Skip to content

What the GrimAge Test Costs, and What You Get for It

Oak
A lab instrument sweeps a laser along a taut DNA strand studded with pale beads, lighting wells in a glowing glass grid below.

A GrimAge test costs roughly $300 to $500 as a consumer product. The Clock Foundation’s GrimAge test and resellers such as myagingtests.com sit near $325. TruDiagnostic’s TruAge panels, which report GrimAge alongside other clocks, run about $250 to $500 depending on the tier. All of these prices buy the same underlying measurement: an Illumina methylation array run on a blood or buccal sample, followed by a scoring step. The array is the expensive part, and it is a commodity. The score is arithmetic on the array output.

That distinction matters if you want to work with the data. GrimAge is one linear model applied to a methylation matrix. If you obtain the matrix, you can compute dozens of other clocks yourself at no marginal cost. You can also recompute them when better models appear. So the practical question is whether the vendor returns your raw IDAT files. Which clock vendor to buy from is not the practical question.

What GrimAge measures

GrimAge, published by Lu and colleagues in 2019, was trained in two stages. First, the authors built DNA methylation predictors of seven plasma proteins.

  • adrenomedullin
  • beta-2 microglobulin
  • cystatin C
  • growth differentiation factor 15 (GDF-15)
  • leptin
  • plasminogen activator inhibitor 1 (PAI-1)
  • tissue inhibitor of metalloproteinases 1 (TIMP-1)

They also built a methylation predictor of smoking pack-years. Second, they regressed time-to-death on those eight surrogates plus chronological age and sex, using a Cox elastic net model.

The output is expressed in years, which is a presentation choice rather than a biological claim. The quantity people report is AgeAccelGrim: the residual after regressing GrimAge on chronological age. A value of +4 means your score is four years higher than the average person your age in the reference cohort. GrimAge2 adds methylation surrogates for C-reactive protein and hemoglobin A1c, and generally performs slightly better in mortality and morbidity association tests.

Two features of this design deserve attention. GrimAge is trained on mortality, not on chronological age, which is why it outperforms first-generation clocks like Horvath’s 2013 multi-tissue predictor at predicting hard outcomes. It also carries a heavy smoking signal. If you have never smoked, a large share of the discriminative power that made GrimAge famous does not apply to you, and the remaining signal is dominated by the inflammatory and metabolic surrogates.

The measurement stack, concretely

The physical assay is the Illumina Infinium MethylationEPIC array. Version 1 interrogates about 850,000 CpG sites. Version 2, now standard, covers roughly 935,000 and dropped a substantial number of v1 probes. That probe churn is a real problem for clock computation, because most published clocks specify exact CpG identifiers. Missing probes must be imputed from a reference mean or the clock must be recalibrated, and both choices shift your score.

Raw output arrives as paired IDAT files, one for each color channel, plus a sample sheet mapping sentrix barcodes to positions. In R, the standard entry point is minfi::read.metharray.exp() followed by normalization. We prefer sesame over minfi for v2 arrays because its masking of unreliable probes and its openSesame() pipeline handle the v2 manifest more cleanly. For clock work specifically, use the normalization the clock was trained under when you can: Horvath’s calculator expects noob background correction followed by BMIQ, while Levine’s PhenoAge and DunedinPACE were developed on differently processed data.

Then you need the coefficients. This is where GrimAge differs from its peers. Horvath’s clock, PhenoAge, Hannum and DunedinPACE all have publicly available or licensable weights. R packages such as dnaMethyAge and methylclock implement them directly. GrimAge weights are not published in full. To get a GrimAge score you upload a beta-value matrix and a sample annotation file to the Clock Foundation’s online calculator and agree to its terms. Practically, that means a CSV with CpG identifiers in rows and samples in columns. A second CSV carries age, sex, and tissue. The calculator returns a table with GrimAge, its component surrogates, and several other clocks.

Reliability, and why single measurements mislead

Individual CpG probes on the EPIC array are noisy. Test-retest correlations for single sites are often poor, and clocks built as unweighted sums of a few hundred such probes inherit that noise. The practical consequence is that the same blood draw split into two aliquots can produce scores differing by a year or more, and repeat draws weeks apart differ by more still.

Two mitigations work. The first is principal-component versions of the clocks, which project methylation onto PCs before applying weights and substantially improve test-retest reliability. The second is repeated sampling. If you intend to track change over time, a single baseline is close to useless. Take two or three samples at baseline, then repeat on a fixed schedule, and treat the noise band as roughly a year for PC-based scores and wider for the original implementations.

