The Golden Retriever
Lifetime Study.
3,044 dogs. 14 years.
The longest-running veterinary study of its kind.
What these dogs built
3,044 Golden Retrievers built the most detailed map of canine health ever created.
They did not know that, of course. They knew the car ride to the vet. They knew the treat after the blood draw. They knew their person brought them here every year, and that was enough.
But what they built, one appointment at a time, over 14 years, is something that did not exist before them. Annual veterinary exams. Blood draws. Fur and nail samples. Questionnaires that tracked everything: what they ate, where they slept, what chemicals were on the lawn, whether they swam in lakes or drank from hoses, whether they were anxious during thunderstorms. Every measurable detail of a dog's life, recorded year after year, so that when the patterns finally emerged they would come from data instead of guesswork.
Most health studies take a snapshot. One blood draw. One survey. One moment in time. This study watched. For years. Through puppyhood and middle age and the slow afternoons of old age. Through moves and new babies and job changes and the thousand small shifts that make up a life. The families kept showing up. The dogs kept wagging through it. Annual participation stayed between 74 and 87 percent through the first five years. Nine years in, overall retention was still 86 percent. That is not a retention rate. That is devotion.
What they built together is a dataset unlike anything else in veterinary medicine. And it is just getting started.
Why Golden Retrievers
They were chosen for a reason nobody wanted to be true.
The cancer burden in the breed is staggering. A 1998 Golden Retriever Club of America survey of 1,444 dogs found that 61 percent of the 420 reported deaths in that cohort were cancer-attributed. The GRLS cohort itself has reported 70 percent of its deaths to cancer through 2021, and hemangiosarcoma alone accounts for roughly 70 percent of those cancer deaths. The breed that defined the American family dog for half a century has one of the highest cancer mortality rates on earth.
The current median Golden lifespan sits around 10 to 12 years. Veterinary geneticists broadly agree the breed is dying younger than it once did. Whether the cause is narrowing genetics, environmental exposure, diet, breeding practices, or some combination, nobody has untangled yet. The result was a breed in crisis.
In 2012, the Morris Animal Foundation decided to stop guessing and start watching. They enrolled 3,044 Golden Retrievers from across the United States between June 2012 and April 2015 and committed to following every one of them for life. Not to test a hypothesis. To build the foundation for every hypothesis that would come after.
The study asks one enormous question: what happens to Golden Retrievers across a full life, and what can that teach us about why some of them get sick and others do not?
Fourteen years later, the answers are arriving.
What the study is finding
Hemangiosarcoma dominates by far.
By Hillman et al. 2023, 233 of 3,044 enrolled dogs (7.65 percent) had been diagnosed with hemangiosarcoma as of September 2022, with an incidence rate of 1.10 cases per 100 dog-years. A Research Square preprint with a December 2024 cutoff reports 490 of 3,044 dogs (16.1 percent) diagnosed, the incidence rate climbing to 1.69 per 100 dog-years. Visceral hemangiosarcoma is the most common form. Lymphoma is second among GRLS cancers. Mast cell tumors and osteosarcoma, despite their reputation in the breed, have appeared far less often than the field expected (10 and 8 cases respectively at the 2021 cutoff). The four cancers appear at different ages, in different patterns, with different risk profiles. A dog that avoids lymphoma at six is not safe from hemangiosarcoma at ten. The study is mapping when each cancer peaks and what distinguishes dogs who develop one from dogs who develop another. That kind of resolution is only possible when you have watched thousands of dogs from puppyhood through their entire lives.
Spay and neuter timing matters more than the field assumed.
Goldens spayed or neutered at six months or younger showed elevated orthopedic injury rates in the GRLS data. Gonadectomy at any age was associated with increased overweight and obesity (Simpson et al. 2019). The cancer link, well-documented in earlier work by Hart and colleagues on Goldens, Vizslas, and Labradors, was a separate finding; the GRLS contribution to the broader reassessment of routine early spay/neuter recommendations comes from the orthopedic and obesity results, not a cancer signal in the GRLS itself.
Environment is harder to read than anyone hoped.
Luethcke et al. 2022 tested chemical plants, dumps, incinerators, railroads, landfills, coal plants, transmission lines, ozone, particulate matter, and secondhand smoke against lymphoma risk in the GRLS cohort. None of those exposures correlated. Tindle et al. 2024 examined radon levels and proximity to active fracking wells; both came back null, with one exception: Goldens with lymphoma lived significantly closer to wastewater wells than matched unaffected dogs (P = .01). A hemangiosarcoma and pollutants case-control study with 100 cases and 200 controls is still ongoing. The honest summary is that the published environmental analyses have come back mostly null. The GRLS is the only study large enough to keep asking.
Diet and behavior are being measured alongside genetics for the first time at this scale.
What a dog eats at age three may influence what happens at age nine. Whether a dog is active or sedentary, anxious or calm, urban or rural. These variables have always been suspected to matter. The GRLS is the first study large enough and long enough to start showing how they interact.
