Age-related changes vary by tissue, activity and disease; mitochondria do not all lose the same percentage of capacity.

Still unresolved: which mitochondrial changes initiate decline and which compensate for it. More mitochondrial material is not necessarily better mitochondrial quality.

Reactive oxygen species are both signals and potential sources of injury.

ROS includes oxygen-derived molecules such as superoxide and hydrogen peroxide. Not all ROS are free radicals: hydrogen peroxide has no unpaired electron. Controlled production helps regulate signalling and adaptation. Excessive or misplaced production can damage proteins, lipids and DNA when defence and repair cannot keep pace. The molecular species, location and duration all matter. [7]

Still unresolved: how to target harmful redox processes without suppressing useful signalling. A generic antioxidant label does not identify the right molecule, tissue or dose. [7]

Protein quality depends on synthesis, folding and disposal.

Proteostasis is the maintenance of a working protein population. Chaperones assist folding; the ubiquitin–proteasome system degrades many selected proteins. Macroautophagy encloses larger cargo in a double-membrane autophagosome that fuses with a lysosome, whose enzymes digest the contents. Recycling supplies building blocks as well as removing dysfunctional material.

Disposal is not simply “protein breakdown” as a harmful process. Controlled degradation is necessary. A build-up of autophagosomes can mean increased formation or impaired clearance, so counting them alone does not measure autophagic flux —the throughput of the complete pathway.

Still unresolved: when increasing clearance helps, when the downstream lysosome is the bottleneck, and whether a treatment can improve tissue function without disrupting necessary protein turnover.

Genetic regulation does not establish a programme for death.

Evolutionary theory asks why maintenance has limits. Mutation accumulation proposes that selection is less effective against harmful effects expressed late in life. Antagonistic pleiotropy describes genetic effects that benefit early survival or reproduction but impose later costs. Disposable soma theory emphasises trade-offs in investment between maintenance and other demands. These are explanations of evolutionary pressures, not conscious resource decisions. [2]

Developmental and growth programmes may also remain active in contexts where their effects become maladaptive. But a regulated pathway is not evidence that human ageing evolved for the purpose of killing the organism. The relative importance of these explanations varies across species. [2]

Still unresolved: how conserved pathways interact with the different lifespans, reproductive strategies and disease patterns of humans.

Growth and maintenance need coordination.

Insulin/IGF signalling, mTOR, AMPK and sirtuins connect nutrient availability and energy state to cell behaviour. mTORC1 promotes anabolic processes and can inhibit autophagy initiation. AMPK responds to cellular energy stress and helps shift activity toward restoring energy balance. These pathways are context dependent: building muscle after training and maintaining tissue require anabolic signalling.

Still unresolved: how to influence the right pathway in the right tissue at the right time without compromising repair, metabolic function or immune defence. “Less mTOR” is not a complete health strategy.

A molecular clock measures a pattern.

Epigenetic regulation involves changes in chromatin and DNA-associated marks that affect gene regulation without changing the DNA sequence. Age-associated changes include both increases and decreases in methylation at different sites. A methylation clock combines selected measurements into an estimate; it does not read one master age setting.

DNA is wound around histone proteins to form nucleosomes. Histone modifications, DNA methylation, chromatin-remodelling complexes and transcription factors jointly affect access to regulatory sequences. Methyl groups are normal biochemical components, often donated through S-adenosylmethionine; their position and context matter. Age-associated change can include local gains, local losses and increased variability. Reprogramming experiments test whether some cell states are reversible, but loss of cell identity, abnormal proliferation and delivery remain critical barriers to safe organism-wide interventions. [1] [12]

Still unresolved: which clocks reliably track treatment benefit, how cell composition influences them, and how much of a changed score reflects causally meaningful repair.

A cell can stop dividing and remain biologically active.

Cellular senescence is a typically stable growth-arrest programme that can follow telomere dysfunction, oncogenic signalling or other stress. p53–p21 and p16–RB pathways help enforce arrest. Senescent cells may remain metabolically active and produce a senescence-associated secretory phenotype(SASP), including inflammatory mediators and matrix-remodelling factors. [14]

The response can suppress proliferation of damaged cells and contribute to repair. Persistent accumulation can disturb neighbours and tissue structure. Senescence is heterogeneous: no single marker identifies all senescent cells, and p16 expression is not an exact body-wide senescent-cell count. [14]

Still unresolved: which cells should be removed, at what stage of disease, and whether benefits outweigh disrupting their useful roles.

A necessary cofactor is not automatically an effective supplement.

NAD+ accepts electrons during metabolism, forming NADH; it is also consumed by enzymes involved in signalling and DNA-damage responses, including sirtuins and PARPs. Its concentration and turnover differ among tissues and cellular compartments. Changing a blood concentration does not establish restoration of mitochondrial performance throughout the body.

