Biological Aging: What It Is, How to Measure It, and What Actually Slows It

What “biological age” means

Chronological age is time since birth. Biological age is an estimate of how selected molecular, physiological, or functional measurements compare with patterns seen across populations. Researchers use the concept to study why people of the same chronological age can have different disease, disability, and mortality risks.

There is no single biological-age organ, molecule, or accepted clinical reference standard. The updated hallmarks-of-aging framework describes 12 interacting processes, including genomic instability, epigenetic alteration, loss of proteostasis, impaired autophagy, altered nutrient sensing, mitochondrial dysfunction, cellular senescence, stem-cell exhaustion, inflammation, and dysbiosis [1]. A commercial blood or saliva score samples only part of that system.

A useful distinction is:

• age estimation, which predicts chronological age from biomarkers;
• risk prediction, which estimates future disease, disability, or mortality; and
• treatment response, which changes after an intervention.

A test can perform well at one task and poorly at another. A clock that accurately guesses age does not automatically prove that lowering its score will extend life.

How epigenetic clocks work

Epigenetic clocks use methylation values at selected DNA sites and an algorithm trained on a particular dataset. First-generation clocks were optimized to predict chronological age. Later clocks incorporated clinical outcomes, smoking-related signals, proteins, or repeated longitudinal measurements.

DunedinPACE, for example, was designed to estimate the pace of aging from longitudinal changes in multiple organ-system measures rather than to produce a simple “you are 52” result [2]. GrimAge was trained using methylation surrogates associated with plasma proteins and smoking exposure and has predicted mortality and disease outcomes in cohorts [3].

These clocks are not interchangeable. They use different CpG sites, training populations, laboratory pipelines, and outcomes. A result can also be affected by blood-cell composition, sample handling, normalization, acute illness, and technical batch effects. Repeating a test with another company can therefore create an apparent change that reflects the assay or algorithm rather than biology.

What a consumer biological-age result cannot prove

A single result does not:

• diagnose cellular senescence, mitochondrial disease, frailty, or sarcopenia;
• identify which organ is “older”;
• show that a supplement or drug caused a change;
• prove that a lower score reduced disability or extended life;
• replace standard cardiovascular, cancer, bone, cognitive, or metabolic assessment; or
• justify starting hormones, metformin, rapamycin, senolytics, or a weight-loss drug.

An international Delphi panel reached consensus on several biomarkers that may be useful as outcomes in intervention studies, spanning inflammatory, physiological, functional, and epigenetic domains. The same paper concluded that future research still needs to determine which combinations have the greatest utility [4]. “Useful in intervention studies” is not the same as “validated for routine treatment decisions.”

A better measurement framework

The first five domains can trigger established clinical actions. The last domain is most useful when methods, timing, and interpretation are standardized.

Frailty and sarcopenia are clinically actionable

Frailty describes reduced physiological reserve and vulnerability to stressors. The Fried phenotype uses weight loss, exhaustion, weakness, slow walking, and low activity; three or more criteria defined frailty in the original cohort [5]. Other valid approaches use accumulated health deficits. Frailty is not inevitable and should prompt evaluation for reversible contributors such as medication effects, depression, anemia, pain, undernutrition, neurologic disease, and social isolation.

Sarcopenia focuses on muscle strength, muscle quantity or quality, and physical performance. EWGSOP2 places low muscle strength at the front of case finding and uses muscle measurement to confirm sarcopenia, with poor physical performance indicating severe disease [6]. Thresholds depend on the method and population, so an isolated home grip reading should not be treated as a diagnosis.

These functional measures matter because they connect directly to independence. They are often more actionable than a molecular score: a slower gait or repeated difficulty rising from a chair can lead to strength training, medication review, vision and footwear assessment, nutrition support, and fall-prevention work.

Exercise has human functional-outcome evidence

Physical activity affects cardiovascular risk, glucose control, bone, muscle, mood, sleep, and mobility. It does not need to “reverse an epigenetic clock” to be valuable.

In the LIFE randomized trial, 1,635 sedentary adults age 70 to 89 at risk for mobility disability received either a structured moderate-intensity program or health education. The activity program included walking, resistance, and flexibility work and reduced the incidence of major mobility disability over an average 2.6 years [7].

The SPRINTT trial enrolled older adults with physical frailty and sarcopenia. A multicomponent program built around physical activity, nutrition counseling, and technology support reduced mobility disability in the prespecified lower-function group [8].

A safe program is scaled to baseline function and disease. Aerobic activity, progressive resistance training, balance work, and reduced sedentary time can all matter. Chest pain, fainting, unstable heart or lung disease, severe joint symptoms, or repeated falls may require assessment before intensification.

Nutrition should support health, not force a clock score

Older adults need enough energy, protein, and micronutrients to preserve muscle and recover from illness. The right intake varies with body size, kidney or liver disease, appetite, training, weight goal, and swallowing or dental problems. A universal high-protein target can be inappropriate, especially in advanced kidney disease or when it displaces total energy and fiber.

Unintentional weight loss deserves evaluation. It can reflect malignancy, gastrointestinal disease, depression, medication effects, food insecurity, dental problems, or neurocognitive change. In an older adult with frailty, aggressive calorie restriction may worsen muscle and bone loss even if it lowers a commercial biological-age score.

What calorie-restriction studies actually show

CALERIE randomized 220 adults without obesity to a calorie-restriction target or usual intake for two years. Participants achieved less restriction than the 25% target on average. A post hoc DNA-methylation analysis found a modest slowing on DunedinPACE but no significant change in PhenoAge or GrimAge biological-age estimates [9].

