Biological age: what your blood test says about how fast you're aging.

The problem with chronological age

Two 50-year-olds walk into a clinic. One has a fasting glucose of 82, hs-CRP of 0.3, perfect lipids, and a resting heart rate of 56. The other has a fasting glucose of 118, hs-CRP of 4.2, pre-diabetic HbA1c, and elevated liver enzymes. They are the same chronological age. They are not aging at the same rate.

This observation — that people of the same age can be in vastly different states of health — launched the field of biological age estimation. The goal: develop quantitative measures of how fast a person is aging, using biomarkers rather than birthdays[1].

The science has matured rapidly. Today, validated algorithms can estimate your biological age from standard blood biomarkers with enough accuracy to predict mortality, disease onset, and functional decline years in advance. More importantly, biological age is modifiable — interventions that improve your biomarkers will reduce your biological age, and that reduction predicts real improvements in health outcomes.

How biological age is measured

There are two broad categories of biological age clocks: those based on blood biomarkers (clinical chemistry clocks) and those based on DNA methylation patterns (epigenetic clocks). Both have strengths. They measure different but overlapping dimensions of aging.

Blood biomarker clocks

These algorithms use standard clinical blood markers — the kind you get from a routine blood panel — to estimate biological age. Their advantage is accessibility: you do not need specialized testing.

Epigenetic clocks

These use patterns of DNA methylation (chemical modifications to DNA that change with age) to estimate biological age or aging pace. They require specialized laboratory testing (methylation arrays) but are generally considered more precise.

KDM Biological Age

The Klemera-Doubal method (KDM) was one of the first rigorous approaches to estimating biological age from clinical biomarkers. Originally published in 2006, it uses a statistical framework to combine multiple biomarkers, weighting each by how strongly it correlates with chronological age and mortality[2].

A widely used implementation of KDM Biological Age uses markers from NHANES data including albumin, alkaline phosphatase, blood urea nitrogen, creatinine, C-reactive protein, HbA1c, systolic blood pressure, total cholesterol, and forced expiratory volume (FEV1). In validation studies, each year of KDM age acceleration is associated with increased risk of mortality, cardiovascular disease, and age-related morbidity[1].

PhenoAge

PhenoAge was developed by Morgan Levine at Yale in 2018 and represents a significant advance in blood-based biological age estimation. Rather than being trained to predict chronological age, PhenoAge was trained in two stages: first, a composite score was derived from nine blood biomarkers that best predicted mortality risk (using NHANES III data, n=9,926); then this composite was recalibrated against chronological age to produce a biological age estimate[1].

The nine biomarkers used in PhenoAge:

• Albumin
• Creatinine
• Glucose
• C-reactive protein (log-transformed)
• Lymphocyte percentage
• Mean cell volume (MCV)
• Red blood cell distribution width (RDW)
• Alkaline phosphatase
• White blood cell count

PhenoAge was validated in NHANES IV (n=11,432) and showed strong prediction of all-cause mortality, cardiovascular mortality, cancer mortality, and diabetes incidence. Each year of PhenoAge acceleration (biological age exceeding chronological age) was associated with[1]:

• 9% increased risk of all-cause mortality
• 9% increased risk of cardiovascular mortality
• 7% increased risk of cancer mortality
• 20% increased risk of diabetes

Epigenetic clocks: GrimAge and DunedinPACE

GrimAge

GrimAge, developed by Ake Lu and Steve Horvath in 2019, is an epigenetic clock that predicts mortality with remarkable accuracy. Unlike first-generation clocks trained to predict chronological age, GrimAge was trained on DNA methylation surrogates of smoking pack-years and seven plasma protein levels associated with mortality[3].

GrimAge acceleration (being epigenetically older than your chronological age) has been associated with increased risk of coronary heart disease, cancer, all-cause mortality, and reduced time to death. It is currently considered one of the most predictive epigenetic clocks available.

DunedinPACE

DunedinPACE (Pace of Aging Computed from the Epigenome) was developed from the Dunedin Study — a longitudinal birth cohort study that has followed 1,037 individuals born in Dunedin, New Zealand in 1972-1973, with biological measurements at ages 26, 32, 38, and 45[4].

What makes DunedinPACE unique is that it measures the pace of aging — how fast you are aging right now — rather than cumulative biological age. A DunedinPACE of 1.0 means you are aging at the average rate (one biological year per calendar year). A DunedinPACE of 0.85 means you are aging at 85% of the average rate. A DunedinPACE of 1.2 means you are aging 20% faster than average.

Because it measures pace rather than cumulative state, DunedinPACE is more responsive to interventions. A caloric restriction trial (CALERIE) showed that participants randomized to 25% caloric restriction had a significantly slower DunedinPACE after 2 years compared to controls[5].

Which biomarkers drive biological age?

Across all validated clocks, certain biomarkers consistently emerge as the strongest predictors of aging rate:

The pattern is clear: chronic low-grade inflammation (inflammaging), metabolic dysfunction (particularly insulin resistance), declining organ function (liver, kidneys), and immune deterioration (immunosenescence) are the primary drivers of biological age acceleration[6].

Can you reduce your biological age?

Yes. And the evidence is growing stronger.