Still unresolved: who might benefit clinically, which outcomes improve, and what sustained use means. Increasing a substrate, altering a pathway and improving long-term health are separate claims.

A repair response becomes a persistent background signal.

Inflammaging describes chronic, low-grade inflammatory activity associated with ageing. Sources can include tissue injury, altered immune regulation, senescent-cell signals and changes in microbial exposure. Acute inflammation recruits defence and initiates repair; failure to resolve the response can keep tissues under stress. A raised inflammatory marker does not identify a single cause.

Still unresolved: how to reduce maladaptive inflammatory activity while retaining host defence and healing.

Less effective defence can coexist with more inflammation.

An older immune system is not simply switched off. Changes in the diversity and responsiveness of immune cells can weaken responses to unfamiliar pathogens while inflammatory activity persists. Prior infections, health conditions and tissue environments influence this balance. A more inflammatory state is not necessarily a more protective immune state.

Still unresolved: which immune changes should be targeted, how to select participants and which clinical endpoint best tests the proposed benefit.

Stem cells depend on the tissue around them.

A stem cell must balance remaining available with producing specialised progeny. Its niche includes neighbouring cells, extracellular matrix and local signals. Regenerative failure can reflect intrinsic damage, altered activation, changes in the niche or systemic signals. More activation is not always better: excessive demand can compromise long-term maintenance.

Still unresolved: which circulating and local factors are causal and how their effects differ among organs. Muscle, cartilage, peripheral nerves and skin have very different repair capabilities; a single age-related healing percentage would obscure those differences.

Reserve connects molecular biology to everyday life.

Stairs, shopping and balance place demands on interacting systems. Muscle force, power, oxygen delivery, coordination and confidence all affect the margin between an everyday task and a person’s current capacity. Illness and inactivity can narrow that margin; training and adapting the task can widen it. This is a functional model, not a numerical test of biological age.

Capacity does not decline as a universal straight line. People begin with different reserves and experience different exposures, diseases, opportunities and training histories. A measure such as VO₂max integrates heart, blood, lungs and muscle; a lower value cannot be attributed solely to mitochondrial ageing.

Stability requires continuous adjustment.

Homeostasis is the active regulation of internal conditions—glucose, temperature, blood pressure, oxygen delivery—not an unchanging state. Ageing can reduce the range of disturbances a system can absorb and the speed with which it recovers. Several systems may compensate at rest and fail only during infection, exertion or a fall. This is why resting blood tests and maximum functional performance answer different questions.

Immune ageing includes changes in the output of new lymphocytes, accumulated antigen experience, altered clonal diversity and changes in innate immune signalling. Thymic involution reduces the tissue supporting new T-cell development; persistent infections and lifelong exposures shape the repertoire. Meanwhile, senescent cells, damaged molecules and altered tissue barriers can sustain inflammatory stimulation. Reduced defence against a new pathogen can therefore coexist with excessive background inflammation. The magnitude is not uniform across individuals or populations. [1] [14]

Stem-cell exhaustion likewise means more than simply “running out of stem cells”. Cells may remain present but activate poorly, differentiate differently or encounter an inhibitory niche. Quiescence—a reversible resting state that preserves a reserve—is different from senescence. Fibrosis changes the extracellular scaffold; altered immune and vascular support can impair regeneration even when the stem-cell genome is intact. A functional explanation needs the cell and the tissue around it. [20]

Memory belongs to a network.

The hippocampus lies in the medial temporal lobe and contributes to forming and retrieving episodic memories and representing spatial relationships. The hippocampal formation includes the dentate gyrus, cornu ammonis fields and subiculum. Its circuits interact with the entorhinal cortex and wider cortical networks. Attention, executive control, vocabulary and memory do not reside in one interchangeable brain region. A change in hippocampal volume cannot describe every aspect of cognition.

Synapses are specialised communication sites between neurons. Their strength and structure can change with activity— synaptic plasticity. Neurons also depend on axonal conduction, myelin, vascular supply, glial support and mitochondrial energy. Normal ageing can affect these processes without implying that dementia is inevitable. Neurodegenerative disease introduces additional pathology and should not be treated as a synonym for ordinary ageing. Meta-analysis of hippocampal subfield volumes.

A plausible pathway is not yet a proven mediator.

Brain-derived neurotrophic factor (BDNF) is a signalling protein involved in neuronal survival and plasticity. Mature BDNF activates the TrkB receptor, which recruits intracellular pathways influencing synapses and gene expression. Its precursor, proBDNF, has distinct receptor interactions and should not be treated as biologically identical. A blood assay is also not a direct sample of hippocampal BDNF: platelets, sample handling, molecular form and tissue exchange affect what it measures.