That is an important research signal, not proof that calorie restriction makes an individual younger or extends human lifespan. The clocks disagreed, the analysis was not the trial’s original primary endpoint, and participants were carefully selected. Any calorie deficit should preserve nutrition, lean mass, bone health, and quality of life.

A frequently cited eight-week diet-and-lifestyle pilot randomized 43 healthy men and reported a lower methylation-age estimate in the intervention group [10]. Its small, narrow sample and combined intervention make it unable to identify which component mattered or whether the clock difference produced a durable health benefit.

Cellular senescence and senolytics

Cellular senescence is a real biological process. Damaged or stressed cells can stop dividing and produce signals that affect surrounding tissue. Animal experiments show that manipulating senescent cells can change age-related phenotypes. Translating that result to general treatment in healthy humans remains unresolved.

The often-cited dasatinib-plus-quercetin study in idiopathic pulmonary fibrosis was an open-label pilot with 14 participants. It assessed feasibility and physical-function measures in people with serious lung disease, not longevity in healthy adults [11]. It cannot establish that self-administered dasatinib, quercetin, or fisetin extends life.

Dasatinib is a prescription cancer drug with potentially serious adverse effects and interactions. Over-the-counter flavonoids are not proven substitutes for a validated senolytic therapy. No senolytic regimen has an established general anti-aging indication.

Rapamycin, metformin, and NAD precursors

Rapamycin

Rapamycin extends lifespan in several animal models and affects nutrient-sensing and autophagy pathways. Human studies have explored immune response, skin, periodontal outcomes, and short-term safety using different rapalogs and regimens. They have not established that off-label rapamycin prolongs human life or prevents multimorbidity in otherwise healthy adults.

Potential harms include mouth ulcers, infection, impaired wound healing, lipid and glucose changes, low blood counts, lung toxicity, and drug interactions. Animal lifespan results cannot determine a safe human anti-aging dose.

Metformin

Metformin improves clinical outcomes for defined populations with type 2 diabetes and is studied in several other settings. Observational comparisons between people taking metformin and people not taking it are vulnerable to selection, disease severity, and treatment differences. Metformin is not proven to slow aging in people without a standard indication, and it can cause gastrointestinal effects, vitamin B12 deficiency, and rare lactic acidosis in high-risk settings.

NAD precursors

Nicotinamide riboside and nicotinamide mononucleotide can change NAD-related biomarkers. In a randomized crossover trial, nicotinamide riboside increased blood NAD-related measures and was generally tolerated over six weeks [12]. The trial did not show that the supplement extended lifespan or prevented disability. Raising a pathway biomarker is not the same as improving a patient-centered outcome.

Weight-loss and hormone therapies are not generic anti-aging drugs

Treating obesity can improve cardiovascular, metabolic, mobility, and quality-of-life outcomes for eligible patients. In SELECT, semaglutide 2.4 mg reduced major cardiovascular events in adults with established cardiovascular disease and BMI at least 27 kg/m² who did not have diabetes, while adverse-event discontinuations were more common than with placebo [13]. That is a disease-outcome indication in a defined population, not proof that semaglutide reverses biological aging.

Likewise, testosterone or menopausal hormone therapy should be considered for accepted symptom or disease indications after evaluating benefits, contraindications, and alternatives. A low commercial clock score does not establish hormone deficiency, and a high score does not justify hormone treatment.

How to use a biological-age test responsibly

If a person chooses testing despite the limitations:

1. Record the exact company, assay, algorithm version, tissue, collection conditions, and date.
2. Avoid testing during acute illness or immediately after a major medication or weight change.
3. Do not compare values from different clocks as if they were the same unit.
4. Look at the laboratory’s repeatability and validation population, not only testimonials.
5. Predefine what result would change an established clinical action.
6. Do not start or stop prescription therapy solely because the number moved.
7. Interpret change alongside function, symptoms, standard risk factors, and adverse effects.

A one-year numerical change over a short interval may fall within biological and technical variation. Repeated testing can create false precision and encourage unnecessary supplements or restrictive diets.

A practical healthy-aging review

The highest-value review is personalized, not a universal “longevity panel.” It may include:

• blood pressure and tobacco exposure;
• guideline-based lipid, diabetes, kidney, and cancer screening;
• medication reconciliation and anticholinergic or sedative burden;
• vaccination and infection prevention;
• weight trajectory and signs of undernutrition;
• activity, strength, gait, balance, falls, and daily function;
• sleep duration, sleep apnea symptoms, mood, hearing, and vision;
• bone-health assessment when age and risk indicate it; and
• social connection, caregiving strain, food access, and advance care goals.

Testing should follow age, symptoms, family history, medications, and risk factors. Routine fasting insulin, homocysteine, sex hormones, inflammatory cytokines, whole-body DXA, or repeated epigenetic testing is not required for every adult over 40.

Bottom line

Biological aging is scientifically meaningful but cannot be compressed into one definitive personal age. Epigenetic clocks are valuable research tools, and some predict health outcomes, but their disagreement and technical variability limit routine individual treatment use.

The strongest current strategy is to improve measurable health and function: prevent tobacco exposure, manage blood pressure and lipids, treat diabetes and other diseases, stay physically active, preserve strength and mobility, maintain adequate nutrition, sleep well, vaccinate, and address sensory, cognitive, and social risks. Experimental drugs and supplements should not displace those interventions.