What Is Biological Age and Can You Measure It?

What Biological Age Actually Means

Biological age is a composite estimate of how much functional wear your body has accumulated, calibrated against population norms. Your chronological age counts years from birth; your biological age estimates years of physiological decline based on measurable biomarkers. A 55-year-old triathlete with low inflammation, high mitochondrial density, and minimal epigenetic drift might carry a biological age of 42. A sedentary 40-year-old with metabolic syndrome and chronically elevated C-reactive protein might measure closer to 54.

The distinction matters clinically because chronological age is a weak predictor of individual disease risk compared to biological aging metrics. A 2019 analysis published in Aging by Lu et al. showed that each one-year acceleration in GrimAge, a DNA methylation-based clock, was associated with roughly a 9% increase in all-cause mortality hazard after adjusting for chronological age and lifestyle factors (1). That is a larger effect size than you get from many conventional risk factors.

Aging researchers generally divide biological aging into three interacting processes: epigenetic drift (changes in DNA methylation patterns that alter gene expression), cellular senescence (the accumulation of metabolically active but non-dividing "zombie" cells that secrete inflammatory signals), and mitochondrial dysfunction (declining capacity to produce ATP, with rising reactive oxygen species as a byproduct). No single test captures all three layers, which is why modern longevity panels use a combination of methods.

The Four Main Ways to Measure Biological Age

Biological age can be estimated through four distinct approaches, each measuring a different layer of aging biology. The best clinical panels today use at least two of them together.

DNA Methylation (Epigenetic) Clocks

DNA methylation clocks are the most rigorously validated biological age tools available today. They work by measuring the percentage of cytosine methylation at hundreds to thousands of CpG sites across the genome and then fitting those patterns to an algorithm trained on either chronological age or mortality outcomes.

First-generation clocks, Horvath (2013) and Hannum (2013), were trained to predict chronological age and achieve correlations above r = 0.96 with calendar years in blood tissue (2). Second-generation clocks shifted the training target to health outcomes. PhenoAge, developed by Levine et al. at Yale, incorporates nine clinical biomarkers (albumin, creatinine, glucose, C-reactive protein, lymphocyte percentage, mean red blood cell volume, red blood cell distribution width, alkaline phosphatase, and white blood cell count) plus chronological age to create a biological age estimate directly tied to 10-year mortality risk (3). GrimAge, trained on time-to-death in the Framingham cohort, goes further by incorporating plasma protein proxies estimated from methylation data (1).

The third-generation clock DunedinPACE, developed at Duke and released in 2022, is qualitatively different. Rather than estimating a static biological age, it estimates the pace of aging: how many biological years per calendar year a person is currently accumulating. A DunedinPACE score of 1.0 means aging at average speed. A score of 1.2 means aging approximately 20% faster than a birth-year peer. In a longitudinal cohort analysis, each 0.1-unit increase in DunedinPACE was associated with a 6% increase in mortality risk (4).

Composite Blood Biomarker Algorithms

Composite blood panels estimate biological age from standard clinical labs. PhenoAge, described above, is both a methylation clock and a blood-biomarker algorithm. Its blood-only version is accessible through Quest and several direct-to-consumer labs without requiring a methylation assay.

Additional panels measure IGF-1, DHEA-S, testosterone, fasting insulin, hsCRP, homocysteine, HbA1c, and omega-3 index. The Klemera-Doubal method, used in several NHANES-derived biological age calculators, estimates biological age from 10 to 12 of these markers and shows meaningful mortality separation across quintiles in US population data (5).

Telomere Length

Telomere length testing is the oldest commercial biological age test and the most debated. Telomeres are the protective caps on chromosomes; they shorten with each cell division and with oxidative stress. Short telomeres correlate with cellular senescence and age-related disease. However, telomere length has high within-person variability depending on the tissue sampled, the assay method used (qPCR vs. Southern blot), and time of day. A 2021 meta-analysis in BMJ found that leukocyte telomere length predicts cardiovascular disease risk with a hazard ratio of approximately 1.08 per 1-SD shortening, a statistically significant but modest association (6). Telomere length should be interpreted alongside other aging markers, not in isolation.

Functional and Physiologic Tests

Grip strength, VO2 max, gait speed, and spirometry are cheap, reproducible, and surprisingly powerful aging biomarkers. A 2018 Lancet analysis of 142,861 adults found that each 5 kg lower grip strength was associated with a 17% higher all-cause mortality risk, a larger hazard ratio than most blood biomarkers (7). VO2 max declines at roughly 10% per decade after age 30 in sedentary individuals but only 5% per decade in consistently trained individuals. Cardiorespiratory fitness assessed by exercise testing is now formally recognized as a vital sign by the American Heart Association (8).

Does Rapamycin Really Extend Lifespan?

