Review of critically appraised topics on adverse food reactions of companion animals

Cutaneous adverse food reactions (CAFR) in our patients can be the most frustrating element of dermatology for the veterinarian and the owner, as well as for the pet. We can negatively impact the human-animal bond with restrictive diets, and even affect the long-term behavior of our pets. Yet food triggers play an important role not only in truly food-allergic or dietary intolerant patients, but also in our atopic patients.

Elimination and provocation trials remain the gold standard, and currently the only, definitive way to diagnose adverse food reactions in dogs and cats. A set of critically appraised published topics on adverse food reactions in companion animals summarizes what is known about food allergies. Each of the 8 topics and the conclusions will be reviewed herein. For each topic, a literature search lead to a review of relevant papers, with inclusion and exclusion criteria based on the subject. The conclusions drawn are based on a meta-analysis of the relevant data.

Part 1: Duration of elimination diets

Recommendations have varied between 3 and 12 wk for the duration of elimination diets in dogs and cats. This review included selected articles on dogs and cats with nonseasonal pruritus after the exclusion of other relevant causes of pruritus and a complete or marked (> 50%) reduction of clinical signs after feeding an elimination diet which then had recurrence of signs after provocation with previous diets or ingredients.

Of 209 dogs, 50% had achieved a marked reduction in symptoms after 3 wk on an elimination diet; 85% had returned to normal after 5 wk, and 95% had complete remission of signs when the elimination diet was extended to 8 wk. Fewer than 5% of dogs needed an elimination diet of up to 13 wk. In 40 cats, 50%, 80%, and 90% of cats achieved remission at 4, 6, and 8 wk, respectively.

The conclusion drawn was that in at least 80% of dogs and cats, an elimination dietary trial should last a minimum of 5 wk in dogs and 6 wk in cats. Increasing the duration of the trial to 8 wk will increase sensitivity of diagnosis to more than 90% of cases in both species. In practice, we may also elect to perform provocation trials to confirm AFR soon after remission is achieved, even if remission is less than 8 wk.

With this information, we can closely monitor our pets so as to minimize the impact of an elimination diet on the human-animal bond.

Part 2: Common food allergen sources in dogs and cats

Knowing the most common offending food allergens in dogs and cats is important in selecting elimination diets and allows for sequential food challenges to aid in the diagnosis of CAFRs. In a review of selected publications, once a CAFR was diagnosed, only cases with a positive re-challenge were included. Positive challenges were considered the only solid evidence for identifying an offending allergen.

At least 1 offending allergen was found in each of the 297 dogs and 78 cats included. The most frequently reported food allergens involved in CAFRs in dogs were beef, dairy products, chicken, wheat, and lamb. Less commonly reported sources were soy, corn, egg, pork, fish, and rice. In cats, at least 1 offending allergen was found in each patient. The most frequent sources causing CAFRs in cats were beef, fish, chicken, wheat, corn, dairy products, and lamb.

Although there are limitations to this data, and true prevalence numbers are difficult to identify, it can be assumed that for a dog living in Australia, Europe, or North America, the allergens most likely to be involved in a CAFR are beef, dairy products, chicken, wheat, and lamb. For cats, beef, fish, and chicken are the most likely triggers.

Part 3: Prevalence of cutaneous adverse food reactions in dogs and cats

Selected articles ranged from 1990 to 2005 and included 22 articles on CAFR in dogs, 9 in cats, and 3 which were pertinent to both species. Diagnosis was made after feeding an elimination diet and challenging with offending allergens.

The prevalence of CAFRs in dogs and cats was found to be variable depending on the presentation and diagnosis. For dogs with any disease presenting to the veterinarian, prevalence of a CAFR was < 1%. For dogs with skin disease in general, prevalence was < 6%. In dogs presenting with any type of allergic skin disease and dogs with skin lesions suggestive of atopic dermatitis, prevalence rates jumped to 20% and 29%, respectively. In cats, prevalence was low in the general population (< 1%) and up to 5% in cats with skin disease, 10% in those with allergic skin disease, and 12% to 21% in cats with pruritus.

