Grain in dog food — what the research actually says

In the waiting room of a small-animal practice sits the owner of an eight-year-old German Shepherd. For weeks the dog has seemed tired, panting unusually long after a short walk. The tentative diagnosis after echocardiography: dilated cardiomyopathy — a progressive enlargement of the heart chambers, otherwise seen mostly in genetically predisposed breeds like Great Danes or Dobermans. The dietary history: for three years a high-priced “grain-free” premium food, with declared main protein from game and “natural” pulses as the carbohydrate source. Cases like this have triggered a wave of investigations in the United States since 2018; their findings have barely reached European consumers. The gap between research and advertising is considerable.

“BEG” — not just grain-free

The researcher who has named the phenomenon most clearly is Lisa M. Freeman. She is a Diplomate of the American College of Veterinary Internal Medicine in the nutrition specialty, teaches at the Cummings School of Veterinary Medicine at Tufts University in Massachusetts, and led the first clinical case series there on diet-associated DCM. In November 2018 she coined a term on the popular Tufts blog Petfoodology that has shaped the debate ever since: BEG diets — Boutique, Exotic ingredients, Grain-free.¹

The point was not semantic. Freeman wanted to show that the public short-hand “no grain” misses the mark. Strikingly often in her case files, suspect foods showed three features at once: they came from small boutique brands, contained unusual protein sources (kangaroo, bison, duck), and were grain-free — with the carbohydrate gap closed by high proportions of pulses (peas, lentils, chickpeas, beans). In a follow-up piece in February 2023 she drew a sober interim balance: the cause remained unknown, she wrote, but the problem had not gone away.²

What the FDA actually found

The public starting shot came in July 2018 from a letter by the United States Food and Drug Administration. It reported a cluster of dilated cardiomyopathy in dogs of non-predisposed breeds — Golden Retrievers, Labradors, mixed breeds — and named a common thread: diets with high proportions of pulses or potatoes, almost without exception marketed as grain-free. From the start of records in 2014 to November 2022 the agency collected 1,382 case reports. On December 23, 2022 the FDA announced it would no longer publish regular updates because adverse-event numbers alone do not establish causality — the investigation itself officially continues.³

A 2024 Narrative Review in the journal Veterinary Sciences summed up the state of research: an environmentally driven taurine deficiency, an impairment of methionine metabolism by high pulse content, and altered gastrointestinal absorption are all under discussion. None of these hypotheses is conclusively proven; several are mutually compatible.⁴ In Journal of Animal Science an 18-month prospective study in healthy adult dogs appeared in 2025, comparing different carbohydrate sources — including grain. All animals remained clinically normal with respect to DCM markers.⁵

Genetics: what the species can do, what a single breed cannot

The evolutionary background of the debate was laid by a 2013 study in Nature by Erik Axelsson and colleagues at Uppsala University in Sweden. Their genome-wide comparison of 60 dogs from 14 breeds with 12 wolves identified 36 regions under positive selection since domestication — three of them affecting starch metabolism. Dogs typically carry four to thirty copies of the AMY2B gene encoding pancreatic α-amylase, versus two copies in wolves — averaged across breeds, a roughly fivefold increase — and accordingly produce more of the enzyme that breaks down starch.⁶ A follow-up study in Heredity dated the copy-number expansion to the spread of agriculture 5,000 to 7,000 years ago.⁷

So much for the species. But breed-specific genetics complicate the picture. In the Doberman two mutations are now identified that predispose to DCM independently of diet: a 16-base-pair deletion in the PDK4 gene (DCM1) and a missense variant in the titin gene TTN (DCM2).⁸ In the Boxer a separate, also genetically anchored heart disease exists — arrhythmogenic right-ventricular cardiomyopathy (ARVC), caused by an 8-base-pair deletion in the 3’ untranslated region of the striatin gene with a penetrance of approximately 72 percent.⁸ A 2025 PLOS One study added a further wrinkle: in British Dobermans the RNF207 variant was associated with DCM, while PDK4 and TTN were not — the genetic architecture varies within the breed across sub-populations.⁹

In practical terms: anyone keeping a dog of a breed with known DCM predisposition is doubly safe with a conservative diet. Anyone with a genetically unencumbered breed nonetheless has no license for any recipe at all.