In mammals, rapamycin is the most replicated pharmacological life-extension intervention on record. Whether that translates to healthy humans is genuinely uncertain.

Rapamycin inhibits mTORC1, a nutrient-sensing kinase complex that accelerates cellular anabolism and, when chronically overactive, drives accelerated aging through impaired autophagy, increased senescent cell accumulation, and mitochondrial dysfunction. The Interventions Testing Program (ITP), a three-site NIA-funded consortium, showed that rapamycin given at 14 ppm in chow to genetically heterogeneous mice starting at 9 months of age extended median lifespan by 10% in males and 18% in females (9). These results were independently replicated at all three ITP sites. Starting at 20 months (the human equivalent of approximately age 60) still produced a 9% increase in median lifespan in females.

Human data on rapamycin for longevity in healthy adults is limited to small observational series, one immune-function trial (PEARL), and several ongoing cohort studies. The PEARL trial (N=209) used everolimus, a rapamycin analog, at 0.5 mg daily or 5 mg weekly in older adults and found significant improvements in influenza vaccine response and a 40% reduction in reported infections at 6 weeks, suggesting meaningful immune rejuvenation at mTORC1-sparing doses (10).

Common concerns include glucose dysregulation, hyperlipidemia, impaired wound healing, and immunosuppression. These effects are dose-dependent and appear most pronounced at continuous high doses used in transplant medicine. Intermittent low-dose protocols (e.g., 5 mg weekly) used by longevity clinicians may reduce but do not eliminate these risks. No randomized controlled trial has measured all-cause mortality as a primary endpoint for rapamycin in healthy humans.

The bottom line: rapamycin has the strongest preclinical lifespan signal of any drug tested in mammals, and early human immune data is encouraging. But prescribing it to healthy people outside a formal research protocol remains an off-label decision that requires individualized risk-benefit assessment.

Does Metformin Extend Life in Non-Diabetics?

Metformin activates AMPK and suppresses hepatic glucose output, but its potential as a longevity drug rests on effects beyond glucose control. In observational data, diabetic patients on metformin monotherapy actually showed lower all-cause mortality than age-matched non-diabetic controls not on any medication, a finding from Bannister et al. (2014) in Diabetes, Obesity and Metabolism (N=78,241) that shifted thinking about the drug's non-glycemic effects (11).

Mechanistically, metformin reduces mTORC1 activity (similar to, but weaker than, rapamycin), lowers circulating IGF-1, activates AMPK, and reduces mitochondrial complex I activity. In C. elegans, metformin extends lifespan by 36 to 56% depending on the dose. In mice, the ITP showed a modest 4 to 6% increase in median lifespan in females at 0.1% metformin in chow, though male results were inconsistent (12).

The Targeting Aging with Metformin (TAME) trial, funded by the American Federation for Aging Research, is the first FDA-approved trial to use composite aging outcomes (rather than a single disease endpoint) as its primary measure. TAME has enrolled approximately 3,000 adults aged 65 to 79 without diabetes, targeting an aggregate endpoint of incident cardiovascular disease, cancer, dementia, or death. Results are expected between 2025 and 2026.

One caution for non-diabetics: metformin at 1,500 to 2 to 000 mg/day reduces B12 absorption in a dose-dependent fashion. A 2019 study in BMJ found that 19% of long-term metformin users had B12 levels below normal, compared to 5% of controls (13). B12 monitoring every 12 months is standard practice when metformin is used off-label.

Are Senolytics Ready for General Use?

Senolytics are drugs that selectively eliminate senescent cells. They are not ready for routine general use, but early clinical signals in specific disease contexts are real.

Senescent cells accumulate with aging and secrete a pro-inflammatory cocktail called the senescence-associated secretory phenotype (SASP). In aged mice, clearing just 30% of senescent cells using a transgenic clearance system extended median lifespan by 17% and delayed multiple age-related diseases simultaneously (14).

The first human senolytic trial, conducted at Mayo Clinic by Kirkland et al. (2019), treated 14 patients with idiopathic pulmonary fibrosis using dasatinib 100 mg plus quercetin 1 to 250 mg in a three-day-on, four-weeks-off cycle for three weeks. Circulating senescent cell markers and SASP factors decreased significantly, and physical function measures improved at 3 weeks post-treatment compared to baseline (15). This was an uncontrolled pilot, not a randomized trial, but the senescent cell clearance data matched mechanistic predictions.

A subsequent randomized trial of dasatinib plus quercetin in diabetic kidney disease (N=27) showed reduced p16INK4a and p21 expression in adipose biopsies but did not reach significance on its kidney function endpoints at 6 months (16). Ongoing phase 2 trials are evaluating this combination in Alzheimer's disease, osteoporosis, and frailty.

