Grain-Free Dog Food and DCM: What the Science Actually Says

First, your dog's biology

To understand why this controversy matters, you need to understand what your dog actually is, metabolically. Not what the marketing says. What the biology says.

Dogs are not wolves. But they are not humans either.

In 2013, a team led by Erik Axelsson at Uppsala University published a landmark paper in Nature that changed how we think about dog domestication. They compared the genomes of dogs and wolves and found that one of the strongest signals of selection during domestication involved genes related to starch digestion.

Specifically, dogs carry an average of about 7 copies of a gene called AMY2B, which codes for pancreatic amylase, the enzyme that breaks down starch. Wolves carry about 2 copies. Each additional copy increases the dog's ability to digest starch by roughly 5.4%. This expansion happened alongside human agricultural development, around 7,000 years ago.

This is the finding that the grain-inclusive camp uses to argue that dogs are well-adapted to eating grains and starches. And they are partially right. Dogs can tolerate starch far better than wolves.

But here is what that same research also shows, and what the marketing glosses over.

None of this means grains are bad for dogs. Whole grains like brown rice, oats, and barley have been part of the domestic dog's diet for thousands of years. They provide digestible energy, fiber, and micronutrients. The AMY2B expansion is real evidence of adaptation. The question is not whether dogs can handle starch. They can. The question is what happens when you replace the grains they have co-evolved with over millennia with a class of ingredients that entered the commercial dog food supply at scale less than 15 years ago.

That is where the legumes come in.

The rise of grain-free (and what actually replaced the grains)

Grain-free dog food went from a niche category to roughly 40% of the US pet food market between 2008 and 2017. The marketing pitch was simple and effective: wolves do not eat wheat, so your dog should not either. It resonated with dog owners who were increasingly skeptical of big pet food companies and attracted to brands that positioned themselves as more "natural" or "ancestral."

But here is what the marketing did not tell you. Removing grains from kibble creates a formulation problem. Extruded kibble needs starch to bind. Without wheat, corn, or rice, manufacturers needed a replacement that could hold the kibble together, provide bulk, and keep costs manageable.

The answer was legumes. Peas, lentils, chickpeas, and their derived forms: pea protein, pea fiber, pea starch, lentil flour. These ingredients offered something grains did too, but with an added bonus for the manufacturer. Pea protein concentrate is roughly 80% protein by weight, which meant brands could push "high-protein" claims on the front of the bag while using a cheap plant source instead of expensive animal protein to hit that number.

By 2018, it was common to find grain-free kibbles with three, four, or five pulse-derived ingredients in the top ten. Peas at position 2, pea protein at position 5, lentils at position 8, pea fiber at position 11. Aggregated, the total pulse content in some formulas exceeded 40% of the diet. This is split-ingredient labeling at work.

These ingredients have existed in human food for centuries. But in commercial extruded dog food, at these concentrations, fed as a sole diet for years? That is a 10 to 15 year experiment. And nobody ran the trial before it went to market.

July 2018: the FDA announces an investigation

On July 12, 2018, the FDA's Center for Veterinary Medicine announced it had begun investigating reports of dilated cardiomyopathy in dogs eating certain pet foods. DCM is a condition where the heart muscle weakens and the chambers enlarge, reducing the heart's ability to pump blood. It leads to congestive heart failure and can cause sudden death.

DCM has a well-documented genetic component in certain breeds. Doberman Pinschers, Great Danes, Boxers, and Irish Wolfhounds are genetically predisposed. What alarmed veterinary cardiologists was that they were seeing DCM in breeds that had never been associated with it: Golden Retrievers, Labrador Retrievers, mixed breeds, Miniature Schnauzers, Shih Tzus. And many of these dogs shared one dietary pattern.

What the taurine theory got right and wrong

The initial hypothesis was straightforward: grain-free diets were causing taurine deficiency, and taurine deficiency causes DCM. This was based on established veterinary science. Taurine is an amino acid critical for cardiac function, and taurine-deficient DCM had been documented in cats (leading to the mandatory taurine supplementation in cat food starting in the 1980s) and in certain dog breeds, particularly Golden Retrievers.

But as more data came in, the picture got complicated.

Most dogs with diet-associated DCM had normal whole-blood taurine levels. A 2019 study by Adin and colleagues found that simple taurine deficiency explained only a minority of cases. The dogs' hearts were failing, their diets were pulse-heavy, but their taurine levels were fine.

This did not disprove the dietary connection. It meant the mechanism was more complex than a simple nutrient deficiency. Something about these diets was damaging the heart, but it was not the straightforward taurine story everyone expected.

Possible explanations that researchers began exploring:

• Reduced bioavailability of sulfur amino acids (the precursors to taurine) from pulse proteins, even when absolute levels appeared adequate
• Bile acid alteration from high-fiber pulse diets, potentially increasing taurine loss through fecal excretion
• Direct effects of pulse compounds on myocardial cell function, unrelated to taurine
• Interactions between multiple nutritional factors that no single-nutrient model could capture

For years, none of these hypotheses had direct evidence. Then, in 2025, something changed.

The Freeman 2025 discovery: phospholipidosis

Dr. Lisa Freeman at Tufts University, a board-certified veterinary nutritionist (DACVIM-Nutrition) who had been studying diet-associated DCM since the beginning of the investigation, published a study that shifted the conversation.

Her team measured urinary concentrations of a lipid metabolite called di-22:6-bis(monoacylglycerol)phosphate, or di-22:6-BMP, in four groups of dogs: dogs with DCM eating high-pulse diets, dogs with DCM eating low-pulse diets, healthy dogs eating high-pulse diets, and healthy controls eating low-pulse diets.