There is also a biological confound that people underweight. Blood methylation reflects cell composition, so a shift in your neutrophil-to-lymphocyte ratio from a recent infection moves your score. Estimate cell fractions with Houseman deconvolution (minfi::estimateCellCounts2 or sesame’s estimateLeukocyte) and report them alongside the clock. If your immune cell proportions moved between timepoints, the clock delta is partly compositional.

Is it worth buying

Our view is that a standalone GrimAge test is a poor purchase and a methylation array with returned IDAT files is a reasonable one. The score by itself tells you a population-level mortality association, and you cannot act on it as an individual without clinical context. The array, in contrast, is a durable dataset that you can rescore against every clock published in the next decade.

The interventional evidence is thinner than the marketing suggests. A recent systematic evaluation of how epigenetic aging biomarkers respond to human longevity interventions found that responsiveness varies considerably across clocks and trials, and that several widely marketed clocks move little.1 Some interventions do show effects: a secondary analysis of a randomized trial of cognitive behavioural therapy for insomnia reported changes in epigenetic aging measures.2 Life history also registers in these measures. Adults with documented childhood maltreatment show faster epigenetic aging on several clocks.3 Methylation signatures also appear in specific physiological states, including a distinct profile in women with diminished ovarian reserve.4

Interpretation is where things get delicate. An elevated AgeAccelGrim is a statistical signal, not a finding about your body, and it does not identify a cause or indicate any action. If your score is high and you want to know why, the productive next step is conventional measurement. That means inflammatory markers, metabolic panels, and a conversation with a physician who can order and interpret them. There is also a use question worth taking seriously. Biological-age scores are already being discussed as inputs to institutional decisions, with real potential for algorithmic age discrimination, so be deliberate about who holds your score and under what agreement.5

Questions people also ask

How much does a methylation test cost? A consumer epigenetic age test runs about $250 to $500. A bare EPIC array run through a research core or commercial lab is often cheaper per sample, particularly in batches. That price assumes IDAT files are returned with no scoring service.

How much does the TruAge epigenetic age test cost? TruDiagnostic’s TruAge products have generally been priced between about $250 and $500, with the higher tiers adding more clock outputs and reports. Check current pricing directly, since it changes.

How much is the Horvath test? Horvath’s 2013 clock is a published set of coefficients, not a product, and the Clock Foundation’s online calculator has historically computed it without charge once you supply a methylation matrix. Your cost is the array itself.

What is the GrimAge test and how does it work? It runs a methylation array on your DNA, computes methylation-based surrogates for seven plasma proteins and smoking pack-years, then combines those with age and sex through a mortality-trained model to output a value in years.

Is epigenetic testing worth it? The raw methylation data is worth having because it is reusable and cheap to rescore. A single clock number, purchased alone and interpreted alone, gives you little you can act on.

Footnotes

  1. Raghav Sehgal, Daniel Borrus, Jenel F. Armstrong, et al. Responsiveness of epigenetic aging biomarkers to longevity interventions in humans. Nature Medicine, 2026. https://doi.org/10.1038/s41591-026-04562-9 ↩

  2. Judith E Carroll, Cynthia D Kusters, Hash Brown Taha, et al. Cognitive behavioural therapy for insomnia and epigenetic ageing: secondary analysis from a randomised controlled trial. The Lancet Healthy Longevity, 2026. https://doi.org/10.1016/j.lanhl.2026.100861 ↩

  3. Daniel W. Belsky, Ariel Bellatin, Kellie Courtney, et al. Epigenetic Clock Analysis of Biological Aging in Adults With Documented Childhood Maltreatment. JAMA Network Open, 2026. https://doi.org/10.1001/jamanetworkopen.2026.28033 ↩

  4. Kristina W. Olsen, Juan Castillo-Fernandez, Andrew Cho Chan, et al. Identification of a unique epigenetic profile in women with diminished ovarian reserve. Fertility and Sterility, 2021. https://doi.org/10.1016/j.fertnstert.2020.09.009 ↩

  5. Henry Bair. Biological-Age Scoring and the Risk of Algorithmic Age Discrimination: An Institutional Ethics Framework. HEC Forum, 2026. https://doi.org/10.1007/s10730-026-09606-8 ↩

Pothos builds longitudinal molecular profiles of individuals: whole-genome sequencing, single-cell RNA sequencing, and plasma proteomics, delivered as a detailed report, every data file, and stored cells. Build your profile.