T. Tapp, C. Griffin, W. Rosenkrantz, R. Muse, and M. Boord Comparison of a Commercial Limited-Antigen Diet Versus Home-Prepared Diets in the Diagnosis of Canine Adverse Food Reaction* Tiffany Tapp, DVM, DACVD a,c Craig Griffin, DVM, DACVD b Wayne Rosenkrantz, DVM, DACVD a Rusty Muse, DVM, DACVD a Mona Boord, DVM, DACVD b a Animal Dermatology Clinic 2965 Edinger Avenue Tustin, CA 92780 b Animal Dermatology Clinic 5610 Kearny Mesa Road San Diego, CA 92111 c Current address: Dermatology for Animals 1480 S. County Trail East Greenwich, RI 02818 ABSTRACT The objective of the present study was to compare the efficacy of a commercial limitedantigen diet (Eukanuba Veterinary Diets Response Formula FP/Canine, The Iams Company) with home-prepared diets in the diagnosis of adverse food reaction in dogs. The study was conducted in two phases. The first phase utilized 28 dogs fed a home-prepared diet to enable a diagnosis of adverse food reaction. Dogs diagnosed from this phase were entered into the second phase in which these dogs were fed the commercial limited-antigen diet. Owners of 10 of the 28 dogs quit Phase 1 before it could be completed, and one case was eliminated because extended treatment with steroids was required. Eight of the remaining 17 (47%) dogs were diagnosed with an adverse food reaction. A reaction occurred in four of *Funding for this study was provided by The Iams Company, Dayton, Ohio. eight dogs fed the test diet in Phase 2. These results demonstrated the difficulties encountered by owners and veterinarians attempting to feed dogs home-prepared diets. In addition, the results of the study suggest that the occurrence of adverse reactions to fish could make it a less desirable ingredient as a limited antigen in canine diets. INTRODUCTION Adverse reactions to food are often classified according to the presence or absence of an immunologic reaction. Food intolerance is a general term describing an abnormal physiologic response to ingested food or food additives that does not include an immunologic mechanism, whereas food hypersensitivity involves an immune-mediated response. 1 5 Both immediate and delayed reactions have been reported in dogs, although an immunologic reaction has not been proven in most clinical reports but has in experimental models. 6,7 Evidence sug- 244
Veterinary Therapeutics Vol. 3, No. 3, Fall 2002 gests most dogs are sensitive to a single or a few dietary substances. 7,8 In this study, adverse food reaction was defined as a nonseasonal pruritic dermatitis that responds to an elimination diet and flares on challenge. No attempt was made to define an immunologic basis to the reaction. Adverse food reactions account for an estimated 1% to 5% of canine dermatoses and are responsible for up to 23% of nonseasonal allergic dermatoses. 2,6 There is little evidence that dogs with atopic dermatitis are predisposed to adverse food reactions. However, a higher percentage of dogs exhibiting an adverse food reaction do have atopic dermatitis. 5 Considering the threshold effect seen in allergic skin diseases, it is possible that adverse food reaction is more prevalent but is underdiagnosed because clinical signs disappear when other allergies are treated. No age, breed, or sex predilection has been conclusively documented, although Rosser reported that the soft-coated wheaten terrier, Dalmatian, West Highland white terrier, collie, Chinese shar-pei, Lhasa apso, cocker spaniel, miniature schnauzer, and Labrador retriever breeds may be at greater risk for developing an adverse food reaction. 9 Historically, the diagnosis of adverse food reaction was made based on an elimination diet and subsequent flare when challenged with the previous diet. 3,9 The length of time recommended for the elimination diet varies from 3 to 10 weeks. 3,9,10 The objective of the elimination diet is to provide dietary substances to which the animal has not been exposed and which are free of additives and preservatives. In recent publications, fish has been suggested as a good protein source. 8,11,12 This finding may be supported by the results of three studies in which none of 78 dogs with adverse food reactions were found to be allergic to fish; however, fish was not the routine challenge material in these dogs. 8,11,13 Another recent review suggested that fish is not highly antigenic and indicated that 68% of all reactions in 253 cases reported in the veterinary literature were to beef, dairy, or wheat. 4 The use of commercial fish diets or home-cooked diets has not been thoroughly evaluated, although some reports suggest the commercial fish diets were effective in eliminating signs of adverse food reaction. 