Science and society · Document reading
A dog-aging announcement, checked against its paper
The dog-aging claim traveling through public discussion traces back to a research highlight the National Institute on Aging published on September 3, 2020.
By Dale Whitmore, Records editor · 9 min read · Published · Updated
What did the NIA dog-aging announcement claim?
The dog-aging claim traveling through public discussion traces back to a research highlight the National Institute on Aging published on September 3, 2020. Checked against its paper, the NIA highlight holds up on population and method: 104 Labrador retrievers, DNA methylation, a new dog-to-human age formula. It leaves out three things the paper’s record shows: a clause on anti-aging interventions, six of the seven grants, and a declaration that two authors hold a patent on animal methylome profiling.
The NIA highlight states that researchers used epigenetics, specifically DNA methylation, to compare aging in dogs and humans, and that the work updated “the previous ‘one dog year equal to seven human years’ formula.” That is the claim as published: an update to a popular formula, not a new law of biology.
The highlight also preserves a specific example pairing: an 8-week-old puppy compared to roughly a 9-month-old baby, and a 12-year-old senior Labrador compared to roughly a 70-year-old adult. These are the only age-equivalence figures the announcement gives, and they apply strictly to the study’s own methylation-based calculation for the dogs it sampled.
The announcement attributes the underlying science to a named paper, cited on the NIA page as Wang T, et al., “Quantitative translation of dog-to-human aging by conserved remodeling of the DNA methylome,” published in Cell Systems in 2020, volume 11, issue 2, pages 176 to 185.e6, doi 10.1016/j.cels.2020.06.006. That citation is the anchor for everything else in the announcement. The same paper is indexed on PubMed at PubMed’s record for the study.
Two organizational details belong in any record of this claim before comparison to the paper begins.
- The NIA highlight is a research highlight, a summary format standing separate from the peer-reviewed paper itself.
- The formula update is described as tied to a specific data set and a specific method, DNA methylation clocks, a claim scoped to that measurement alone.
Match the announcement to the primary paper
The primary paper backing the NIA announcement is the Wang et al. study indexed at PubMed under record 32619550, published in Cell Systems in 2020 with the title “Quantitative translation of dog-to-human aging by conserved remodeling of the DNA methylome.” Matching the announcement to this record means checking whether the summary and the indexed study describe the same population, method and outcome.
On population, the NIA highlight states that researchers collected DNA methylation data from blood samples of 104 Labrador retrievers, ranging from puppies to 16 years of age, and compared these against methylation patterns from 320 humans, ages one to 103. That detail, specific to Labrador retrievers, is a fact carried from the announcement, and it should be checked against the paper’s own methods before assuming it generalizes.
On method, both the announcement and the underlying citation describe DNA methylation, a chemical modification affecting gene activity, as the measured signal. The NIA highlight calls this an “epigenetic clock” approach: methylation patterns that change with age in a consistent way, used to build a cross-species comparison.
On outcome, the announcement’s claim is that this method produced a “new formula to calculate dog-to-human age,” replacing the earlier seven-to-one multiplier with an approach based on developmental gene methylation patterns shared between species.
The paper’s own abstract is tighter than either the title or the highlight. It describes characterising the methylomes of 104 Labrador retrievers across a 16-year age range, using oligo-capture sequencing at more than 150 times coverage within mammalian syntenic blocks. The abstract reports that comparison with human methylomes reveals a nonlinear relationship, one that translates dog-to-human years and lines up the timing of major physiological milestones.
One phrase in that abstract does not appear in the NIA highlight at all. The paper says the conserved changes are sufficient to translate age and the effects of anti-aging interventions across multiple mammals. The announcement drops the interventions clause. That is a difference worth logging, because it runs the opposite way from the usual pattern where a press summary overstates a paper.
Compare population, outcome and study status
The population behind this claim is narrower than the phrase “dog aging” suggests. The NIA highlight specifies 104 Labrador retrievers, spanning puppies to 16 years of age, compared against 320 humans ranging from one to 103 years old. A single breed sample is the base for the age-translation formula described in the announcement.
The design is worth naming before the outcome. Blood samples were profiled and methylation patterns compared across two species at different ages, with a function fitted to the alignment. The age-translation comparison described here assigns no aging treatment. The formula describes a correspondence between two sets of measurements, a statistical alignment, and the announcement stops short of naming a cause of ageing in either species.
The outcome measured is methylation pattern alignment, standing apart from any health or longevity outcome. The announcement describes researchers finding “similar DNA methylation patterns between the dogs and humans, especially in the early and late life stages,” with changes in shared developmental genes used to line up the two species’ epigenetic timelines. That is a molecular-pattern comparison, and the announcement does not present it as a claim about disease risk, mortality risk or physical aging symptoms.
The two example age pairs, an 8-week-old puppy set against roughly a 9-month-old baby, and a 12-year-old Labrador set against roughly a 70-year-old adult, are outputs of this methylation-based calculation specific to the sampled data. The NIA highlight does not extend these pairs into a general-purpose age chart for all dogs.
