LetsGo
Login
Bioage11/16/202612 min read

What does biological age measure, and which markers actually track healthy aging?

Second-generation biological age clocks like PhenoAge and GrimAge predict human mortality and disease risk better than chronological age, but physiological measures like VO2max and ApoB remain more directly actionable.

A guide to epigenetic clocks, composite blood algorithms, and functional physiological benchmarks.

LetsGo Editorial•Public LetsGo editorial
What does biological age measure, and which markers actually track healthy aging?
Bioage

LetsGo Care Editorial

What does biological age measure, and which markers actually track healthy aging?

A single biological age number looks definitive on a lab report, yet two different tests taken on the same morning can disagree by a decade.

Chronological age counts the years since birth; biological age estimates the cumulative decline in tissue structure and physiological reserve across organ systems. In human clinical research, biological age is not a single cellular readout. It is an algorithmic score derived from either epigenetic DNA methylation patterns, clinical blood chemistry panels, plasma proteomics, or functional physical assessments. First-generation epigenetic clocks were trained to predict chronological years, whereas second- and third-generation clocks predict time-to-death, time-to-disease, or the current rate of multiorgan deterioration. Understanding how these algorithms are constructed reveals why an epigenetic clock score can improve without an obvious change in physical performance, and why established physiological and metabolic markers remain the primary targets in clinical practice.

Key facts

  • First-generation epigenetic clocks (Horvath, Hannum) predict chronological age from CpG site methylation, whereas second-generation clocks (PhenoAge, GrimAge) predict human morbidity and all-cause mortality.
  • DunedinPACE is a third-generation blood DNA methylation measure trained on a longitudinal birth cohort to quantify the current pace of multiorgan decline per chronological year.
  • In the 2-year human CALERIE randomized controlled trial (220 non-obese adults), caloric restriction slowed the DunedinPACE aging rate by 2% to 3%, even though first- and second-generation clock ages did not significantly shift.
  • White blood cell composition shifts during acute infection or exercise can alter blood-based DNA methylation readouts by several years without reflecting true tissue rejuvenation.
  • Established clinical biomarkers—specifically VO2max, grip strength, ApoB, HbA1c, and blood pressure—have stronger, directly validated causal links to human healthspan outcomes than standalone methylation scores.

Who this is for

Adults aged 40 to 65 evaluating commercial biological age tests, standard longevity blood panels, or starting point physiological assessments to guide lifestyle and medical interventions.

In this article

  1. What it is
  2. Why it is used
  3. How it helps
  4. Who should try it
  5. Measurements that show improvement
  6. Products and services

What it is

Biological age (often shortened to BioAge) is a statistical summary of physiological wear and tear compared against a reference population. Because no single molecule captures the aging of all tissues, researchers use machine-learning models—most commonly elastic net regression—to compress hundreds or thousands of biological variables into an age equivalent or a rate-of-aging multiplier.

In human studies, four main categories of biological age algorithms are used:

  1. First-generation epigenetic clocks: The Horvath multi-tissue clock (353 CpG dinucleotide sites) and Hannum blood clock (71 CpG sites) measure the addition or removal of methyl groups on cytosine bases in DNA via DNA methyltransferases (DNMT1, DNMT3A, DNMT3B) and TET demethylating enzymes. These clocks were trained strictly to predict chronological age. Paradoxically, a clock that perfectly predicts chronological age in everyone would be useless for measuring biological variation, because it would ignore differences in health status.

  2. Second-generation mortality clocks: PhenoAge (Levine) and GrimAge (Lu) solved this limitation by training DNA methylation models on clinical phenotypes and mortality data rather than birth certificates. PhenoAge first constructs a clinical 'phenotypic age' from nine routine blood markers—albumin, creatinine, serum glucose, high-sensitivity C-reactive protein (hs-CRP), lymphocyte percentage, mean cell volume, red cell distribution width (RDW), alkaline phosphatase, and white blood cell count—plus chronological age, and then maps 513 CpG sites to that score. GrimAge uses 1,030 CpG sites that act as DNA methylation surrogates for smoking pack-years and seven plasma proteins (adrenomedullin, beta-2 microglobulin, cystatin C, GDF-15, leptin, PAI-1, and TIMP-1). In large human observational cohorts spanning thousands of participants, a higher GrimAge acceleration (biological age minus chronological age) strongly predicts time to coronary heart disease, cancer, and all-cause mortality.