8,11,13 Many previous studies have suggested a home-cooked diet is the most accurate way to test for an adverse food reaction because it eliminates preservatives and additives. 2,3,6,9 Because fish was found in only 33 of 269 dog foods evaluated, 11 many investigators do not attempt to test for a fish allergy. This approach may reflect the reports that fish is considered to be used infrequently in dog foods. Unfortunately, the description of how to diagnose food reactions is oversimplified in scientific reports. However, textbooks do allude to the difficulty that may be encountered in instituting and completing a thorough hypoallergenic diet trial. 3,6 A number of clients that start a trial do not finish, although the dropout rate is not really known. In one study, 327 of 449 dogs evaluated by intradermal testing for atopic dermatitis and flea allergy were fed a home-prepared hypoallergenic diet trial for at least 3 weeks because they failed to have a significant response to flea control or were not deemed flea-allergic. The protein sources most commonly used were horse meat, mutton, and fish. 14 The study was completed by all dogs that were unresponsive to flea control and 10 (3.1%) were determined to have food allergies. Many clients are not willing to perform diet trials and a home-cooked elimination diet is very difficult for many clients to maintain. In addition, the nutritional soundness of these home-prepared diets is often inadequate. 15 Previous studies have suggested that commercial diets may be only partially successful in diagnosing adverse food reactions. 7,9,10 The lack of double-blinded placebo-controlled challenges 245
T. Tapp, C. Griffin, W. Rosenkrantz, R. Muse, and M. Boord and trials has presented some difficulty in accurately assessing the role of adverse food reactions in veterinary dermatology and determining the most common allergens. 4 The present study was designed to compare the efficacy of a commercial limited-antigen diet (Eukanuba Veterinary Diets Response Formula FP/Canine, The Iams Company) with a home-prepared diet in the diagnosis of adverse food reaction in dogs. MATERIALS AND METHODS Test Subjects Each dog in the study was suspected of having an adverse food reaction. Suspicion was based on the presence of a nonseasonal pruritus or the onset of pruritus in a dog less than one year or greater than 7 years of age. Flea allergy was ruled out as a cause of the pruritus because either there was no dorsolumbar involvement or the condition was unresponsive to aggressive flea control. Age, breed, and sex were not used as selection criteria for dogs in the study. Because there is often a variation in appearance and distribution of lesions in adverse food reactions, lesions and specific lesion locations were not required criteria. Clients were informed regarding the requirements for performing a thorough dietary trial and provocative challenge. Only those clients willing to commit to this procedure were included. The study was divided into two phases. Phase 1 consisted of the initial evaluation (history, physical examination) and the selection of a home-prepared diet based upon the dog s dietary history. In addition, during Phase 1 any pyoderma or Malassezia dermatitis was controlled with appropriate therapy. Dogs in Phase 1 were evaluated by one of the clinicians in the facility, and all subsequent exams of that patient were made by the same clinician whenever possible. All dogs were graded on a scale of 1 to 5 for pruritus, erythema, scaling/crusting, pustules, and papules. Location of pruritus, presence of ear involvement, and results of cytologic testing were also recorded. Diets Home-prepared diets consisted of a 50:50 mixture of one protein and one carbohydrate source. The sources of protein (canned tuna in water, pinto beans, tofu, ostrich, peanut butter) and carbohydrates (potato, sweet potato, quinoa, oats) were variable among dogs and were based on the dog's previous dietary history. Care was taken to select items to which the dog had not previously been exposed. The dogs were fed approximately 3 cups of the prepared mixture per 9.1 kg (20 lb) body weight. If palatability appeared to be producing anorexia, owners were allowed to alter cooking methods to improve palatability (e.g., baked potatoes instead of boiled potatoes) and were permitted to add certain flavor enhancers (salt, molasses, flavored oils), if necessary. Owners were given a diary for recording their dog s clinical signs and diet at home. Any dog with a diagnosed bacterial