A blood sample can now date a cat to a median error of 0.83 years, roughly ten months, using 34 methylation sites rather than teeth. That figure comes from a clock built on 130 cats aged 0.21 to 20.9 years, and every feline age study since is judged against it. Here is what the number is worth, and where it stops being worth anything.
What an epigenetic clock actually counts
DNA methylation is a methyl group parked on a cytosine that sits next to a guanine, a CpG site. In the 2021 cat study, 1,379 CpG sites shifted their methylation level with age at p below 10-8. The clock does not read all of them. A penalised regression model selected 34 sites, weighted each one, and added them into a single estimate in years.
On blood, the correlation with real age was R = 0.97 and the median absolute error 0.83 years. Two limits follow immediately. The clock estimates chronological age, not health. And the gap between predicted and true age, called epigenetic age acceleration, predicts disease in people; in cats it is a research question, not a treatment trigger.
Three published clocks, three error bars
Not every "cat age" number in the literature is the same measurement. The table below keeps them apart, because mixing them is how a 0.8-year test gets quoted as a 4-year test.
| Clock | CpG sites | Sample | Reported error |
|---|---|---|---|
| Cat blood clock (Raj 2021) | 34 | Blood, 130 cats | MAE 0.83 years, R 0.97 |
| Human-cat chronological clock | 563 | Blood | Trained across both species |
| Human-cat relative-age clock | 540 | Blood | Age as a fraction of lifespan |
| Cat clock applied to cheetahs | 34 | Cheetah blood | MAE 1.38 years, r 0.79 |
| Pan-mammalian clock on cheetahs | Array of 37,492 probes | Blood, liver, skin | MAE 3 to 4.42 years |
The cheetah rows matter for pet owners: the same 34-site clock that hits 0.83 years on a domestic cat drifts to 1.38 years on a related species, and the general mammalian clock is off by three to four years. Precision is tied to the population the model was trained on.
What the at-home swab actually delivers
Commercial age kits ask for an oral swab, two of them, mailed to a lab, with results in four to six weeks. The US list price is about $109, and the first version sold for $96 in late 2022. The maker states accuracy within 12 months of true age, with 75 percent of samples landing within 5 months, and explicitly does not recommend the test for cats under one year.
Two caveats sit under that claim. The published 0.83-year figure is for blood, not cheek cells, so the swab is a different sample on the same principle. And the accuracy figures are the laboratory's own; no independent trial of the cat swab has been published. What you buy is a birthday estimate plus a life-stage label, not a breed report and not a disease screen.
The seven-year rule versus the real curve
The familiar conversion, one cat year equals 15 human years, two equals 24, then four added per year, is a rule of thumb the clock paper itself calls crude. It is still useful because it is fast and it tracks the shape of the curve: growth is front-loaded, ageing is roughly linear after year two.
| Cat age | Human equivalent | What is happening |
|---|---|---|
| 1 year | 15 | Growth plates closing, adult teeth in place |
| 2 years | 24 | Social maturity, weight stabilises |
| 7 years | 44 | Mature adult stage begins |
| 10 years | 56 | Senior stage, twice-yearly checks |
| 15 years | 76 | Most cats now have measurable kidney change |
Why an exact age changes the medical schedule
The 2021 AAHA/AAFP life stage guidelines split a cat's life into four stages: kitten to 1 year, young adult 1 to 6, mature adult 7 to 10, senior over 10, with annual exams for all and every six months from the senior stage. A cat rehomed as "about eight" that turns out to be 13 just moved two stages and gained a screening schedule.
The numbers behind that shift are large. Chronic kidney disease affects up to 40 percent of cats over 10 and up to 80 percent over 15, so a five-year error changes whether you are screening or monitoring. Knowing the age also settles insurance eligibility, anaesthetic risk conversations, and how a cat lifespan expectation should be framed. If you only want a rough figure, the cat age calculator does the conversion for free.
Where the number fails
Three failure modes are documented. Below one year, methylation moves too fast and non-linearly for an adult model; in the cheetah test set, two animals near sexual maturity were overestimated by 3.1 and 5.6 years. Second, tissue matters: the same clock applied to liver and skin instead of blood lost accuracy badly. Third, a single reading is a snapshot; methylation drifts with cell composition, and a sick cat's blood profile is not a healthy cat's.
That last point is why the test is best read as a range with a midpoint, not a birthday. A result of "11 years, plus or minus one" is honest; "born 14 March 2015" is marketing.
If your cat has no birthday
A veterinary exam still does most of the work. Tooth eruption is reliable up to about 6 months, wear and tartar give a rough band after that, and lens sclerosis, iris atrophy, and muscle mass over the spine add supporting detail. In an adult cat the combined estimate is usually good to two or three years, which is enough for feeding and vaccination and not enough for a kidney screening decision.
A swab is worth the money when the gap actually changes something: a shelter cat with an unknown history, an insurance application, or a cat crossing the senior line where twice-yearly bloodwork starts. For a cat with a known birthday, the test adds a data point and little else. Either way it is a screening tool, and the plan that follows still belongs to a veterinarian.
This article does not replace a veterinary examination. Age estimates inform screening schedules; they do not diagnose anything.
Frequently asked
How accurate is a cat DNA age test?
The peer-reviewed cat clock reports a median absolute error of 0.83 years on blood from 130 cats, with a correlation of 0.97. Commercial oral-swab kits claim accuracy within 12 months and do not recommend testing cats under one year, where growth makes the estimate unreliable.
Does an epigenetic clock measure biological age?
It estimates chronological age. The gap between predicted and calendar age is called epigenetic age acceleration and is linked to disease in humans, but no cat study has shown it should change a diagnosis or a treatment plan yet.
Can I estimate my cat's age without a test?
Yes, with limits. Tooth eruption is reliable to about 6 months, and after that wear, tartar, lens clouding, iris atrophy and spine muscle mass combine into a band roughly two to three years wide in adults. Weighing the cat monthly and scoring body condition adds a trend line a single estimate cannot give.