  3. Third-generation pace-of-aging clocks: DunedinPACE was developed using longitudinal human data from the Dunedin Study cohort, tracking 19 biomarkers across cardiovascular, metabolic, renal, hepatic, immune, dental, and pulmonary systems over two decades in adults followed from age 26 to 45. Instead of estimating how old a person looks biologically, DunedinPACE estimates how fast their organs are deteriorating right now, expressed as biological years per chronological year (where 1.0 is average).

  4. Alternative molecular and functional clocks: GlycanAge measures immunoglobulin G (IgG) glycosylation patterns, which reflect systemic chronic inflammation and inflammaging. Telomere length assays measure TTAGGG nucleotide repeats at chromosome ends, though telomere length has high assay variability and weaker predictive power for mortality in humans than epigenetic clocks. Functional physiological composites combine cardiorespiratory fitness (VO2max), neuromuscular strength, arterial stiffness, and metabolic panels into a clinical BioAge.

Why it is used

In clinical trials and preventive medicine, chronological age is an unmodifiable variable, and waiting 20 to 30 years to measure hard clinical endpoints like myocardial infarction, stroke, dementia, or death makes testing preventive interventions in midlife adults impractical. Biological age markers are used as surrogate endpoints to evaluate whether an intervention alters the trajectory of physiological decline within months rather than decades.

For example, academic clinical trial centres run short multimodal feasibility trials—such as 8-week programmes combining structured exercise, targeted nutritional compounds, and health coaching in adults aged 50 to 80, or pharmacological trials testing sirtuin-6 (SIRT6) activators in pre-frail older adults—to see whether composite aging biomarkers and functional frailty scores respond before launching larger outcome trials. In a clinical setting, an elevated biological age relative to chronological age identifies hidden subclinical dysregulation (such as low-grade inflammation, declining renal filtration, or insulin resistance) when individual lab values still sit near the upper edge of standard reference ranges.

However, a critical non-obvious distinction governs how physicians interpret these numbers: an improvement in a surrogate clock does not automatically prove a reduction in disease risk unless the underlying causal pathways have improved. Because whole-blood epigenetic clocks read DNA extracted from leukocytes, any transient shift in the ratio of naive T cells, exhausted T cells, monocytes, and neutrophils will change the methylation score. A person recovering from a viral illness or changing their acute training load can see a 3-to-5-year swing in a first-generation blood methylation clock purely from leukocyte trafficking, even though the structural integrity of their vascular endothelium, myocardium, and skeletal muscle has not changed. Conversely, proven interventions can improve hard physiological function without moving older epigenetic clocks.

How it helps

Tracking biological age and its component markers helps stratify starting point risk and measure the physiological response to lifestyle, nutritional, or pharmacological interventions across specific cellular pathways.

In human randomized controlled trials, lifestyle modifications have demonstrated measurable effects on third-generation clocks and clinical composites. The CALERIE trial, a 2-year randomized controlled trial in 220 healthy, non-obese adults aged 21 to 50 assigned to 25% caloric restriction (achieving roughly 12% actual restriction) versus ad libitum intake, showed a 2% to 3% slowing in the DunedinPACE pace of aging, alongside improvements in cardiometabolic markers (lower hs-CRP, improved insulin sensitivity, and reduced blood pressure). Notably, standard first- and second-generation clocks (Horvath, Hannum, PhenoAge, GrimAge) did not show statistically significant changes in that same trial, illustrating that pace-of-aging measures are more sensitive to medium-term metabolic interventions in humans than cumulative state clocks.