or yeast dermatitis or otitis was treated with an appropriate antimicrobial. Owners were permitted to shampoo the dogs and institute flea control as needed. For ethical reasons, any dog showing extreme pruritus was treated with oral glucocorticoids during the first 4 weeks of the dietary trial period. Any dog requiring additional treatments for extreme pruritus was eliminated from the study. Topical otic preparations were permitted for short-term use (not to exceed the first 4 weeks of the trial). The home-prepared diet was fed a minimum of 6 weeks, up to a maximum of 8 weeks, pending clinical response. For example, if a dog showed greater than 50% clinical improvement by 6 weeks, the dog s treatment and evaluation was terminated at 6 weeks. Conversely, if the degree of clinical improvement was questionable, 246
Veterinary Therapeutics Vol. 3, No. 3, Fall 2002 TABLE 1. Presenting Signs for Dogs Evaluated in Food Study Number (%) with Clinical Sign Number (%) with Clinical Sign Among Clinical Sign Among Dogs Initially Evaluated Dogs Having Food Reactions Pruritus 28/28 (100%) 8/8 (100%) Erythema 28/28 (100%) 8/8 (100%) Scale/Crust 18/28 (64%) 3/8 (38%) Ear involvement 17/28 (61%) 5/8 (63%) Papules 16/28 (57%) 5/8 (63%) Pustules 5/28 (18%) 1/8 (13%) the dog continued participation for the full 8 weeks. Any recurrence of pyoderma or Malassezia dermatitis was treated prior to dietary challenge to preclude any confusion as to the source of the pruritus. Any dog showing exacerbation of pruritus or other clinical signs on challenge followed by remission of those signs when reverted back to the test diet was considered to have had an adverse food reaction and was moved into Phase 2 of the study. This determination was based on physical examination by the clinician as well as input from the owner. At the start of Phase 2, the dogs were challenged by feeding their original food for up to 10 days. During the challenge period, flea control and bathing was continued as it was during Phase 1. Dogs enrolled in Phase 2 were fed the commercial test diet after the clinical pruritus resolved during feeding of the home-prepared diet. The commercial diet was fed according to the manufacturer's guidelines for a period of 8 weeks. Any dog showing an exacerbation of signs while being fed the trial diet was withdrawn from the study. Dogs were examined at the completion of Phase 2, and the owners were asked to pursue challenges to individual components of the commercial test diet to help determine the specific ingredient(s) to which their dog reacted. Dogs with residual pruritus were assessed for atopy. During both phases of the study, owners were instructed to grade each diet according to ease of preparation, palatability, quality of the dog s stool, vomiting, and response to therapy. RESULTS Twenty-eight dogs were enrolled in Phase 1. Presenting signs most commonly observed for the original 28 dogs included pruritus, erythema, and ear involvement. Papules, scaling/ crusting, and pustules were seen less frequently (Table 1). Of the original 28 dogs enrolled, 17 completed Phase 1. Ten owners (35.7%) terminated participation during Phase 1 because of the difficulty they encountered feeding the diet to their dog, and one dog was eliminated because steroids were required beyond the initial 4 weeks. Of the 17 dogs that completed Phase 1, nine showed no exacerbation of clinical signs on challenge with their previous diet and eight were confirmed to have an adverse food reaction when challenged and then returned to the home-prepared diet. Of the eight dogs that continued into Phase 2, four maintained clinical improvement when their diet was changed from the home-prepared diet to the commercial test diet (Table 2). Three of the other four dogs that reacted to the test diet did so within 1 to 7 days. Clinical signs worsened for one dog on the test diet over a 6-week period. This dog was subse- 247
