Common Confounders of Dietary Elimination Trials Contain the Antigens Soy, Pork, and Beef
Nutritionists and dermatologists recommend avoiding flavored over-the-counter (OTC) products and medications during dietary elimination trials because those products are thought to contain common proteins that may confound the trial. The objective of this study was to determine if there are soy, pork, and beef antigens in flavored OTC products and medications and, if so, could those antigens be identified. Seven products, three OTC products and four veterinary therapeutics, were tested using enzyme-linked immunosorbent assays (ELISA) for the presence of soy, pork, and beef antigens, in addition to positive and negative controls. All OTC test products produced ELISA results in agreement with their ingredient lists. ELISA testing of veterinary therapeutic products did not agree with either their ingredient lists or product inserts because of other ingredients not listed. Veterinarians should contact manufacturers of oral therapeutics prior to prescribing them to determine other ingredients. Likewise, manufacturers should be contacted regarding “natural and artificial flavors.” Lastly, gelatin capsules may contain either beef or pork proteins and should not be administered during a trial. In conclusion, flavored medications contain the common antigens soy, pork, and beef although they may or may not be listed on the ingredient list or product insert.
Introduction
The primary clinical sign associated with food allergies is nonseasonal pruritus, but may also include dermatologic signs such as otitis externa, bacterial pyoderma, and other primary or secondary skin lesions.1–4 Concurrent gastrointestinal signs associated with food allergies include vomiting, diarrhea, flatulence, and increased frequency of defecation.2,3 Gastrointestinal signs are often intermittent and chronic in nature. In a 2001 study, challenge diets caused concurrent gastrointestinal signs in up to 62% of dogs.4 Another study revealed approximately 67% of dogs with food allergies had concurrent gastrointestinal signs.3
Currently, the definitive diagnosis of food allergies in canines is based on results of a dietary elimination trial followed by a food challenge test. Dietary elimination trials are lengthy (3–13 wk) and require a client’s strict adherence to the prescribed diet.2,4,5 After a successful dietary elimination trial (i.e., resolution of dermatologic and/or gastrointestinal signs), a food challenge test is performed where the dog is fed their previous diet.4,5 The previous diet should cause the original clinical signs to reoccur if the dog has food allergies. Refeeding the successful food from the dietary elimination trial should resolve the clinical signs once again. In general, gastrointestinal signs resolve more quickly than dermatologic signs and often allow for a shorter dietary elimination trial (2–4 wk).4,5 After confirming a diagnosis of food allergy, provocation tests with individual foods can be used to determine the specific allergen(s) in the dog’s previous diet. One protein (e.g., meat, dairy, or grain) can be added to the elimination diet over 1–2 wk periods, and the dog can be monitored for reoccurrence of the original clinical signs with different food additions. The process is time consuming, and once clinical signs reoccur, only the elimination diet must be fed until the clinical signs resolve before starting a new provocation test.
When a dog develops a food allergy, it is most often after chronic exposure to common dietary proteins.5 Therefore, diets used in dietary elimination trials must contain either novel protein sources (i.e., food proteins never before consumed by the patient) or hydrolyzed protein sources (i.e., low molecular weight peptides with reduced antigenicity).5 The diet also cannot contain any amount of the more commonly used protein ingredients in commercial pet foods. The animal proteins commonly used in pet foods are beef, egg, chicken, turkey, fish, lamb, and pork.6 The most common grain proteins are soy, rice, corn, and wheat.6 Dogs have been reported to have food allergies to beef, dairy products, lamb, egg, chicken, wheat, and soy.1,5 Known and unknown consumption of antigens (i.e., common proteins or proteins with prior exposure) during the dietary elimination trial either nullify or confound the results of such a trial and could preclude an accurate diagnosis.5,7 Either intentional or known consumption of suspected antigens occurs when the owner gives an unapproved treat, supplement, or medication during a dietary elimination trial. In contrast, either unknown or unintentional consumption of an antigen occurs when a treat, supplement, or medication contains ingredient residues from previous production runs (i.e., ingredient cross contamination during manufacturing) or when not identified on the ingredient list or product insert.7
Many confounders of dietary elimination trials exist. Veterinary nutritionists and dermatologists recommend avoiding flavored over-the-counter (OTC) products and medications during dietary elimination trials.5,8 Those flavored products include vitamin and mineral supplements, medications (e.g., heartworm, flea, arthritis, etc.), pill administration treats, toothpaste, and dental products.2,5,8 Those products may contain common proteins in the flavorings that may confound the results of the dietary elimination trial, although no such direct testing has been published to date. Those common confounders could be a major obstacle in performing a proper dietary elimination trial and may preclude the diagnosis of food allergies in dogs.