Smaller human trials have tested multimodal diet, sleep, exercise, and relaxation guides over 8 weeks (such as a pilot RCT in 43 healthy adult males aged 50 to 72), reporting an average reduction of roughly 2 to 3 years on the Horvath clock compared to controls. While encouraging, sample sizes in these short-duration lifestyle trials remain small, and long-term durability is not yet established.

At the mechanistic level, interventions that improve biological age markers typically act on conserved hallmarks of aging demonstrated in both animal models and human tissues:

  • Nutrient sensing and metabolic signaling: Caloric restriction, protein moderation, and exercise downregulate chronic mTORC1 signaling and insulin/IGF-1 pathways while activating AMPK and sirtuin deacetylases (SIRT1, SIRT3, SIRT6), which support DNA repair and mitochondrial biogenesis.
  • Senescence and inflammaging: Visceral adiposity and cellular damage accumulate senescent cells that secrete pro-inflammatory cytokines (IL-6, TNF-alpha, IL-1beta) and matrix metalloproteinases (the senescence-associated secretory phenotype, or SASP). Lowering visceral fat and improving cardiorespiratory fitness reduce systemic hs-CRP and shift IgG glycosylation toward less inflammatory profiles in humans.
  • Epigenetic maintenance: Exercise alters DNA methylation patterns specifically within human skeletal muscle tissue at promoter regions of genes governing mitochondrial oxidative phosphorylation (such as PGC-1alpha), directly increasing mitochondrial density and aerobic capacity.

Who should try it

Measuring biological age is most useful for high-agency adults aged 40 to 65 who want an objective starting point before starting a structured exercise, nutrition, sleep, or medical prevention programme, and who understand the difference between exploratory molecular clocks and validated clinical biomarkers.

A staged approach avoids the common trap of paying for expensive molecular assays while ignoring foundational physiology:

  • Tier 1 (Everyone in midlife): Start with clinical blood chemistry (including the nine PhenoAge inputs, lipid particles, and glycemic markers) paired with objective functional testing (VO2max, grip strength, body composition via DEXA, and blood pressure). These metrics have established reference ranges, low measurement error, and direct clinical playbooks.
  • Tier 2 (Data-driven trackers): Add a second- or third-generation epigenetic test (such as GrimAge or DunedinPACE) or a glycomics test if you are already addressing Tier 1 markers and want to track longitudinal trends in systemic aging rate over 6- to 12-month intervals.

Trade-offs and limitations must be weighed plainly. Commercial epigenetic and proteomic biological age tests cost several hundred dollars per test and are rarely covered by insurance. Direct-to-consumer methylation tests have technical test-retest variability (often 1 to 3 years of noise on standard clocks, though newer computational methods using principal components analysis—PC clocks—substantially reduce technical noise). Epigenetic tests should not be taken during or immediately after an acute infection, surgery, or vaccination, as acute immune activation skews leukocyte populations and invalidates the starting point. Furthermore, high anxiety around health metrics is a practical contraindication; an unexpectedly high epigenetic age score is not a clinical diagnosis and should never prompt unverified polypharmacy. The largest opportunity cost is spending capital on repeated methylation panels instead of investing in DEXA scans, cardiopulmonary exercise testing (CPET), or structured coaching.

Measurements that show improvement

Because no single clock captures every tissue, a rigorous healthy aging assessment tracks a matrix of molecular, cardiometabolic, and functional markers. Meaningful improvement requires movement in the right direction across multiple independent domains:

Cardiorespiratory and Neuromuscular Function (Established Human Outcomes):

  • VO2max: Measured via cardiopulmonary exercise testing in mL/kg/min. Higher values strongly correlate with lower all-cause mortality across large human cohorts (moving from the bottom 25th percentile to above average is associated with a roughly 50% or greater relative reduction in mortality risk). Direction that matters: increase or slower age-related decline.
  • Grip strength and leg press strength: Measured by handheld dynamometry (kg) and force plates. Reflects neuromuscular integrity and predicts disability and mortality in older adults. Direction that matters: increase or maintenance.
  • Resting heart rate (RHR) and heart rate recovery (HRR): Lower RHR and faster drop in heart rate 1 minute post-peak exercise (>18-20 bpm drop) indicate stronger vagal tone and autonomic health.
  • Heart rate variability (HRV, specifically nocturnal RMSSD): Reflects parasympathetic nervous system recovery. Direction that matters: stable or upward trend relative to personal starting point.