The results: dogs with DCM on high-pulse diets had significantly elevated di-22:6-BMP levels compared to controls. The metabolite concentrations were positively correlated with the pulse content of the diet (r = 0.52).

di-22:6-BMP is an established biomarker for phospholipidosis, a condition where phospholipids accumulate abnormally inside cells, particularly in lysosomes. In humans and other animals, phospholipidosis is a known mechanism of drug-induced organ damage, including cardiac damage. Finding this biomarker elevated in dogs with diet-associated DCM, and correlated with pulse intake, provides the first biologically plausible mechanism connecting pulse-heavy diets to heart disease.

The study won the AKC Canine Health Foundation's inaugural Canine Health Discovery of the Year Award in December 2025.

It is a single study with 53 dogs. It needs replication. But it is the first time anyone has identified a credible biological pathway from pulse-heavy diet to cardiac damage that does not depend on taurine deficiency.

The industry response: follow the money

While independent researchers at Tufts, Cornell, and UC Davis were publishing peer-reviewed work on diet-associated DCM, the pet food industry was funding its own counter-narrative.

In 2020, a group of authors employed by BSM Partners, a pet care research and consulting firm, published a literature review in the Journal of Animal Science concluding there was no established link between grain-free diets and DCM. The paper stated:

"The authors declare no real or perceived conflicts of interest."

Veterinary critics immediately raised concerns.

The authors eventually updated their disclosure statement, but by then the paper had been widely cited by grain-free brands as evidence their products were safe.

This is the paper that still gets cited when someone on a pet food forum says "the DCM thing was debunked." It was not debunked.

An industry-funded literature review with initially undisclosed conflicts of interest concluded the evidence was insufficient. That is a different statement.

BSM Partners has since continued publishing research that questions the diet-DCM link, including studies challenging blood-based taurine diagnostics and criticizing research published by competitors (most recently, attacking a metabolomics study on The Farmer's Dog products). Their position is consistent: grain-free diets have not been proven to cause DCM. This is technically true. Causation has not been proven. But the pattern of evidence, the biological mechanism now identified, and the systematic failures of disclosure deserve to be laid out transparently so you can evaluate the claims yourself.

December 2022: the FDA walks away

On December 23, 2022, the Friday before Christmas, the FDA posted a brief update to a three-year-old webpage. Not a press release. Not an announcement. A paragraph.

The agency stated it had "insufficient data to establish a causal relationship" between DCM reports and pet food products. It would not release further updates "until there is meaningful new scientific information to share."

1,382 reported cases. 16 named brands. Four and a half years of investigation. And the conclusion was: we do not have enough data, and we are done looking.

The pet food industry treated this as exoneration. The grain-free market stabilized. Reformulations slowed. Multiple reports suggest the FDA's update came only because of mounting Freedom of Information Act requests, not because the agency had planned to communicate.

But veterinary cardiologists did not stop seeing cases. Dr. Freeman's phospholipidosis research was published two years later. Academic work at UC Davis, Cornell, and the University of Saskatchewan continued. The FDA may have stopped updating its webpage, but the science did not stop.

The investigation started with a roar and ended with a whimper. But "insufficient to prove causation" is not the same as "safe." It means nobody funded the definitive study. It means the reporting system was flawed from the start (voluntary adverse event reports with no denominator data). It means the question is still open.

The distinction that matters: grain-free vs high-legume

This is where Sniff diverges from every other rating system, and it is the single most important conceptual move in our methodology.

Every other scoring system treats these as the same thing. Either grain-free gets blanket-penalized (which unfairly punishes well-formulated grain-free foods) or it gets ignored entirely (which pretends the FDA investigation and the Freeman discovery never happened).

Here is how Sniff handles it.

What we do not know (and what we will do when we find out)

We are not pretending this is settled science. Here is what remains genuinely uncertain.

If future research demonstrates that pulses at any concentration are safe for all dogs, we will remove the penalties, update the rubric, publish a changelog explaining exactly what changed and why, and re-score every affected product. The methodology is versioned for exactly this reason. We would rather update a position based on new evidence than defend an old one out of stubbornness.

And if future research strengthens the association, we will tighten the scoring accordingly. The rubric follows the evidence. That is the commitment.

Where Sniff stands

We apply the Lindy principle. Named animal proteins have anchored canid nutrition for roughly 15,000 years. Whole grains have been part of the domestic dog's diet for at least 7,000, backed by documented genetic adaptation. High-pulse extruded kibble has existed at scale for about 10 to 15 years. The burden of proof rests on the newer formulation.

We do not penalize grain-free. We penalize high-legume.

These are different things, and treating them as the same thing is either lazy analysis or deliberate misdirection.

We score mitigation. A brand that includes taurine, organ meats, or marine protein alongside pulse ingredients is making a visible effort to address the concern. That effort is reflected in the score.

We publish everything. Every weight, every threshold, every cap trigger, every citation. If you think our pulse penalty is too aggressive, you can read the rubric, identify the specific line you disagree with, and tell us why. If you make a compelling case, we will update it. That is not a slogan. That is how the versioning system works.

We are precautionary, not alarmist. A WATCH flag on a pulse ingredient is not "this food will kill your dog." It is "the evidence is not settled, the investigation was not completed, a plausible biological mechanism has been identified, and we think you should know." What you do with that information is your decision.

Frequently asked questions

Related reading