Grain-free diets – good or bad?

Introduction

The general term “grain” refers to dry seeds harvested from flowering plants containing either one seed leaf (monocotyledon or monocots) or two seed leaves (dicotyledon or dicots); these can be further classified as cereals, minor cereals (grasses), or pseudocereals (non-grasses) ( Figure 1 ). Examples of dietary grains and their classifications are shown in Box 1 . Gluten, a protein mixture of glutenins and gliadins, is specifically found only in wheat, barley, rye and triticale; oats in themselves are gluten-free, but may be contaminated with wheat during harvesting or processing¹.

Gluten can also be found in some processed sauces, medications and supplements, and processed meats 1. All of this terminology and background information is important to keep in mind when considering grain-free diets for companion animals, because pet owners may voice a desire to feed a “grainfree” diet when in fact they are specifically looking to remove gluten or individual grains without actually considering the entire list of grains to be inappropriate. Especially with respect to dogs, many owners appear to be most concerned with removing cereals (i.e., corn/maize, rice, wheat) from the diet. Individual owners may, however, consider some or all of the minor cereals to be acceptable, and many actually prefer to include pseudocereals in their pet’s diet. Ensuring the veterinary team and the owner are both referring to the same individual food items when using the term “grain-free” will abate any future confusion and increase the likelihood of recommending dietary ingredients with which the owner is comfortable ( Figure 2 ).

¹ Corn gluten meal (CGM) is a byproduct of corn (maize) processing that is used in some countries as an animal feed, but the phrase is misleading; corn contains neither gliadin nor glutenin.

Grain-free diets for cats and dogs have recently increased in popularity, representing 29% of the United States “pet specialty” market sales in 2015 2 and 19% of dog food and 15% of cat food purchases in 2016 3. Many reasons for this feeding trend are espoused, including a desire to feed a more biologically “ancestral” diet, to avoid unnecessary blood glucose fluctuations caused by diets with high-carbohydrate levels, to improve overall diet digestibility and quality, and to avoid food allergies. This article reviews what is known about grain consumption in both cats and dogs related to these points.

Biological ancestral diets

A biological ancestral diet refers to the type of diet that would be eaten by the intended species if it were to live in the wild. For pet dogs, this generally results in a desire to feed the same diet as a wolf, while domestic cats are compared to their wildcat lineage. In winter months, gray wolves preferentially kill and consume large ungulates every 2-3 days, although they often encounter fluctuating food availability 4. After killing their prey, a pack of wolves immediately consumes internal organs, followed by large skeletal muscles. Over the next 48 hours they consume bone, tendon, cartilage and hide, leaving the rumen and unbreakable bones behind. In summer months, their diet becomes more varied, and includes rodents, birds, invertebrates and plant material. The typical macronutrient intake of wolves is 54% metabolizable energy (ME) protein, 45% ME fat and 1% ME carbohydrate 4, while domestic dogs have a preference for dry or canned diets containing 30% ME protein, 63% ME fat and 7% ME carbohydrate ( Box 2 ) 5.

Wildcats have been shown to have a preference for rabbit prey, secondarily consume rodents, and round out the remainder of their diet with insectivores, reptiles, birds and arthropods, based on availability. The main dietary components of a free-ranging feral cat are reported as 78% mammalian, 16% birds, 3.7% reptiles and amphibians, and 1.2% invertebrates, with prey availability also dictating consumption preferences 6. The daily macronutrient intake of feral cats is 52% ME protein, 46% ME fat and 2% ME carbohydrate ( Box 2 ) 6. When domestic cats were offered a variety of dry and canned diets, their preferred macronutrient distribution was 52% ME protein, 36% ME fat and 12% ME carbohydrate ( Box 2 ) 7. Grain-free dry diets for cats contain less carbohydrate compared to grain-containing diets (22.4 ± 5.6%ME vs 30.1 ± 7.7%ME; calculated using 3.5 Kcal/g energy factor; P <0.001) 8. Similar data are not yet available for dog diets.

In addition to feeding a diet that mimics the macronutrient profile typically selected by the species, proponents of grain-free diets argue domestic cats and dogs need this type of diet due to their carnivorous nature. Wolves are classified as generalist carnivores, based on their ability to feed on a diverse array of foods, but typical consumption of prey. Their canine and incisor teeth are used to subdue prey, slash the hide and muscle, and grasp and hold prey while the carnassial pair of teeth (upper fourth premolar and lower first molar) have two shearing edges to trap and cut food in a self-sharpening motion. The back of the lower carnassial tooth and the upper first molar acts as a crushing or grinding surface. The dentition of dogs is very similar, and while some consider them to be carnivores, they have been classified as omnivores by the National Research Council 9. In support of this classification is whole-genome resequencing data reporting that three genes with key roles in starch digestion (AMY2B, MGAM and SGLT1) were targets of selection during dog domestication 10. After domestication, positive selection continued to affect AMY2B copy number in dog breeds based on their level of habitual starch consumption 11.

