Pace of Aging Tests Explained

You sent off a finger-prick of blood, waited a few weeks, and got back a number like 1.05. No units. No age in years. Just a decimal that looks like it wants to be a ratio. Which is a slightly strange thing to have paid for. It’s also, it turns out, the product of a small New Zealand city whose residents kept showing up for checkups for decades. Since the question underneath is about living longer and better, Humanic Blue is looking at where that decimal comes from.

That decimal is a rate estimate. It does not report an age in years. The blunt version comes from the lab that built it: this is a speedometer, not an odometer. An odometer tells you how many miles the car has done. A speedometer tells you how fast it’s moving right now.

So the number is biological years accrued per calendar year. The scale was set so that 1.0 equals the average pace measured in roughly a thousand New Zealanders born in 1972–73 and tracked from age 26 to 45 by the Dunedin Study. Above 1.0 is faster than that group. Below it is slower.

A 1.05 means about five percent faster than the reference group; a 0.9, about ten percent slower. None of that is the same thing as a “biological age” of 38 or 52.

The question is how one blood sample can estimate a pattern researchers originally observed over twenty years—and what that estimate can tell an individual.

What does a pace of aging score of 1.05 actually mean?

The Duke laboratory behind DunedinPACE calls it a speedometer: one blood sample estimates how fast aging-related physiology is changing. A value of 1.0 is not a medical definition of healthy. It is the average pace in the original Dunedin reference group.

Age complicates the dashboard. A 2026 preprint pooling nearly 38,000 people suggests expected scores rise with age and differ by sex. In its worked examples, 1.05 was ordinary for an 80-year-old woman, while 0.99 was faster than expected for a 25-year-old man. The age-norming method has not yet been peer reviewed, so its exact percentiles are preliminary. The durable lesson is simpler: the same decimal can mean different things at 25 and 80. The distribution and age-norming calculations

How did a birth cohort in New Zealand become the ruler?

The Dunedin Study enrolled about 1,000 people—91% of eligible babies born in one New Zealand city in 1972–73—and has followed them since age three. Researchers repeatedly measured ordinary indicators across the heart, lungs, kidneys, liver, metabolism, immune system, teeth and other systems.

By age 45, the slowest measured pace was 0.40 biological years per calendar year and the fastest was 2.44. Faster agers also tended to show more cognitive difficulty, brain-aging signs and sensory-motor decline. The score is therefore anchored to repeated human function and physiology, not a molecular mist someone found behind the laboratory fridge. Its limitation is equally clear: one unusually complete city cohort became the original ruler for everyone else.

How does a blood sample stand in for twenty years of measurements?

Waiting twenty years is a tough consumer proposition. So the team trained a model using chemical marks on DNA to estimate each person’s measured rate of change from one blood sample. DunedinPACE uses 173 DNA sites selected partly for reliable measurement—an unusual acronym that actually explains the product: Pace of Aging Calculated from the Epigenome.

Repeat runs of the same sample agree closely. Fresh samples from the same living person are a trickier customer. Small 2026 studies found that meals, stress, altitude and pollution could move clocks over hours or days, with DunedinPACE among the less stable for these short-term biological repeats. A laboratory can measure one tube precisely while the person attached to the next tube keeps having a life. The technical and biological reliability estimates

Is pace of aging the same as biological age?

No. Mixing the two up is the most common misread of these reports, and the 2015 paper draws the line: “Pace of Aging and Biological Age represent two different approaches to quantifying aging. Pace of Aging captures real-time longitudinal change in human physiology across multiple systems and is suitable for use in studies of within-individual change.”

The speedometer-and-odometer comparison is an analogy. Age clocks produce age-like estimates from patterns in biological measurements; pace clocks estimate a rate of change. Neither directly measures a single quantity called accumulated damage or all the changes occurring in a person’s body.

Clocks such as Horvath and Hannum were built mainly to estimate calendar age. PhenoAge and GrimAge incorporate health- or mortality-related information, but also report age-like years. A pace measure instead reports a rate. The two can disagree. You can carry a lot of accumulated wear and be aging slowly now. You can also look young on paper while aging fast. Belsky’s 2015 analysis found the correlation between the two in the Dunedin cohort was 0.38: related, not interchangeable.

One terminology trap: Google’s own autocomplete pairs “pace of aging” with a wrist-worn wearable brand. A wearable can’t read DNA methylation. If no blood or saliva left your body, whatever the app calls its number, it is not DunedinPACE and it isn’t benchmarked to the Dunedin cohort.

How would you explain the difference between an estimated age and an estimated rate of aging?

What does a faster DunedinPACE predict?

Across four outside cohorts in the UK and US, faster scores tracked with illness, disability and mortality. Some associations remained after accounting for GrimAge, although they weakened. Independent studies involving roughly 6,300 people across the US, Ireland and Northern Ireland also linked faster pace with mobility, grip, cognition and mortality, while a German cohort connected it with new metabolic syndrome.

That makes the evidence Supported for population prediction across several countries and research teams. It does not make one person’s decimal a diagnosis, and none of these cohorts shows that deliberately lowering the score lowers risk. The cohorts and exact estimates

Can the number move, and does moving it matter?

This is where the speedometer has an advantage—and marketing sees an open highway.

In CALERIE, adults were assigned to calorie restriction or their usual diet for two years. DunedinPACE slowed 2–3%; PhenoAge and GrimAge did not. The trial did not measure whether that clock change prevented disease or death. A 2026 analysis pooling 51 intervention studies also found DunedinPACE was the clock most likely to move, although its senior author disclosed consulting fees from TruDiagnostic and another clock company.

The most revealing follow-up may be a 21.5-year Italian study of about 700 people. Baseline DunedinPACE predicted mortality; change in the score alone did not clearly do so, although combining the starting value and change improved models. A higher starting score was linked to earlier death. A later change in the score, considered on its own, was not clearly linked to mortality in that analysis. Promising for responsiveness; unknown for clinical benefit. The intervention and survival estimates

What would show that a slower pace score reflects a benefit beyond the score itself?

Where a pace of aging score falls short

The limits come largely from the people building and evaluating the measure:

• One original ruler: 1.0 was benchmarked to 45-year-old New Zealanders. International validation helps, but it does not change where its 1.0 reference value came from.
• Age and sex matter: scores tend to rise with age and may change differently in men and women. Current age-norming estimates are still preprint evidence.
• One reading is one reading: excellent repeatability on the same tube does not guarantee stability across fresh samples after ordinary short-term exposures.
• Prediction is not validation for treatment: researchers still lack consensus on when an aging biomarker is ready for clinical use. A good predictor can be a poor treatment target.
• Inventors have a commercial stake: DunedinPACE is licensed to TruDiagnostic, several authors are inventors, and Belsky has additional advisory roles. The algorithm is also public for researchers.

The conflicts do not erase the evidence. They make independent replication and modest claims particularly important.

What do you get from a consumer pace-of-aging test?

The main US consumer route identified here is TruDiagnostic’s blood test, which bundles DunedinPACE with age-style and organ reports. The company says the outputs should support rather than replace clinical judgment and should not be used alone for diagnosis or treatment. Mixing a rate with several ages in one dashboard is convenient—and almost custom-built for unit confusion.

The company’s mortality marketing also cited a 7.4-year German study whose published outcome was new metabolic syndrome, not death. Ask for the mortality analysis before repeating that claim. US regulatory history does not independently validate the score for diagnosis or treatment. Researchers, meanwhile, can use the public DunedinPACE code on methylation-array data.