Nara
← Evidence library

Nara evidence guide

Does grain-free dog food cause heart disease?

A 2026 Dog Aging Project study found grain-free feeding did not predict new owner-reported cardiac diagnoses, but formulation-level DCM concerns remain. The evidence, Nara’s interpretation, and the limits of what can be concluded remain visibly separate.

Evidence confidenceModerate
NutritionReviewed September 21, 2026

Published September 21, 2026 by Nara. Evidence is separated from Nara’s interpretation, and limitations remain part of the conclusion.

The best current population evidence does not support "grain-free" as a stand-alone cause of heart disease in dogs. A large 2026 Dog Aging Project analysis found no statistically significant association between grain-free feeding and new owner-reported cardiac diagnoses over its follow-up period. But that does not prove every grain-free formula is equally safe, because the study did not isolate individual ingredients or specific heart diseases such as dilated cardiomyopathy (DCM). 1

The more useful question is not simply whether a food contains grain. It is what replaces the grain, how the whole diet is formulated and tested, and what is happening in the individual dog.

What did the 2026 Dog Aging Project study find?

The new study is important because it asks a much broader question than many of the earlier case reports: across a large longitudinal cohort, did grain-free feeding predict a new owner-reported cardiac diagnosis?

The answer was no. Across multiple statistical models, the researchers did not find a statistically significant relationship between grain-free diet status and new owner-reported cardiac diagnoses. In an incident-only model that included measured confounders, the odds ratio was 1.05, with a 95% confidence interval of 0.91 to 1.21 and p = 0.52. 1

That is reassuring evidence against a simple rule that grain-free food, as a category, causes heart disease.

It is also not the final word on diet-associated DCM. The outcome was a broad owner-reported cardiac diagnosis rather than veterinarian-adjudicated DCM, and the analysis could not tell us whether a particular food with a particular amount of peas, lentils, potatoes or another ingredient changes risk. The study also highlighted a basic measurement problem: owners do not always classify their dog's food as grain-free or grain-inclusive correctly. 1

So the result is strongest for this conclusion: the grain-free label by itself is a poor predictor of cardiac risk.

Why was grain-free food linked to DCM in the first place?

In 2018, the U.S. Food and Drug Administration began investigating reports of DCM in dogs eating certain diets. Many reported foods were labelled grain-free and contained peas, lentils, other pulses or potatoes high in the ingredient list. More than 90% of products reported to the FDA in its 2019 case summary were grain-free, and 93% contained peas and/or lentils. FDA investigation

Those reports were a safety signal, not an incidence study. Owners and veterinarians were more likely to report cases once the issue became public, the dogs were not randomly assigned to diets, and the denominator of healthy dogs eating the same foods was unknown.

The FDA has also been careful not to reduce the issue to grain avoidance. It has received reports involving both grain-free and grain-containing diets, has said that legumes and pulses are not known to be inherently dangerous, and has no definitive information that the diets associated with reports are inherently unsafe and need to be removed from the market. 2

What evidence still supports a diet-associated DCM problem?

The newer Dog Aging Project result does not erase the clinical and experimental evidence that some diets or formulations may matter.

One influential 2018 study described 24 Golden Retrievers with both taurine deficiency and DCM. Twenty-three of the 24 affected dogs were eating diets that were grain-free, legume-rich, or both. After treatment that included diet change and taurine supplementation, 23 of 24 dogs showed echocardiographic improvement and normalized whole-blood taurine. 3

That is clinically meaningful, but it cannot tell us that removing grain caused the disease. These were selected affected dogs rather than a random sample, and diet change and taurine were given together.

A later prospective study followed dogs with DCM or subclinical cardiac abnormalities after they changed diet. Dogs that had been eating nontraditional diets showed small improvements in several echocardiographic measurements after the change. Most also received taurine, and none had low plasma or whole-blood taurine at enrollment. The study supports the idea that a diet-responsive form of cardiac dysfunction can exist without proving that the grain-free label is the causal feature. 4

Another small study of dogs with subclinical abnormalities found larger improvements in some cardiac measurements after dogs eating grain-free diets were changed to grain-inclusive foods. Again, that is evidence worth taking seriously, but it came from a selected clinical population rather than healthy dogs randomly assigned to grain-free versus grain-inclusive feeding. 5

What happens when specific ingredients are tested more directly?

Controlled feeding studies make the picture more complicated.

In a 2023 short-term feeding experiment, dogs eating a research diet containing peas developed several subclinical cardiac changes, including reduced stroke volume and cardiac output and increased NT-proBNP, while a lentil-containing diet did not produce the same pattern. Taurine status did not fall. 6

That study is useful because it points away from a simple grain/no-grain explanation. But it lasted only 28 days and measured early physiological changes rather than clinical DCM.

A longer randomized study published in 2025 fed 60 healthy dogs one of four complete diets for 18 months. The diets included grain-free formulations containing potatoes and peas or potatoes alone, as well as grain-inclusive formulations with and without peas. The researchers reported no clinically significant between-group differences in cardiac troponin or NT-proBNP, taurine remained normal, and the tested formulations supported normal cardiac function over the study period. All authors were current or former employees of Hill's Pet Nutrition, an important commercial context when interpreting the result. 7

Another controlled study found that nutritionally balanced diets containing up to 45% whole pulses did not impair cardiac function or sulfur-amino-acid status over 20 weeks in healthy adult dogs. 8

Taken together, these studies do not support the idea that peas, lentils or potatoes are automatically cardiotoxic. They do support the more cautious possibility that ingredient amount, processing, amino-acid availability, the rest of the formulation, the individual dog, or interactions among these factors may matter.