Despite the limitations of this review, and the inconsistent prevalence rates based on presentation, the prevalence of CAFR in dogs and cats was high enough to justify ruling out CAFR in any patient presenting for allergic or pruritic skin disease. It is estimated that one-third of dogs with atopic dermatitis have concurrent CAFR; diet therefore plays an integral role in many of our allergic patients.

Part 4: Can we diagnose adverse food reactions in dogs and cats with in-vivo or in-vitro tests?

Elimination diets and provocation trials are inconvenient and frustrating. Testing for food-specific reactions would potentially eliminate years of suffering for our patients and owners. Laboratory tests of blood, hair, and saliva are attractive to owners and veterinarians as they are an easier alternative to elimination diets.

From a search of more than 700 articles, 23 papers were chosen for review, which included a wide variety of testing: food-specific IgE in serum of dogs and cats, canine food-specific IgG, lymphocyte proliferation assays, intradermal testing and gastroscopic food testing, patch testing, colonoscopic testing, canine fecal IgE, hair, and saliva testing.

Most of the studies reviewed involved serum allergy testing, and the results were inconsistent. Two studies showed low repeatability of IgE and IgG results, and 1 study found high concentrations of food-specific IgE in a large number of dogs with atopic dermatitis which did not improve after an elimination diet. When the serum test results were correlated with food-provocation outcomes in dogs with CAFRs, accuracy and negative predictabilities of the tests varied widely. Intradermal testing revealed positive results in dogs with no clinical food sensitivities and negative results in dogs with known CAFRs. Lymphocyte proliferation tests were reported as 80% accurate with 100% positive predictability and 50% negative predictability, but currently these tests are only offered in research facilities, and commercial use is limited by volatility of the sample. Patch testing was found to be accurate and negative predictability was excellent. However, this test cannot be used to diagnose CAFR, simply to guide the choice of elimination diet. There was unsatisfactory accuracy with gastroscopic testing, fecal food-specific IgE as well as hair and saliva testing.

The conclusion from this review was that while patch testing with food ingredients might be useful in some selected dogs (to choose the ingredients for an elimination diet), all other tests cannot be recommended for the clinical diagnosis of CAFRs in dogs and cats. Serum allergy testing for food is widely available, but at this time is not reliable and cannot be recommended for dogs and cats. Currently, an elimination diet with provocation trials is the only recommended diagnostic procedure.

Part 5: Discrepancies between ingredients and labelling in commercial pet foods

Elimination diet trials are often conducted with commercial diets and rely on the ingredients on the label to be accurate and complete. Studies have been conducted worldwide, looking for the presence of unlabelled potentially allergenic ingredients, as well as to demonstrate the absence of components written on the label.

The percentage of mislabeled pet foods varied from 0% to 83% (median 45%). In studies testing pet foods advertised as “novel/limited ingredients” or “restrictive” diets the percentage of mislabeling varied from 33% to 83%. Most hydrolyzed protein samples, surprisingly, were not cross-contaminated. Missing ingredients were found in 50% of studies (8/16).

A patient’s reactivity to a food is multifactorial and includes the type of allergen recognized by the patient’s immune system, the amount of mislabeled allergen (i.e., threshold), and the degree of hypersensitivity of the patient itself. Review of the literature suggests that mislabeling of pet foods is common, even in those advertised as novel or restricted. Therefore, using a commercial diet for an elimination diet trial is unreliable.

Part 6: Prevalence of noncutaneous manifestation of food reactions in dogs and cats

Common features of adverse food reactions in dogs are atopic dermatitis and urticaria. Common reaction patterns in cats with AFRs include miliary dermatitis, eosinophilic granuloma complex including oral lesions, and non-inflammatory alopecia. Gastrointestinal signs are often associated with food in dogs and cats, but this review looked at a wide variety of non-cutaneous adverse reactions.

In reviewed studies, improvement of a clinical sign with an elimination diet, and recurrence after re-challenge with the previous diet was considered strong evidence of an AFR. A large number of dogs and cats with AFR were reported to exhibit vomiting and/or diarrhea. In 395 dogs, 93% had diarrhea, 2% had vomiting, and 5% had both. In 40 cats, 45% had diarrhea, 38% had vomiting, and 18% had both. Broadly, AFR-associated vomiting was more common in cats than in dogs. Vomiting and diarrhea were the non-cutaneous signs reported in more than 20% of dogs and cats with AFR.

