longevity researchBiological age test: which one to take and how to read it
A biological age test estimates how fast your body is aging compared to your birthday. Here's how each major test works and which one is worth ordering.
A 45-year-old man with poor sleep, high ApoB, and untreated insulin resistance can have a biological age north of 55. Another 45-year-old, with the same birthday, can test at 38. The gap between those two men is not destiny. It is the cumulative effect of choices that biological age tests are designed to make visible.
The category has matured fast since Steve Horvath published the first DNA methylation clock in 2013 (Horvath 2013). What was once a research curiosity is now a consumer market with at least a dozen tests sold direct-to-buyer, ranging from $40 finger-prick kits to $500 saliva-based methylation panels. Most of them measure something real. Almost none of them tell you what to do next.
What a biological age test actually measures
There are four broad categories, and the differences matter.
DNA methylation clocks read the pattern of methyl groups attached to your DNA at specific sites called CpG islands. These patterns shift in predictable ways as you age, and the rate of shift correlates strongly with mortality risk. GrimAge (Lu et al. 2019) and DunedinPACE (Belsky et al. 2022) are the current state of the art, both validated against mortality and disease incidence in large cohorts.
Blood-based clocks like PhenoAge (Levine et al. 2018) and the Klemera-Doubal BioAge use standard lipid, metabolic, and inflammation markers from a routine blood draw. The math weights nine to twelve variables (albumin, creatinine, glucose, CRP, lymphocyte percentage, mean cell volume, red cell distribution width, alkaline phosphatase, white cell count) against chronological age to produce a biological age estimate.
Telomere length tests measure the protective caps on chromosomes. The biology is real but the test-retest variability is high enough that single measurements are not useful for tracking change. The category has lost favor in research circles for individual prognostication.
Functional age tests ask the body to perform: VO2 max, grip strength, balance time on one leg, sit-to-stand reps. These correlate strongly with mortality but require equipment and effort that most consumer tests skip.
The practical implication: methylation is the gold standard for prediction strength, blood-based clocks are the best cost-to-signal compromise, and functional tests should be folded in as a free sanity check.
What "GrimAge" and "DunedinPACE" are predicting
The strongest validation data sits with these two methylation clocks, and they predict different things.
GrimAge predicts time to death. It was trained directly on mortality and lifespan, which is why it correlates harder with all-cause mortality than any other epigenetic clock. A 2019 Aging paper showed each one-standard-deviation reduction in GrimAge corresponded to a 27% lower mortality risk in the test cohort (Lu et al. 2019).
DunedinPACE predicts the speed of aging, measured in years of biological aging per year of chronological aging. A score of 1.0 means you are aging at population norm. A score of 1.2 means you are aging 20% faster. A score of 0.85 means you are aging slower than your peers. The Dunedin team published evidence that the score is modifiable in middle-aged adults through caloric restriction (Belsky et al. 2022), which is the cleanest published evidence that biological age is not fixed.
The two clocks correlate with each other but answer different questions. GrimAge tells you where you stand. DunedinPACE tells you which direction you are moving.
The blood-based shortcut: PhenoAge
For most men, the right entry point is not a $300 methylation kit. It is a standard blood panel run through the PhenoAge calculator.
Morgan Levine's 2018 paper (Levine et al. 2018) introduced PhenoAge using nine biomarkers from a routine annual physical: albumin, creatinine, glucose, log-CRP, lymphocyte percent, mean cell volume, red cell distribution width, alkaline phosphatase, and white blood cell count, plus chronological age. The formula produces a biological age that correlates roughly 0.7 with GrimAge across published cohorts, which is impressive given the input data costs about $60 to generate.
The actionable variables PhenoAge weights heaviest are glucose, CRP, and creatinine. Men with elevated fasting glucose (see fasting glucose optimization), elevated inflammation, or kidney stress will see PhenoAge run ahead of their chronological age, and these are the same three buckets that high-cost methylation tests typically flag.
Free online calculators (Yale's Aging.AI, Bioage Labs, MyDNAge) take a standard CMP and CBC and produce a PhenoAge number in under a minute. The signal is real, the cost is trivial, and the inputs are already in most men's chart.
A biological age number without a plan is a vanity metric. With a plan, it's the only longevity feedback loop that gives you a quarterly signal instead of waiting for disease.
How to actually use the result
A biological age number without a downstream protocol is a vanity metric. The framework that produces useful behavior change has four parts.
Establish a baseline. Test once under representative conditions: fasted, well-slept, no acute illness, no training in the 24 hours prior. This number is your reference point, not a verdict.
Identify the dominant lever. If PhenoAge runs hot, look at which biomarker is pulling the average up. Glucose and CRP point to metabolic and inflammatory drivers. Creatinine elevation in lean men can be benign muscle mass, but in sedentary men it can indicate kidney strain. Alkaline phosphatase elevations track liver and bone turnover. The specific outlier tells you where the leverage is.
Intervene on one thing for 90 days. Biological age responds to changes the same way biomarkers do, on weeks-to-months timescales. A 90-day window is long enough to register a real shift on methylation and blood-based tests, short enough to maintain compliance.
Retest under the same conditions. Same lab, same kit version, same time-of-day. Methylation tests have inherent technical variability of about 1.5 years on retest of the same sample. Treat any change smaller than 2 years as noise. Treat changes of 3+ years as signal that the intervention is working, but verify with a third test 6 months later.
The men who get the most out of biological age testing treat it the way endurance athletes treat lactate threshold testing. Not as a verdict on their fitness, but as a periodic calibration of where to push next. For a structured walkthrough of which levers move biological age the fastest, see how to reduce biological age and the biological age calculator.
Common pitfalls
A few patterns trip up first-time testers reliably.
Single-test fixation. One result, especially from a methylation test, can be off by 1 to 2 years for reasons that have nothing to do with aging. Treating a single number as gospel produces overreaction in both directions.
Time-of-day artifacts. Cortisol, glucose, and inflammation markers swing through the day. Morning fasted draws are the only condition under which blood-based clocks produce stable, comparable numbers.
Acute confounders. A bout of intense training 24 hours before a draw can elevate CRP, creatinine, and white cell count enough to add 2 to 3 years to PhenoAge. So can a vaccine, a virus, or a single night of bad sleep.
Brand shopping. Different methylation clocks use different CpG site panels and different reference cohorts. A GrimAge of 42 and a TruDiagnostic estimate of 39 are not contradictory. They are different instruments. Pick one provider and stay with them for longitudinal comparisons.
A practical protocol
- First test, low cost. Pull a standard CMP + CBC + hsCRP from any lab, then run the numbers through PhenoAge using a calculator like Aging.AI or BioAge Labs. Cost: $60 to $120 depending on insurance.
- If you want the gold standard, layer on methylation. A single GrimAge or DunedinPACE test from TruDiagnostic, Elysium, or myDNAge runs $250 to $500. Treat this as the once-a-year deeper read.
- Run a 90-day intervention. Pick the single biggest lever (sleep, metabolic, training, alcohol). One variable at a time is the only way to attribute change to a cause.
- Retest at 6 months. Same lab, same time of day, same fasting state. Compare delta. Treat anything within 2 years of baseline as noise.
- Hold the protocol or pivot. If the number moved in the right direction by more than 2 years, the lever is real for you. If it didn't, the next lever is the one to test.
The men who treat biological age testing as a quarterly steering signal, rather than a one-time vanity check, get more out of it than the average longevity-clinic patient. The number is feedback. The plan is what produces the change.
Want a structured view of your biology before ordering a test? Start with the PrimalPrime biological age calculator to map where your levers actually are.