Why "grain-free" may be the wrong label to focus on

"Grain-free" tells you what is missing from a food. It does not tell you what replaced it.

Two grain-free foods can have very different ingredient patterns, protein sources, pulse concentrations, fiber profiles, amino-acid availability, energy density and formulation histories. Two grain-inclusive foods can also be very different from each other.

This helps reconcile apparently conflicting research. A large population study can find no general grain-free association while a clinical series or controlled feeding study still finds a problem with a narrower diet pattern. Those are different questions.

Nara's interpretation is that diet-associated DCM is better treated as a formulation-level question than a binary grain-free question.

What should an owner do with a grain-free food?

For a healthy dog doing well on a nutritionally complete grain-free food, the 2026 Dog Aging Project study is reassuring. It does not support panicking or assuming that the word "grain-free" means the dog's heart is being damaged.

It also does not make every grain-free product interchangeable. Useful questions include:

  • Is the food nutritionally complete for the dog's life stage?
  • Who formulated it, and how does the manufacturer validate nutrient adequacy and consistency?
  • What ingredients replace the grains, and how prominent are pulses or potatoes in the formula?
  • Is the dog maintaining appropriate body and muscle condition?
  • Has the dog developed exercise intolerance, fainting, coughing, unexplained weakness or another clinical concern that deserves veterinary assessment?

If a dog already has DCM or another meaningful cardiac abnormality, the exact diet history becomes clinically relevant. A veterinarian or veterinary cardiologist may recommend a diet change and, in selected cases, taurine testing or other diagnostics. The published diet-responsive DCM studies do not justify starting taurine supplementation solely because a healthy dog eats grain-free food. 3 4

For broader food comparison, see how to check whether a dog's food meets calorie and protein needs and how Nara estimates carbohydrates in dog food.

Nara's interpretation

The newest evidence makes a simple "grain-free causes DCM" message harder to defend. It is more accurate to say that a diet-associated form of cardiac dysfunction appears to exist, but grain-free status alone does not identify which diets cause it or which dogs are susceptible.

That is why Nara would rather preserve the exact food and formulation context than collapse a diet into a grain-free yes/no flag. Over time, the dog's actual food, calorie intake, body condition, activity, symptoms and objective health measurements can be interpreted together.

Build your dog's free Snapshot.

Evidence confidence

Moderate overall. Confidence is moderate that grain-free status alone is not a useful general predictor of new cardiac diagnoses: the large 2026 Dog Aging Project analysis is reassuring, and the FDA itself has not concluded that all grain-free diets are causally unsafe. Confidence is also moderate that some diet-responsive cardiac abnormalities are real, because multiple clinical and feeding studies show improvement after diet change or physiological differences between formulations. Confidence is lower on the exact causal ingredient, mechanism and individual long-term risk.

Sources and editorial note

Research and writing were AI-assisted. Sources were checked on 21 September 2026. This article has not been independently reviewed by a veterinarian. The evidence includes observational cohorts, selected clinical cases and controlled feeding studies that answer different questions; some feeding research also has pet-food-industry employment, ingredient-industry support or other commercial relationships. Those limitations are reflected in the interpretation above.

  1. O'Brien J, Bishop M, Dog Aging Project Consortium, Tolbert MK, Ruple A. Multiple Modelling Methods Suggest That Grain-Free Diets Are Not Predictive of New Owner-Reported Cardiac Diagnoses Over Two Years in the Dog Aging Project, Despite Owner Misclassification of Diet. Veterinary Medicine and Science. 2026;12(5):e71213.
  2. U.S. Food and Drug Administration. Questions & Answers: FDA's Work on Potential Causes of Non-Hereditary DCM in Dogs.
  3. Kaplan JL, Stern JA, Fascetti AJ, et al. Taurine deficiency and dilated cardiomyopathy in golden retrievers fed commercial diets. PLoS ONE. 2018;13(12):e0209112.
  4. Freeman LM, Rush JE, Adin D, et al. Prospective study of dilated cardiomyopathy in dogs eating nontraditional or traditional diets and in dogs with subclinical cardiac abnormalities. Journal of Veterinary Internal Medicine. 2022;36(2):451-463.
  5. Haimovitz D, Vereb M, Freeman LM, et al. Effect of diet change in healthy dogs with subclinical cardiac biomarker or echocardiographic abnormalities. Journal of Veterinary Internal Medicine. 2022;36(3):1057-1065.
  6. Quilliam C, Reis LG, Ren Y, Ai Y, Weber LP. Effects of a 28-day feeding trial of grain-containing versus pulse-based diets on cardiac function, taurine levels and digestibility in domestic dogs. PLoS ONE. 2023;18(5):e0285381.
  7. Morris EM, Stiers CA, Hancock LB, Gross KL. Different carbohydrate sources in dog foods supported overall health and cardiac function: an 18-mo prospective study in healthy adult dogs. Journal of Animal Science. 2025;103:skaf225.
  8. Singh P, Banton S, Raheb S, et al. The Pulse of It: Dietary Inclusion of Up to 45% Whole Pulse Ingredients with Chicken Meal and Pea Starch in a Complete and Balanced Diet Does Not Affect Cardiac Function, Fasted Sulfur Amino Acid Status, or Other Gross Measures of Health in Adult Dogs. Journal of Nutrition. 2023;153(5):1461-1475.