Cardiometabolic and Organ Function (Clinical Blood Markers):

  • Apolipoprotein B (ApoB): Measures the total number of atherogenic lipoprotein particles (LDL, VLDL, IDL) in mg/dL or g/L. Directly causal in human atherosclerosis. Direction that matters: decrease to target based on cardiovascular risk.
  • HbA1c and Fasting Insulin: Measure 90-day average glycemia and hepatic/peripheral insulin sensitivity. Lowering elevated fasting insulin and HbA1c reduces glycation end-products and vascular damage.
  • High-sensitivity C-reactive protein (hs-CRP): Measures systemic vascular and hepatic inflammation in mg/L. Values persistently below 1.0 mg/L indicate low systemic inflammatory burden. Direction that matters: decrease.
  • Cystatin C (with eGFR): A filtration marker less confounded by muscle mass than serum creatinine, also weighted heavily as a methylation surrogate in GrimAge. Direction that matters: lower Cystatin C / stable eGFR.
  • Red cell distribution width (RDW) and Albumin: Two of the heaviest-weighted inputs in the clinical PhenoAge calculation. Lower RDW (reflecting uniform erythropoiesis and low oxidative/inflammatory stress) and higher normal serum albumin correlate with lower biological age.

Body Composition (Imaging):

  • Visceral adipose tissue (VAT) mass/volume and Appendicular Lean Mass Index (ALMI): Measured via DEXA scan. Lowering VAT directly reduces inflammatory adipokines, while maintaining ALMI prevents sarcopenia and preserves metabolic disposal of glucose.

Epigenetic and Molecular Clocks (Surrogate Composites):

  • DunedinPACE: Expressed in biological years per chronological year. Direction that matters: reduction below 1.0 (for example, moving from 1.05 to 0.92 over 6 to 12 months).
  • PhenoAge / GrimAge acceleration: Expressed in years relative to chronological age. Direction that matters: negative acceleration (biological age lower than chronological age).

Products and services

These LetsGo products and services sit next to this subject. They are not a prescription. A link is not evidence that the product causes the change discussed above.

Frequently asked questions

Can you take the PhenoAge test using a standard blood panel instead of a DNA methylation test?

Yes. The original PhenoAge algorithm developed by Levine et al. was built first as a clinical chemistry equation using nine routine blood biomarkers (albumin, creatinine, glucose, hs-CRP, lymphocyte percentage, mean cell volume, red cell distribution width, alkaline phosphatase, and white blood cell count) plus chronological age. That clinical score was then mapped onto DNA methylation sites. Calculating clinical PhenoAge from a standard fasting blood draw is inexpensive, highly reproducible, and directly shows which organ system or inflammatory marker is driving the score.

Why did my biological age go up after I started exercising harder?

Intense or unrecovered exercise causes transient muscle microtrauma, acute elevations in inflammatory cytokines (like IL-6 and CRP), and shifts in circulating white blood cell populations. If you draw blood for an epigenetic clock or a clinical PhenoAge panel within 24 to 72 hours of a heavy training block, the acute inflammatory and leukocyte response can artificially inflate your calculated biological age. Always standardize blood draws after 48 hours of rest or light recovery.

How often should biological age markers be retested?