T. Tapp, C. Griffin, W. Rosenkrantz, R. Muse, and M. Boord TABLE 2. Overview of Eight Food-Allergic Dogs in Comparison of Food Reactions for a Commercial Limited-Antigen Diet and Home-Prepared Diets Breed Age Sex Original Commercial Diet West Highland 2.5 yr Female (s) IVD Venison and Potato white terrier (Innovative Veterinary Diets) Golden retriever 1 yr Male (c) IVD Lamb and Potato (Innovative Veterinary Diets) Dalmatian 7 yr Male (c) Natural Choice Lamb and Rice Formula (Nutro Products, Inc.) Milk-Bone biscuits (KF Holdings) Jack Russell terrier 10 mo Female (s) Natural Choice Adult Lamb and Rice Formula (Nutro Products, Inc.) Boston terrier 11 yr Female (s) Iams Lamb Meal and Rice Formula (The Iams Company) Meaty Bones (Heinz) Terrier mix 10 yr Female (s) Moist N Meaty Burger with Cheddar Cheese (Nestlé Purina PetCare Company) Milk-Bone biscuits Lhasa apso 2 yr Female Purina One Lamb and Rice Formula (Nestlé Purina PetCare Company) Milk-Bone biscuits Rawhide African hairless 11 mo Female (s) Science Diet Lamb and Rice (Hill s Pet Nutrition) VitaBones (s) = spayed; (c) = castrated. *Reaction = exacerbation of clinical signs; no reaction = maintained clinical improvement observed with home-prepared diet. This dog was treated for Malassezia dermatitis. This dog was treated for bacterial pyoderma. quently diagnosed with coexistent atopy (based on clinical disease and a positive test for allergies by ELISA conducted on a blood sample). Whether this dog s progressive pruritus was related to its coexistent atopy or to the test diet is unknown. However, on challenge with the pure fish ingredient contained in the test diet, the dog s clinical signs reoccurred within 48 hours, leading to the conclusion that the reaction was to the fish in the diet. The remaining three owners whose dogs worsened while they were being fed the test diet opted to perform ingredient challenges to some degree. One of these dogs reacted to challenge using the pure 248
Veterinary Therapeutics Vol. 3, No. 3, Fall 2002 Home-Prepared Diet Response to Test Diet* Ingredient Challenge Pinto beans and quinoa No reaction Pinto beans and oats No reaction Pinto beans and potato Reaction Reaction to fish ingredient Pinto beans and potato Reaction Reaction to canned tuna Tuna and potato No reaction Ostrich and potato No reaction Tuna and potato Reaction Reaction to lamb, rice with butter; no reaction to tuna; would not eat fish Pinto beans and potato Reacted over 6 weeks Reaction to fish ingredient, eggs, and beef fish ingredient, one dog reacted to canned tuna when challenged, and the final dog would not eat the canned fish but did not react to canned tuna, and repeatedly reacted to the test diet. None of the eight dogs was given corticosteroids during any part of the trial. Two of the dogs were treated with ketoconazole for secondary Malassezia dermatitis, one dog was treated with antibiotics for a secondary bacterial pyoderma, and one was treated for both Malassezia and bacterial dermatitis (Table 2). The questions and the mean scores of the owner s responses regarding ease of preparation, palatability, presence of vomiting, and 249
T. Tapp, C. Griffin, W. Rosenkrantz, R. Muse, and M. Boord stool quality of each diet are shown in Table 3. Based on the scores, owners generally felt the test diet was easier to prepare, slightly more palatable, and produced fewer gastrointestinal signs than the home-prepared diets. These data provide additional support for commercial diets having advantages over home-prepared diets. DISCUSSION The results of this study reveal a surprisingly high prevalence of cutaneous reactions to fish among dogs confirmed to have adverse food reactions. In one previous study, 11 the most common ingredients leading to adverse cutaneous reactions were determined to be beef, soy, chicken, milk, corn, wheat, and egg. The researchers in that study concluded that dogs react more frequently to ingredients commonly included in commercial dog foods. The discovery in the present study that 50% of the dogs reacted to the test diet and 37.5% definitely reacted to fish provides new data suggesting that dogs may be developing adverse reactions to fish due to the increased exposure to fish and fish meal in commercial diets and treats. This hypothesis is beyond the scope of the present study; however, findings in the present study emphasize that the selection of the "limited antigen" in home-prepared and commercial diets may be more critical the more widely used the "limited antigen" is in pet food. The dropout rate and feedback from the owners demonstrate the inherent difficulties in performing a well-designed food trial. Thirtyseven percent of the owners dropped out of the study, in most cases within 1 to 3 weeks of entering Phase 1. Their reasons were varied (dog did not want to eat the diet; dog had gastrointestinal signs [vomiting, diarrhea, constipation] on the diet; owner had difficulty complying with the strict diet; owner felt the diet was too expensive; and owner convinced food was not the problem ), but they serve to emphasize the major limitations in conducting home-prepared food trials. Only one dog was eliminated because of the necessity for administering drugs that were not permitted in the protocol. Of the 17 clients remaining in the study, many also reported the same complaints listed above. Although previous studies 7 estimated an 85% success rate using a commercial diet to diagnose adverse food reaction, the present study demonstrates a 37% chance of not making a diagnosis using a homemade diet solely due to owner compliance. This lower success rate may in fact provide a more compelling reason to use a commercial diet as an initial diagnostic tool in the diagnosis of adverse food reactions. Certainly, the nutritional needs of the patients are more thoroughly addressed with the use of commercial limited-antigen diets. 