Biological Age Testing: Do Epigenetic Age Clocks Actually Work?

What's measured

What Epigenetic Clocks Actually Measure

DNA methylation is a chemical tag on the genome (a methyl group on a cytosine base) that turns gene expression up or down. Specific methylation patterns change predictably with chronological age — the basis for every commercial epigenetic clock.

First-generation clocks (Horvath 2013, Hannum 2013) were trained to predict chronological age and do so within roughly ±3 years in healthy adults. Second-generation clocks (PhenoAge, GrimAge, DunedinPACE) were trained against mortality, smoking history, and clinical biomarkers — these correlate better with health outcomes than with chronological age itself.

GlycanAge takes a different approach: it measures the glycan (sugar) structures attached to immunoglobulin G (IgG) in your blood, which shift toward a pro-inflammatory pattern with age.

The honest evidence

Honest Evidence: What These Clocks Can and Can't Tell You

What they do well: predict all-cause mortality and age-related disease risk at the population level, particularly GrimAge and DunedinPACE. People with older biological ages than chronological ages have higher risk; the signal is real.

Where they fall short: commercial test-retest reliability is imperfect — two samples from the same person on the same day can come back 2–4 years apart depending on the clock. Within-person change over months of lifestyle intervention is often inside that noise band, which makes short-term scoring of supplements, exercise blocks, or NAD+ programs unreliable.

Whether you can move the number with intervention is partly answered: the CALERIE caloric-restriction trial showed measurable slowing of DunedinPACE with sustained restriction. Smaller trials of exercise, diet, and lifestyle have shown signals but are not yet definitive.

Brands

TruAge, GlycanAge, and the Major Clocks Compared

• ·TruAge / TruDiagnostic (PACE / OMICm). Uses DunedinPACE under license — the rate-of-aging clock with the strongest published track record. $229–$499 depending on panel. Blood spot kit, ships back, results in 3–5 weeks.
• ·GlycanAge. IgG glycan profile, not methylation. £214 (~$275). Strong link to systemic inflammation; responds to lifestyle changes in observational data; less mortality-validation than methylation clocks.
• ·Elysium Index. Methylation panel, consumer-facing, $499.
• ·MyDNAge. Horvath-style methylation panel. ~$299.
• ·Research-only clocks (Horvath 353, PhenoAge, GrimAge raw) are not directly sold — they're computed from methylation data and reported via TruDiagnostic and academic services.

Who should test

Who Actually Benefits From Testing

Reasonable use: a baseline at 40–55 alongside ApoB, Lp(a), VO2 max, and coronary calcium score, with a re-test no sooner than 12–18 months to track trend, not noise.

Less useful: quarterly testing to grade supplements, or testing in isolation without the cardiometabolic and fitness markers that actually drive longevity. A bad ApoB or Lp(a) matters far more for your next decade than a biological age that runs two years high.

The physician take

Physician Take — Trend, Don't Score

Treat the result as a directional metric. A biological age that drops 5 years after one month of supplements is almost certainly measurement noise. A consistent downward trend across 2–3 annual measurements alongside improving ApoB, VO2 max, and body composition is meaningful.

If you only fund one set of longevity diagnostics, prioritize coronary calcium score, ApoB, Lp(a), and a CPET VO2 max test before an epigenetic clock.