Study status matters for how the claim should be read today. The NIA highlight, dated September 3, 2020, describes the underlying paper as “recently published” at that time, and states plainly that “next steps include testing the formula with other dog breeds, and further exploring epigenetic changes in developmental genes as they relate to aging.” That sentence marks the formula as a starting point pending cross-breed replication, according to the announcement’s own framing, and stops short of a settled cross-breed standard.
| How the announcement frames it | What the announcement’s own text specifies | |
|---|---|---|
| Population | Dogs (general framing) | 104 Labrador retrievers, puppies to 16 years |
| Comparison group | Not specified in headline | 320 humans, ages 1 to 103 |
| Outcome measured | “Dog aging” | DNA methylation pattern alignment |
| Status | Framed as an update | Described as needing cross-breed testing |
Record what the announcement leaves out
The announcement does not extend its formula beyond Labrador retrievers. The NIA highlight names next steps explicitly: “testing the formula with other dog breeds, and further exploring epigenetic changes in developmental genes as they relate to aging.” That sentence is the institute’s own acknowledgment that the single-breed data set has not been checked against other breeds at the time of publication.
The announcement does not claim a health, disease or longevity finding. It describes a molecular comparison method, DNA methylation pattern matching between species, used to build an age-translation formula. Nothing in the NIA text ties this formula to claims about a dog’s risk of illness, expected lifespan change or clinical outcome. Any read of this announcement as evidence for a health or wellness effect goes beyond what the summary states.
The announcement gives two example age pairs and no others. The 8-week-old puppy to 9-month-old infant pairing, and the 12-year-old Labrador to 70-year-old adult pairing, are the only translated ages published in the highlight. A complete age chart, a formula usable for arbitrary ages, or breed-specific adjustments are not part of the published summary.
The announcement does not describe study funding beyond a partial NIA grant citation, “P01AG031862,” and it does not detail sample collection dates, methodology limitations, or statistical methods used to align the methylation clocks. The PubMed record lists seven grants behind the work, where the announcement names only one.
The larger omission is an interest declaration. The paper’s record carries a Declaration of Interests stating that T. Ideker is a co-founder of Data4Cure with an equity interest and a seat on its scientific advisory board, that he sits on the scientific advisory board of Ideaya Biosciences with an equity interest and sponsored research funding, and that Ideker and T. Wang hold a patent entitled “Methylome profiling of animals and uses thereof,” international application PCT/US18/49103 and US application 16/638,454. The record adds that UC San Diego reviewed and approved those arrangements under its conflict of interest policies.
None of that makes the methylation work wrong. It does mean two authors hold a patent on animal methylome profiling, and a public-record file on this announcement that omits the declaration is an incomplete file.
The announcement’s list of contributing institutions, University of California San Diego, the National Human Genome Research Institute, Sanford Burnham Prebys Medical Discovery Institute, University of Pittsburgh School of Medicine, and UC Davis, is stated without describing each institution’s specific role in the study design.
Readers checking this claim against later coverage should treat any generalized “dog years” chart, any breed-specific claim beyond Labrador retrievers, or any health-outcome claim attributed to this announcement as an addition made outside the original NIA text.
Preserve any response or correction with the claim
No correction, update or amended version of this NIA highlight appears in the material available for this record. The page, published September 3, 2020, reads as a stable research highlight, with no visible edit history to suggest an evolving news item.
The NIA highlight itself functions partly as a correction to prior public understanding. It explicitly frames the study as updating “the previous ‘one dog year equal to seven human years’ formula,” positioning the Labrador-based methylation formula as a replacement for that older, widely circulated multiplier. That framing is the announcement correcting a public assumption, not the announcement being corrected.
The PubMed record for the paper carries no retraction, erratum or correction notice. It does carry something the announcement never mentions: a linked commentary. Under Comment in, the record lists “From dog days to human years” by D. Clyde, Nature Reviews Genetics, September 2020, 21(9), pages 508 to 509, PMID 32681033. That’s a published response in the literature, and a record of this claim that stops at the NIA highlight would miss it. A correction issued after today would not appear in this article either, so the PubMed record stays the place to check.
The NIA highlight names its own limitation as an open item rather than a closed finding: cross-breed testing is described as a “next step,” not as work already completed at the time of publication.
The PubMed record also links the Nature Reviews Genetics commentary noted above, and the PubMed record’s own Cited by list runs to work in other species, including epigenetic ageing in marmots and in retired sled dogs. Absence of a dispute in these records does not prove there has been none, and that limit belongs in the file next to everything else.
Anyone tracking this claim forward should check the PubMed record periodically for citation notices, since journal-level corrections are typically appended there rather than to an institutional news summary. Is one dog year equal to seven human years? The NIA highlight says the 2020 study updated that formula using DNA methylation. How old is a 12-year-old Labrador in human years? About 70, by the study’s methylation-based calculation, which covers Labradors only.