15 In conclusion, a limited-antigen diet containing fish and potato was inadequate for diagnosing adverse food reaction in four of eight cases evaluated. Although the sample size was small, there are concerns about the possibility of emergence of fish allergies in canine-food hypersensitive patients. This study provides evidence that dog food manufacturers must continue to develop novel ingredients in limited-antigen diets. Just as lamb is no longer considered novel, it is possible that fish is limited as a novel ingredient for the diagnosis of canine adverse food reaction. However, perhaps the greater piece of information from this study is the difficulty for owners in performing a home-cooked food trial over 6 to 8 weeks. Both owners and dogs had difficulty during the home-cooked phase of the study. This study provides valuable information to suggest that using a commercial diet for a food trial may provide better owner compliance than using a homecooked diet. 250
Veterinary Therapeutics Vol. 3, No. 3, Fall 2002 TABLE 3. Comparison of Median Scores Given by Dog-Owners for Home-Prepared Diet and Eukanuba Response Formula FP/Canine in Eight Food-Allergic Dogs* Feature Home-Prepared Eukanuba Response Formula FP/Canine Ease of preparation 2.375 1.00 Palatability 3.625 3.75 Vomiting 1.75 1.125 Stool 1.75 2.125 *Based on the following questions: How easy was it to prepare the new diet? 1 = very easy; 2 = easy; 3 = difficult; 4 = very difficult. How well did your dog eat the new diet? 1 = not at all; 2 = picks at food, does not finish; 3 = slowly but finishes; 4 = normally; 5 = eagerly. Has your dog had any vomiting with the new diet? 1 = never; 2 = rarely; 3 = occasionally; 4 = often. How would you rate the quality of your dog s stool compared with when the dog was eating the previous diet? 1 = worse; 2 = same; 3 = better. REFERENCES 1. Bindslev-Jensen C, Stahl Skov P, Madsen F, Poulsen LK: Food allergy and food intolerance what is the difference? Ann Allergy 72:317 320, 1994. 2. Reedy LM, Miller WH, Willemse T: Food hypersensitivity. In: Allergic Skin Diseases of Dogs and Cats, 2 nd ed. London: WB Saunders; 1997:173 188. 3. MacDonald JM: Food allergy. In: Griffin CE, Kwochka KW, MacDonald JM eds: Current Veterinary Dermatology. The Science and Art of Therapy. St Louis: Mosby; 1993:121 132. 4. Roudebush P: Hypoallergenic diets for dogs and cats. In: Bonagura J, ed. Kirk s, Current Veterinary Therapy XIII. Philadelphia: WB Saunders; 2000:530 536. 5. Hillier A, Griffin CE: The ACVD task force on canine atopic dermatitis (X): Is there a relationship between canine atopic dermatitis and cutaneous adverse food reactions? Vet Immunol Immunopath 81:227 231, 2001. 6. Scott DW, Miller WH, Griffin CE: The skin immune system and allergie skin diseases. In: Muller and Kirk s Small Animal Dermatology. 6 th ed. Philadelphia: WB Saunders; 2000:543 666. 7. Jeffers JG, Shanley KJ, Meyer EK: Diagnostic testing of dogs for food hypersensitivity. JAVMA 198(2): 245 250, 1991. 8. Paterson S: Food hypersensitivity in 20 dogs with skin and gastrointestinal signs. J Small Anim Pract 36: 529 534, 1995. 9. Rosser EJ: Diagnosis of food allergy in dogs. JAVMA 203(2):259 262, 1993. 10. White SD: Food hypersensitivity in 30 dogs. JAVMA 188(7):695 698, 1986. 11. Jeffers J, Meyer E, Sosis E: Responses of dogs with food allergies to single-ingredient dietary provocation. JAVMA 209(3):608 611, 1996. 12. Fadok VA: Diagnosing and managing the food-allergic dog. Compend Contin Educ Pract Vet 16:1541 1544, 1994. 13. Carlotti DN, Remy I, Prost C: Food allergy in dogs and cats. A review and report of 43 cases. Vet Dermatology 1:55 62, 1990. 14. Carlotti DN, Costargent F: Analysis of positive skin tests in 449 dogs with allergic dermatitis. Pratique Med Chirurgicale Anim Compagnie 27:53 69, 1992. 15. Roudebush P, Cowell CS: Results of a hypoallergenic diet survey of veterinarians in North America with a nutritional evaluation of homemade diet prescriptions. Vet Dermatology 3:23 28, 1992. 251