Cats, however, are obligate carnivores as they are required to obtain several essential nutrients from a diet based on animal tissue 6. In one study, evaluation of domestic cats (Felis catus) revealed genes involved with neural processes (such as behavior and contextual clues related to reward) differ from the wildcat (Felis silvestris silvestris and Felis silvestris lybica) genome; this suggests selection for docility was the major force that first altered the domestic cat genome 12. The authors suggest the modest genetic effect of feline domestication seen in their results are due to a recent divergence from an ongoing mingling with wildcats, a relatively short human cohabitation time, and a lack of clear morphological and behavioral differences from wildcats. In short, there is currently no genetic evidence supporting a divergence in diet-related characteristics between domestic cats and wildcats, although there are slight macronutrient preference differences between domestic and free-ranging feral cats as noted above.

Carbohydrates, blood glucose and diet digestibility

Another common reason some owners prefer feeding a grain-free diet is to limit carbohydrate intake and secondary fluctuations in blood glucose. Although dogs lack salivary α-amylase, which initiates the process of cleaving carbohydrates into oligosaccharides, they have similar carbohydrate digestive and metabolic enzymes found in omnivorous species such as humans. Cats, on the other hand, have numerous differences in their ability to digest, absorb, and metabolize starches and sugars. Details of these metabolic adaptations are beyond the scope of this discussion, but have been recently reviewed 13.

Despite having reduced numbers and types of carbohydrate enzymes, cats are still able to effectively digest and utilize carbohydrates. A study evaluating six different carbohydrate sources found cats had starch digestibility values similar to rats and dogs 14. While cats are capable of digesting carbohydrates, there is still much debate and controversy regarding the long-term effects of high-carbohydrate diets on the development of obesity and diabetes mellitus in this species. There is currently no evidence that dietary carbohydrate content directly impacts on the risk or development of obesity in cats. On the contrary, studies have found cats fed a high-fat or high-protein versus a high-carbohydrate diet gain more fat mass and consume more calories 15. However, some studies do suggest a lower carbohydrate diet better regulates blood glucose concentrations in diabetic cats 16 and can reduce post-prandial insulin and glucose concentrations in healthy cats 15. Whether or not long-term feeding of highcarbohydrate diets contributes to the development of diabetes in cats remains unclear. 

When contemplating the effects of carbohydrates on blood sugar, some consideration should also be made regarding the types of carbohydrates in a diet. In both dogs and cats, carbohydrate sources with higher concentrations of fiber (digestiveresistant starch) and protein tend to have lower glycemic responses 14, 17. For example, corn and brewer’s rice elicit a higher glucose and insulin response in cats than ingredients like peas and lentils 14, so diets with similar carbohydrate contents could have differing metabolic effects. 

Food allergies

Owners may choose grain-free diets for their pets in an effort to avoid food allergies. The term “food allergy” may be defined as an adverse immune response toward food proteins, or an intolerance associated with a hypersensitive immune response, which is repeatable with a dietary challenge 18. Food allergies can be immediate (IgE-mediated), delayed (non-IgE-mediated), or a combination of the two 18. In people, food allergens are water-soluble glycoproteins ranging from 10-70 kD molecular weight and are divided into either class 1 primary sensitizers and/or class 2 sensitizers that are cross-reactive 18. There is a risk for cross-reactivity among food items within a single food family; for example, in humans there is a 75% risk of cross-reactivity among species of shellfish, but legumes have only a 5% risk, whilst grains have a 25% risk 18. Cross-reactivity categories have not yet been verified in cats and dogs, although there does not appear to be cross-reactivity between beef and dairy or soy and wheat in dogs, but there may be reactivity between chicken and eggs 19. For this reason, dietary challenge with specific ingredients should be undertaken to identify actual reactivity, and removing all ingredients within a single food family without such a challenge may be unnecessarily limiting potential diet options. 

In contrast to food allergy, a general food intolerance involves a non-immunologic response to a food that is also repeatable with a dietary challenge 18. A classic example is lactose intolerance, where a deficiency of the enzyme lactase results in an inability to adequately digest foods that contain lactose, with subsequent gastrointestinal signs. It is important to realize that distinguishing between a true allergy versus a dietary intolerance is challenging in companion animals, and it may be preferable to use the term “adverse food reaction (AFR)“.