The gut microbiome is an emerging axis.
A partnership with AnimalBiome, announced in 2022, is characterizing gut bacterial and fungal microbiomes in dogs with and without cancer. A primary peer-reviewed GRLS microbiome publication has not yet appeared. The work is early. But it is the kind of question that only a 14-year study can even ask.
And the research keeps expanding.
A North Carolina State study is examining PFAS exposure and hypothyroidism in the cohort. The hemangiosarcoma case-control study is testing chronic exposure to persistent organic pollutants against HSA risk. Each of these projects draws on the same data the families have been contributing for over a decade. The investment keeps paying forward.
The Oldies
196 of the 3,197 GRLS dogs in our atlas reached age 12 or older without a malignant cancer diagnosis.
Let that settle for a moment. In a cohort where 70 percent of deaths through 2021 were cancer-attributed and median lifespan sits around 10 to 12 years, reaching 12 without cancer is not just old. It is extraordinary.
We call them Distinguished Oldies and they carry a special badge in the atlas. Not because we decided they were important. Because they are, scientifically, among the most valuable Golden Retrievers ever studied.
Understanding what protected them is the path to protecting the ones that come next. Was it genetics? Environment? Diet? Activity level? Some combination that only becomes visible when you compare them against the dogs who were not as fortunate? Those questions are answerable now, because the data exists. Fourteen years of annual records on 196 dogs who beat the odds.
What our 3,197 contribute
The Sniff atlas includes 3,197 Goldens identified as part of the GRLS cohort in the CanVAS canine variant atlas. That is 153 more than the Morris Animal Foundation's original enrollment of 3,044; the additional dogs are GRLS-cohort genotypes that CanVAS identified at the genotype level when harmonizing fifteen public canine datasets together. Either way, this is the densest within-breed population in our atlas, by a wide margin. Each dog has a real genetic position computed from the 77,215-marker typed backbone and 256 principal components.
At this density, things become visible that you cannot see with smaller numbers. Field-line Goldens cluster separately from show-line Goldens. North American lines separate from European imports. The genetic diversity within a breed that most people think of as one uniform population turns out to be structured in ways that may matter for health outcomes.
Every one of these dogs carries trait loci calls at 15 genes governing coat, color, size, and body shape. Every one has carrier status results for variants like MDR1 and progressive retinal atrophy. Every one has a computed neighborhood of the most genetically similar dogs in the atlas.
When the Morris Animal Foundation phenotype data joins these genetic positions, the annual health records, the cancer diagnoses, the environmental surveys, the diet logs, this becomes one of the most deeply characterized cohorts in the history of veterinary genetics.
These dogs already taught us something just by being placed. The within-breed substructure they reveal, the trait distributions across 3,197 genomes, the carrier frequencies computed from a real population rather than estimated from a handful of samples. That is science nobody else has published at this resolution for Golden Retrievers.
The names
The dogs in this study came to us as subject IDs. Numbers in a dataset. That is how research works. It is not how dogs work.
We gave each one a name from a pool of nature, stars, and warm human names. These are placeholders. We know that. A generated name is not the name a dog knew, the one spoken at the back door every evening, the one that made their ears perk up from across the house.
We thought they deserved more than a number to be remembered by. But a placeholder is not a name. It is a promise that the real one has a place here whenever it comes home.
If your dog was part of this study, they have a star in the atlas, a genetic profile, a neighborhood, and a page waiting for the name they were loved by. Your photo replaces the AI portrait. Your memorial line lives on their page permanently. The placeholder steps aside the moment you show up.
We are not yet affiliated with the Morris Animal Foundation, but we have reached out and look forward to coordinating verification with them. In the meantime, email [email protected] with your dog's enrollment number, their name, and anything else you want on their page. We will get back to you.
The Golden Retriever Lifetime Study is conducted by the Morris Animal Foundation. It is the largest and longest observational veterinary study in history. It was made possible by 3,044 families who said yes and kept saying yes, year after year, because they believed their dog's life could teach us something.
Sniff is not affiliated with the Morris Animal Foundation. We are grateful for their work and for every family and every dog that made this study real. The science belongs to everyone. We are doing our best to make sure it is seen.
Morris Animal Foundation, Golden Retriever Lifetime Study ↗
Cohort reference: Labadie et al. 2022, doi:10.1371/journal.pone.0269425. Sources cited in the body: Hillman et al. 2023 (Vet Comp Oncol, hemangiosarcoma cases); Luethcke et al. 2022 (Canine Med Genet, lymphoma and environmental exposures); Tindle et al. 2024 (J Vet Intern Med, radon and fracking proximity); Simpson et al. 2019 (PLOS ONE, gonadectomy and orthopedic injury / obesity); Research Square preprint rs-7064010 (December 2024 HSA incidence update, not yet peer-reviewed); Morris Animal Foundation 2026 program updates (retention, PFAS, HSA case-control). The 1998 cancer mortality figure originates in the Golden Retriever Club of America National Health Survey.