Caloric restriction and weight loss

The CALERIE trial — the first controlled caloric restriction study in healthy, non-obese humans — randomized participants to 25% caloric restriction for 2 years. The intervention group showed a significantly slower pace of aging by DunedinPACE, equating to approximately a 2-3% reduction in aging pace[5].

In overweight and obese individuals, weight loss interventions reduce PhenoAge by approximately 1-2 years, driven primarily by improvements in glucose, CRP, and albumin[7].

Exercise

Both aerobic exercise and resistance training are associated with younger biological age. A meta-analysis found that physically active individuals have approximately 0.5-2 years younger biological age compared to sedentary controls, depending on the clock used[8].

The mechanisms are multiple: exercise reduces CRP, improves glucose metabolism, maintains muscle mass (which supports albumin levels), improves kidney function markers, and enhances immune function. These are precisely the biomarkers that drive biological age calculations.

Sleep

Short sleep duration and poor sleep quality are associated with accelerated biological aging. A study using PhenoAge found that each additional hour of sleep (up to approximately 7-8 hours) was associated with approximately 0.5 years younger biological age[7].

Specific biomarker interventions

Because biological age is calculated from specific biomarkers, improving those biomarkers directly reduces biological age:

• Reduce hs-CRP: Anti-inflammatory diet (Mediterranean), omega-3 fatty acids, weight loss, regular exercise. CRP is heavily weighted in PhenoAge.
• Optimize glucose/HbA1c: Reduce refined carbohydrates, increase fiber, regular exercise, adequate sleep, weight management.
• Maintain albumin: Adequate protein intake (especially important in adults over 50), resistance training to maintain muscle mass.
• Support kidney function: Adequate hydration, moderate sodium intake, blood pressure management.
• Reduce RDW: Address iron, B12, and folate deficiencies. Reduce chronic inflammation.

Lipa's ensemble approach

No single biological age clock is perfect. KDM and PhenoAge use different biomarker sets and different statistical methods. They capture overlapping but distinct dimensions of aging. A person can have an accelerated PhenoAge (driven by inflammation and immune markers) but a normal KDM age (driven more by metabolic and organ function markers), or vice versa.

Lipa uses an ensemble approach: we apply multiple validated algorithms to the biomarkers available on your blood panel and present both the individual estimates and a weighted composite. This approach provides several advantages:

• Robustness: Less sensitive to a single outlier marker or clock-specific bias.
• System-level insight: You can see whether your immune aging, metabolic aging, and organ function aging are aligned or divergent.
• Sensitivity to change: Different clocks respond to different interventions. The ensemble captures a broader range of improvement.
• Actionability: We show you which specific markers are driving your biological age in each clock, giving you targeted areas for improvement.

The Dunedin Study: why longitudinal data matters

The Dunedin Study is one of the most important studies in aging science. By following the same 1,037 people from birth with detailed biological measurements at multiple timepoints, it revealed something that cross-sectional studies cannot: that the pace of aging varies dramatically between individuals even in young adulthood[9].

By age 38, some participants were aging at a rate of nearly 3 biological years per calendar year, while others were barely aging at all. The fast agers not only looked older (assessed by facial photographs rated by independent panels) but also had worse physical function, worse cognitive function, and more signs of cardiovascular disease — at age 38.

This finding is both alarming and empowering: it means that biological aging is not a fixed trajectory. It is variable, and the variation starts early enough to intervene.

Limitations and caveats

Biological age estimation is a powerful tool, but it has limitations:

• Acute illness: A temporary infection can spike CRP and WBC count, artificially inflating biological age. Test when healthy.
• Single timepoints: A single biological age estimate is a snapshot. Trends across multiple tests are far more informative.
• Population calibration: Most clocks were calibrated on US/European populations. Validity in other populations is less established[3].
• Precision: Blood-based clocks have a standard error of approximately 4-6 years. This means a PhenoAge of 42 in someone who is 40 is within the margin of error; a PhenoAge of 50 is not.
• They are estimators, not measurements: Biological age is a statistical prediction based on biomarker patterns. It is not like measuring height.

What the future looks like

The aging clock field is evolving rapidly. Current developments include:

• Organ-specific clocks: Estimates of brain age, heart age, liver age, and immune age from blood biomarkers and imaging[10]
• Proteomic clocks: Using thousands of plasma proteins rather than nine biomarkers, potentially offering much higher resolution[11]
• Integration with wearable data: Heart rate variability, activity, sleep, and blood biomarkers combined
• Intervention tracking: Using biological age clocks as primary endpoints in clinical trials of longevity interventions

The TAME trial (Targeting Aging with Metformin) and other aging-intervention trials are using biological age clocks as outcome measures — which will generate the first randomized evidence for pharmacological biological age reduction in humans[12].

Practical recommendations

1. Get a comprehensive blood panel that includes the PhenoAge biomarkers: albumin, creatinine, glucose, CRP, lymphocyte %, MCV, RDW, alkaline phosphatase, WBC count. Most standard panels include these.
2. Test regularly — twice per year allows tracking of trends and seasonal variation.
3. Focus on the big levers: inflammation (CRP), glucose metabolism, body composition, exercise, and sleep account for the majority of modifiable biological age variation.
4. Do not over-interpret a single result. Trends are more meaningful than any individual number.
5. Treat biological age as a compass, not a GPS. It shows direction (aging faster or slower than expected) rather than a precise destination.