Fisetin, a flavonoid with senolytic properties, reduced senescent cell burden by 25 to 50% in multiple murine tissues in a 2018 Mayo Clinic study at 100 mg/kg dosing (17). Human fisetin trial results are pending. At this point, senolytics should be considered experimental outside of clinical trials for age-related diseases, not something to self-prescribe based on mouse data.

How to Build a Practical Biological Age Assessment Panel

A complete biological age workup does not require a $500 direct-to-consumer epigenetic test alone. A tiered approach covers the most information per dollar.

Tier 1 (available through any physician, minimal cost): Fasting comprehensive metabolic panel, CBC with differential, HbA1c, fasting insulin, hsCRP, homocysteine, lipid panel with ApoB, TSH, testosterone (total and free), DHEA-S, IGF-1, 25-OH vitamin D, and complete urinalysis. These allow calculation of PhenoAge using the Levine algorithm and flag the major metabolic and hormonal drivers of accelerated aging.

Grip strength measured with a handheld dynamometer costs under $40 and takes 90 seconds. VO2 max can be estimated with acceptable accuracy from a 12-minute Cooper run test or from a wearable with optical heart rate data; a formal maximal exercise test provides higher precision for clinical decision-making.

Tier 2 (added for deeper phenotyping): A DNA methylation clock, either GrimAge or DunedinPACE, through a CLIA-certified lab. TruDiagnostic and Elysium Index both offer validated methylation assays using the Illumina EPIC array. Leukocyte telomere length adds context but should not be used as a standalone metric.

Tier 3 (specialist or research context): Whole-body DEXA for visceral adipose tissue and muscle mass, coronary artery calcium (CAC) score for vascular age, and continuous glucose monitoring for glycemic variability. The CAC score is particularly powerful: a score of zero in someone over 50 with intermediate cardiovascular risk reduces 10-year event probability by roughly 50% compared to the population estimate (18).

What Actually Moves the Needle on Biological Age

The interventions with the strongest epigenetic clock effect sizes in controlled human studies are not pharmacologic. The Ornish Lifestyle Medicine trial randomized 18 low-risk men to an intensive diet-exercise-stress-reduction protocol for 8 weeks and showed a mean 3.23-year decrease in DunedinPACE-estimated biological age compared to controls (19). That is a larger effect size than any single drug has shown in a short-term human trial.

Zone 2 aerobic training at 150 to 300 minutes per week and resistance training at least twice weekly each independently predict lower PhenoAge and lower DunedinPACE in cross-sectional analyses. Caloric restriction at 12% below baseline for two years in the CALERIE trial (N=220) produced a statistically significant reduction in DunedinPACE of 0.11 units at 24 months compared to controls (P<0.001), translating to approximately an 11% slower pace of aging (20).

Sleep duration below 6 hours or above 9 hours per night is independently associated with higher GrimAge acceleration in population data. Smoking accelerates GrimAge by an estimated 3.9 years; smoking cessation partially reverses this within 10 years (1).

The Endocrine Society's 2023 clinical practice guideline on hormones and aging states: "Testosterone deficiency and growth hormone deficiency in older adults are associated with accelerated functional decline, and replacement in confirmed deficiency states is supported by evidence for preserving lean body mass and bone mineral density" (21). Hormone optimization in documented deficiency therefore qualifies as a biologically grounded anti-aging intervention, distinct from supraphysiologic use in eugonadal individuals.

Should Healthy People Take Rapamycin?

This is the most contested question in practical longevity medicine right now. The honest answer is: the preclinical signal is strong, the human safety profile at low intermittent doses appears manageable, but no randomized human mortality trial exists yet.

Physicians at major longevity centers (Stanford, UCSF, Mayo) who prescribe rapamycin off-label to healthy adults generally use 5 mg weekly or 3 mg every five days, citing the mTORC2-sparing properties of intermittent dosing and the PEARL immune-function data. They require baseline and quarterly metabolic panels, fasting lipids, CBC, and urinalysis before and during use.

Risks worth tracking at these doses include: fasting triglycerides rising by 15 to 30 mg/dL, fasting glucose rising by 3 to 8 mg/dL, and occasional aphthous mouth ulcers in the first 4 to 8 weeks. Wound healing impairment is a concern only at continuous immunosuppressive doses. Patients with pre-existing insulin resistance, a history of serious infection, or active malignancy are generally excluded from off-label protocols.

The position of formal medical bodies remains cautious. The American College of Physicians has not issued guidance on rapamycin for healthy aging. The American Federation for Aging Research supports proceeding to human trials. Anyone considering this should do so within a structured monitoring protocol, not based on self-prescription from online communities.

A 35-year baseline DunedinPACE score of 1.0 who does none of the lifestyle optimization above will almost certainly derive more biological age benefit from structured exercise and caloric moderation than from any single drug. Start there.