Power is strength expressed at speed.

Strength concerns force production; power is the rate of doing mechanical work and, in a simple translational setting, force multiplied by shortening velocity. Catching a stumble requires force quickly. A maximal slow strength test therefore cannot fully capture the capacity needed to regain balance.

A motor unit comprises a motor neuron and the muscle fibres it activates. Age-associated motor-neuron loss can denervate fibres; collateral sprouting from surviving neurons may reinnervate some of them. The resulting remodelling, altered firing and loss or atrophy of fast fibres can affect rapid force production. Muscle quality also reflects architecture, connective tissue, fat infiltration, calcium handling and energy supply. Loss of muscle mass, strength and power are related but not interchangeable outcomes. Reid et al., longitudinal neuromuscular measurements.

Build usable strength.

Resistance exercise gives muscle and the nervous system a repeated, recoverable demand. Improvements can reflect coordination and force production as well as muscle growth. That is why a strength gain should not be reported as an equal percentage increase in muscle mass.

Useful application: choose progressively challenging movements appropriate to present ability, practise consistently and allow recovery. Track meaningful tasks and strength, not only appearance.

Train the oxygen-delivery and utilisation system.

Useful application: build a sustainable amount of walking, cycling or other aerobic activity, with intensity matched to ability. Strength and aerobic fitness solve overlapping but distinct problems. Everyday activity also creates opportunities to practise function; the LIFE results show why that matters. [23]

Protect a recurring opportunity for recovery.

Useful application: protect sufficient, regular sleep opportunity and address persistent sleep difficulties. The aim is adequate recovery and daytime function, not maximising hours without regard to sleep quality.

Support function and reduce disease risk.

Adequate energy and protein provide resources for maintaining and adapting tissue. Dietary quality and metabolic context matter alongside calories. Restriction that undermines strength, recovery or nutritional adequacy can work against the function a person wants to preserve.

Calorie restriction: CALERIE provides useful human biomarker evidence, but its target of 25% restriction should not be mistaken for an achieved intake or a prescription for every reader. A small change in an ageing algorithm cannot be converted into promised extra years of life. [13]

Give each capacity a reason to stay.

For adults, WHO recommends 150–300 minutes of moderate aerobic activity or 75–150 minutes of vigorous activity per week, plus muscle-strengthening activity on at least two days. Older adults should also practise varied multicomponent activity emphasising balance and strength on three or more days. These are public-health targets, not a pass–fail test: some activity is better than none, and the dose should build from current ability. WHO 2020 recommendations.

A useful programme combines walking or another aerobic activity with progressively challenging resistance exercise, balance practice and everyday movement. Where appropriate, supervised practice of producing force rapidly can address power; maximal lifting load alone does not train every demand. Choose outcomes that matter: rising from a chair, carrying shopping, climbing stairs or walking farther with manageable effort. Recovery after illness deserves particular attention because temporary inactivity can consume an already small reserve.

Nutrition should support that work: sufficient energy, regular protein-containing foods and dietary variety. An older adult losing weight unintentionally needs a different strategy from someone seeking to reduce excess adiposity. Restrictive diets that undermine strength or nutritional adequacy can defeat the goal of preserving function. Sleep deserves regular opportunity and attention to treatable problems such as insomnia or sleep apnoea; neither an isolated bad night nor a consumer sleep score quantifies biological ageing.

Prevention also acts outside an exercise session: avoid smoking, address hypertension and other cardiovascular risks, review hearing and vision, follow relevant vaccination and screening recommendations, and seek assessment when function changes unexpectedly. Social participation and an accessible environment can preserve meaningful activity when intrinsic capacity is limited. Aids and adaptations are ways to support independence, not signs that the biology has been “defeated”. National Institute on Aging: healthy ageing.

Keep the claim attached to the endpoint.

NAD+ precursors, senolytics and mTOR inhibitors target credible biology. Credibility is a starting point for trials. The relevant questions are who was treated, what improved, how long it lasted and what harms were measured. Mouse survival, human metabolic changes and clinical disability are different levels of evidence. [17] [16] [19]

Longevity marketing crosses the line when a molecular change becomes a claim of restored youth, a mouse lifespan gain becomes a human promise, or an exploratory subgroup is presented as the main result. Look for the prespecified primary outcome, absolute effects, uncertainty and independent replication.

Healthspan is a useful organising aim: retain the capacity to live well, reduce preventable disease and respond to challenges. Ageing biology makes those goals more understandable; it does not yet turn them into a single score or guaranteed protocol.