The opposite side — and what is right about it

Criticism of the BEG hypothesis comes largely from the boutique segment of the industry and from a suspicion that cannot be brushed aside: the large manufacturers — Mars Petcare, Nestlé Purina, Hill’s — fund a substantial portion of academic veterinary-nutrition research. Who works for whom is a legitimate question.

The point deserves a differentiated reply. First: the central DCM publications — the Tufts work, the FDA data analyses, the 2024 Veterinary Sciences review — are not financed by a single third-party sponsor, and their conflict-of-interest statements are publicly available. Second: independent studies also show anomalies. A 2024 retrospective study in the Journal of Veterinary Internal Medicine by Coppinger et al. compared 97 Irish Wolfhounds on high-pulse (n = 35, 36 %) versus low-pulse (n = 62, 64 %) diets. On echocardiography no significant differences showed up — but on the ECG, 6/35 (17 %) of the high-pulse group exhibited ventricular premature complexes versus 1/62 (2 %) in the comparison group. That is not proof, but it is a finding that does not fit the “nothing to see here” narrative.¹⁰

The honest balance: the main criticism from the boutique side — “no clearly proven mechanism” — is correct. The conclusion “therefore no reason for caution” does not follow.

What the market is saying

The economic environment is congruent. Future Market Insights values the global grain-free pet-food market (dogs and cats) at around USD 49.1 billion in 2024 with a forecast of USD 80.5 billion by 2034 (CAGR 6.4 percent); the dog segment accounts for roughly 71 percent of this volume.¹¹ In the premium segment the category had double-digit annual growth between 2010 and 2018. Since the FDA warning, growth has slowed perceptibly: Nielsen IQ data show year-on-year growth of 4.8 percent for grain-free dog food as of April 2023 — compared with 9.6 percent in 2018; non-grain-free formulations rose 18.7 percent over the same period.¹² Industry analysts cite three drivers of the cooling: consumer scepticism after media reports on the DCM association, the price gap to grain-inclusive premium lines, and competition from gently cooked and fresh-food vendors who occupy the “natural” selling point differently.

In plain language: the “no-grain” wave is reaching its peak after a decade. The market is correcting slowly — the advertising messages take longer.

What matters in processing

A review in Animal Feed Science and Technology documents that starch digestibility depends heavily on both source and processing. Cereal starches (rice, oats, corn) remain relatively well digestible even in raw form; in tubers and pulses, however, raw digestibility sits clearly below the values seen after cooking or extrusion, because their starch granules need heat and water to be opened up properly.¹³ In practical terms: rice or oats in a properly extruded dry food is physiologically a different matter from unprocessed pulse fractions in a poorly prepared home mix. Problems typically arise where poor processing meets high proportions of poorly disrupted raw starch — clinically visible as chronic diarrhoea, flatulence, or in predisposed animals also as inflammation markers.

What remains

The advertising claim “grain-free = species-appropriate” is unsustainable on the current state of peer-reviewed literature. It was already truncated in evolutionary-biological terms; it has taken a further hit through the open question of diet-associated DCM; and it is losing ground economically. Five robust points for practice:

1. Grain as a carbohydrate source is physiologically unproblematic if it is properly processed. Dogs have adapted genetically to starch.
2. The category grain-free says nothing about quality, balance, or tolerance. The ingredients list does — especially the proportion and ranking of pulses.
3. Boutique brands with exotic protein sources and high pulse content deserve extra caution — not because they are ill-intentioned, but because their risk profiles are less well investigated than those of the large manufacturers.
4. In breeds with known cardiac predisposition (Dobermans and Great Danes for DCM, Boxers for ARVC, among others) a genetic test that clarifies the individual risk is worthwhile — together with a deliberately conservative diet choice.
5. Suspicion of diet-associated DCM is a cardiac emergency. Echo, ECG and a consistent switch to a grain-inclusive standard recipe from a large manufacturer belong to the first-line measures.

Lisa Freeman’s 2018 message, in one sentence: It isn’t just grain. It is about boutique brands, exotic ingredients, high proportions of pulses — and about the fact that the marketing category grain-free carries a safety message that no study backs up, and whose proof has remained outstanding for eight years.

---