In dogs, the most common food ingredients reported in association with an AFR in dogs are beef, dairy, poultry, wheat and egg (Figure 3), while the most common ingredients in cats are beef, dairy, fish, lamb, poultry and wheat (Figure 4). As highlighted in one paper 24 it is important to note these data do not reflect the true prevalence of specific food allergies in the population of cats and dogs, since the animals were not challenged with all possible food allergens, and details of challenge protocols used are often unclear. What can be gleaned from these data, however, is that most food allergens in cats and dogs are related to the animal-derived component, rather than the plant-derived component of the diet. In addition, the likelihood of allergy development increases with exposure, so it is possible the most common allergens will change over time if the general companion animal diet is adjusted to avoid the currently implicated ingredients.

One study compared the individual ingredient composition of grain-containing and grain-free dry cat diets available in the United States 8. The most common animal-sourced ingredients in graincontaining diets are largely poultry, followed by fish and egg. Grain-free diets are more likely to contain equal proportions of poultry and fish, with egg rounding out the rest of the majority of animalsourced ingredients (Figure 5). For plant-sourced ingredients, the grain-containing diets were more likely to contain rice, flax, cranberry, oat, carrot, beet, pea, barley, and blueberry, while the grainfree diets generally contained pea, cranberry, potato, carrot, blueberry, flax, sweet potato, tapioca/ cassava, and apple (Figure 5). Based on this information, both types of diet are less likely to contain some of the most common food allergens in cats, including beef, dairy, lamb, wheat, or corn (Figure 5), but are likely to contain fish and poultry, respectively the third and fifth most common ingredients associated with feline AFR. Essentially, this means the allergenic potential of a diet is not really altered simply by feeding a grain-free diet. 

In addition to removing dietary grains for general food allergies, some owners pursue this type of diet to specifically exclude gluten. 1% of the world’s human population is affected by celiac disease, a multisystem immune disorder that is triggered by ingestion of gluten 1. A syndrome referred to as non-celiac gluten sensitivity, in which people without celiac disease or a wheat allergy describe an improvement in symptoms after maintaining a gluten-free diet, has been reported, although whether this is actually a separate and unique condition has yet to be established 1. Some owners that follow a gluten-free diet themselves may choose to remove gluten in their pet’s diet in order to reduce their own risk of reactivity. In dogs, glutensensitive enteropathy has been reported in Irish Setters 26 (but it has been largely eliminated in this breed) and gluten-responsive epileptoid cramping syndrome has been described in Border Terriers 27. For affected animals, switching to a gluten-free diet, which does not necessarily need to be completely grain-free, will likely be helpful. Glutenspecific conditions have not been described in cats.

Grain free diets and DCM

Very recently, the development of dilated cardiomyopathy (DCM) in dogs eating a high proportion of ingredients consistent with grain-free diets (peas, potatoes, lentils, other legume seeds) has been reported 28. While there are known breed predispositions to canine DCM, the U.S. Food and Drug Administration has received reports from atypical breeds fed on grain-free diets.

Half of these cases had low blood taurine concentrations, and taurine deficiency is a known cause of DCM. Dogs are able to synthesize taurine and do not typically have a dietary requirement for the amino acid, but some individuals or breeds may have conditionally essential requirements. It may be that the ingredients in these diets interfere with the bioavailability of taurine or another nutrient(s) yet to be evaluated. At this point, it is difficult to draw any conclusions regarding an association of grain-free diets with the development of DCM, as the number of cases reported is a very small fraction of dogs fed these types of diets, and more investigation is needed.

So are grain-free diets good or bad?

Cats and dogs are both capable of digesting and metabolizing carbohydrates, including grains. As a general rule, placing both species on this type of diet is not inherently problematic, although there are no specific data suggesting it is truly a more optimal feeding plan. While dry grain-free diets for cats may contain less total carbohydrates compared to grain-containing diets, neither diet type is completely carbohydrate-free and may not reflect the macronutrient profile domestic cats and dogs tend to prefer. Importantly, a diet that is low in total carbohydrates must inherently be higher in protein and fat, which may be an inappropriate profile for animals with some medical conditions, including chronic kidney disease and those requiring dietary fat reduction. Simply switching to a grain-free diet may also not improve clinical signs in animals that are actually suffering from an AFR, since these reactions are more likely to be due to animal-sourced ingredients, and two of the most commonly associated ingredients in feline petfoods are likely to be included in dry grainfree diets. If a cat or dog does have an AFR to a specific individual grain, cross-reactivity to all grains has not been demonstrated in these species – and is likely in only 25% of affected people.

The bottom line is that if a domestic cat or dog is thriving on a well-formulated grain-free diet, this type of feeding plan may be continued, although it may be prudent to bear in mind the comments on DCM above. However, if an animal is switched to a grain-free diet in an attempt to feed a more biological ancestral diet, avoid unnecessary blood glucose fluctuations, improve overall diet digestibility and/or avoid food allergies, itmay not actually be the grain cessation that is responsible for any beneficial effects.