Standard clinical blood markers (lipids, HbA1c, fasting insulin, hs-CRP, and clinical PhenoAge inputs) and functional tests (VO2max, grip strength) can be reassessed every 3 to 6 months to verify whether an exercise, nutrition, or medical guide is working. Epigenetic DNA methylation clocks change slowly and carry inherent assay variability, so retesting more frequently than every 6 to 12 months rarely provides actionable signal over technical noise.

Does a younger biological age on a test guarantee a longer lifespan?

No. Biological age scores are statistical probabilities derived from population cohorts, not individual guarantees. Furthermore, epigenetic clocks do not detect localized structural risks such as an unruptured cerebral aneurysm, a specific genetic cancer predisposition, or non-calcified coronary plaque. They must be paired with standard age-appropriate medical screening and imaging.

Practical takeaways

  • Prioritize functional and clinical markers (VO2max, grip strength, DEXA visceral fat and lean mass, ApoB, HbA1c, hs-CRP, and blood pressure) before spending money on commercial DNA methylation clocks.
  • Calculate your clinical PhenoAge from a routine comprehensive metabolic panel, complete blood count with differential, and hs-CRP to get an inexpensive, actionable starting point.
  • If you use an epigenetic test, choose a second-generation clock (GrimAge, PC-PhenoAge) or a third-generation pace clock (DunedinPACE) rather than a first-generation chronological clock.
  • Standardize testing conditions: draw blood fasted in the morning, avoid vigorous exercise for 48 hours prior, and postpone testing for at least 3 to 4 weeks after any acute viral or bacterial infection.
  • Re-evaluate interventions over 6- to 12-month windows rather than reacting to single-test fluctuations that may sit within the lab's margin of error.

Sources and evidence notes

  • Horvath S. Genome Biology (2013). Established the first multi-tissue human DNA methylation clock across 353 CpG sites predicting chronological age.
  • Levine ME et al. Aging (2018). Developed PhenoAge using nine clinical blood biomarkers and 513 CpG sites to predict human phenotypic aging and all-cause mortality.
  • Lu AT et al. Aging (2019). Developed GrimAge using DNA methylation surrogates for plasma proteins and smoking pack-years, demonstrating strong prediction of lifespan and healthspan in human cohorts.
  • Belsky DW et al. eLife (2022) & Waziry R et al. Nature Aging (2023). Developed the DunedinPACE pace-of-aging blood methylation algorithm and demonstrated in the human CALERIE randomized controlled trial that 2 years of caloric restriction slowed DunedinPACE by 2% to 3%.
  • Higgins-Chen AT et al. Nature Aging (2022). Showed that standard epigenetic clocks exhibit technical noise across replicate samples and introduced principal-component (PC) clocks to improve longitudinal test-retest reliability.
  • Mandsager K et al. JAMA Network Open (2018). Retrospective cohort study of 122,007 adults demonstrating a strong, inverse, graded association between cardiorespiratory fitness (VO2max) and long-term all-cause mortality with no observed upper limit of benefit.

Safety note

Blood draws carry minor risks of bruising, lightheadedness, or localized infection. Maximum cardiorespiratory exercise testing (VO2max) requires medical clearance in individuals with known cardiovascular, pulmonary, or metabolic disease, or those experiencing chest pain, unexplained shortness of breath, or dizziness.

Disclaimer

This article is for educational and informational purposes only and does not constitute medical advice, diagnosis, or treatment. Always consult a qualified physician or healthcare provider before starting any new exercise, nutrition, supplement, or diagnostic testing guide, or before making changes to prescribed medications.

Biological age is a valuable framework for compressing complex physiological data into a trackable metric, provided you understand what sits inside the algorithm. Second- and third-generation epigenetic clocks like GrimAge and DunedinPACE offer genuine insight into systemic mortality risk and the pace of multiorgan decline in humans, whereas clinical chemistry composites and direct physiological measurements—VO2max, muscle strength, body composition, and atherogenic lipids—tell you exactly what to fix. Using functional and clinical markers as your primary compass, and molecular clocks as a secondary longitudinal check, turns biological age from a novelty score into a rigorous tool for preventive health.