In dogs, non-cutaneous signs included anaphylaxis, increased frequency of defecation, symmetrical lupoid onychodystrophy, and sneezing. In cats, conjunctivitis, salivating, flatulence, hyperactive behavior, and respiratory signs were reported.

Part 7: Signalment and cutaneous manifestations of dogs and cats with adverse food reactions

Pruritus is widely accepted as the main symptom that affects pets with a cutaneous AFR. Data reviewed from 825 dogs with CAFR showed the age of onset to vary between 1 and 13 y, with a mean onset of 2.9 y. Over 22% of dogs showed symptoms before 6 mo of age, and an onset before 1 y of age was found in 38% of dogs. No sex predisposition was found. There was an over-representation of German shepherds, Labrador or golden retrievers, and West Highland white terriers (WHWT). Pruritus was the dominant symptom in 94% of cases in dogs, which was most often generalized, but also included ventral and ear pruritus. Contrary to previous beliefs, perianal pruritus affected only a minority of dogs (4% to 25%) with CAFR. Between 11% and 70% of dogs had recurrent or chronic pyoderma, and 3% to 69% had otitis externa. Concurrent atopic dermatitis was reported in 13% to 100% of dogs and pyotraumatic dermatitis (hot spots) was found in under 10% of cases. Malassezia dermatitis and perianal fistulae were reported in a surprisingly small number of dogs.

In 210 cats, the age of onset also varied widely ranging between 4 mo and 15 y. Only 9% of cats had symptoms before 6 mo of age and 23% by 1 y. Other than domestic shorthair, Persians, Siamese, and Burmese cats were included. Nearly all cats were pruritic, and face/head pruritus was found in over 50% of cats. Other areas of pruritus were the ears, ventrum, and feet. As with dogs, the perineum was less frequently pruritic. Most common in cats were a presumed self-induced, often symmetrical alopecia (40% to 100%), a head and neck erosive/ulcerative/crusted dermatitis (30% to 65%), miliary dermatitis (21% to 40%), and variants of eosinophilic disease (6% to 23%).

In summary, CAFRs affected cats and dogs of nearly all ages, both sexes, and the onset of clinical signs in dogs was earlier than in cats. Almost 40% of dogs with CAFR developed symptoms by 1 y of age, and 4 of 10 dogs with CAFR were German shepherds, WHWT, or Labrador or golden retrievers. Generalized pruritus was a sensitive, although non-specific symptom of CAFR in dogs and cats. Head and neck pruritus were common in cats and ears, ventrum, and feet were frequently affected in both dogs and cats. The perineum was rarely affected in CAFR, but recurrent bacterial and yeast infections, otitis externa, and atopic dermatitis were common manifestations of CAFR, and could all exist in the same patient.

Part 8: Storage mites in commercial pet foods

Dogs with year-round atopic dermatitis are often sensitized to Dermatophagoides house dust mites (HDM). Storage mites (most commonly Tyrophagus), which can be found in dog foods, can cause flare-ups in symptoms in dogs with HDM sensitivity. Tyrophagus mites can trigger allergy flare-ups in laboratory beagles sensitized with Dermatophagoides HDM. A flare-up of symptoms in an allergic dog challenged with a new diet may be related to storage mites rather than a true food allergy and can lead to misinterpretation of the trigger.

Evaluating the studies that looked at storage mite populations in various brands of dog food the authors concluded that storage mites are found in low populations in newly bought kibble. Storage mites will not proliferate rapidly if food is kept indoors in low humidity, but will proliferate rapidly if kept in warm environments with high humidity (i.e., the garage), and that dogs sensitized to HDM antigen will trigger to storage mite contamination.

Recommendations for dogs sensitized to HDM include storage of dog food in an airtight container, kept cool and indoors, in low humidity in order to decrease the population of storage mites. Additionally, when performing elimination diet trials and provocation testing, fresh dog food should be used in order to reduce the possibility of storage mite contamination.