The objective of this study was to determine if there are soy, pork, and beef antigens in flavored OTC products and medications and, if so, could these antigens be identified using enzyme-linked immunosorbent assays (ELISAs). The hypothesis was that flavored OTC products and medications do not contain the antigens soy, pork, and beef unless listed on the ingredient list or product insert.
Materials and Methods
Seven test products, including three OTC products (one flavored vitamin and mineral supplementa and two pill administration treatsb,c) and four veterinary therapeutics (two heartworm preventativesd,e and two arthritis supplementsf,g) were tested during March 2011 for the presence of soy, pork, and beef antigens (Table 1). The test products and controls were purchased from January 22, 2011 to March 3, 2011. A negative control product (unflavored vitamin and mineral supplementh) was selected based on the ingredient list. The positive control products (three dry pet foods) were selected based on the ingredient lists specifically containing a soyi, porkj, or beefk product, regardless of the product name.
Ingredients in bold font are discussed in the text.
BHA, butylated hydroxyanisole; BHT, butylated hydroxytoluene; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid.
A 100 g sample of each product was sealed in a clean plastic bag and was sent to an independent laboratory to test for the presence of soy, pork, and beef antigensl. The laboratory was accredited in meat speciation and allergen testing and was blinded to the samplesm. The laboratory routinely performed ELISA testing in quadruplicate for the presence of antigens in food products, feed, drugs, food additives, and unknown biologic samples.
ELISA for Soy
A US Department of Agriculture approved, amplified, double sandwich ELISA utilizing specific antisoy trypsin inhibitor and other soy protein antibodies that was previously used in the study by Raditic et al. (2011) was selected.7 The ELISA was positive for soy protein and quantitative when compared with a known soy flour protein concentration gradient between 2.5 and 25 parts per million (ppm). The lower limit of detection for the assay was the value of the lowest positive control, which was 2.5 ppm soy flour protein. A sample with <2.5 ppm soy should not be interpreted as having no soy but simply as containing <2.5 ppm. The level of soy protein in a test sample varied according to ingredients and manufacturing. Soy protein that had been either significantly treated or altered with high temperatures, pressure, fermentation, or hydrolysis may not have been detected.
ELISAs for Pork and Beef
An amplified, double sandwich ELISA utilizing a biotinylated beef or pork antibody approved by the US Department of Agriculture was used.9 The beef ELISA test was previously used in the study by Raditic et al. (2011).7 The assay was based on antibodies raised to heat-resistant, species-specific, muscle-related glycoproteins. The ELISAs were specifically designed to detect cooked meat species in cooked meat, meat products, meat and bone meals, and heat processed foods. Test samples with a mean absorbance value minus three standard deviations >0.250 were considered positive. All other samples were negative.
Specificity was based on species-specific reagents tested against a panel of meat samples for cross reaction and was found to produce negative responses to heterologous species samples.9 The cooked pork ELISA was positive for striated muscle of pork only. The cooked beef ELISA was positive for both beef and bison striated muscle because these are closely related species. The lower limit of detection for pork was 4% and beef was 1% in a mixed extract of the target and a nontarget species. There were limitations to the quantification of muscle in a sample because muscle antigens are heat labile. Temperature and cooking times of a pet food product could result in lower antibody/antigen reaction for both pork and beef. Therefore, the intensity of the ELISA reaction was related to the extent of sample processing and the level of antigen present.9
Results
Control Products
The negative control product contained <2.5 ppm soy and was negative for both pork and beef (Table 2). The positive control products were positive for soy, pork, and beef protein, respectively. Hence, the positive control products confirmed the use of the ELISAs to detect the presence of soy, pork, and beef protein in cooked pet foods.
ELISA, enzyme-linked immunosorbent assay. +, indicates a positive test result; −, indicates a negative test result.
Test Products
All test products contained <2.5 ppm soy with the exception of one heartworm producta and one arthritis supplementg that tested positive for soy (3.4 and 9.2 ppm, respectively). The arthritis productg listed “soy protein isolate” and “soybean unsaponifiables” on the ingredient list. Additional information from the manufacturers of the veterinary therapeutic products is presented in Table 3. The heartworm productsd,e only listed therapeutic ingredients on their product inserts; however, after contacting the manufacturers, other ingredients were identified including “soy protein” as a binding agent in the manufacturing of one productd and “soy” as part of the “artificial beef flavor” for the other heartworm producte. That said, the second heartworm producte contained <2.5 ppm soy.
Ingredients in bold font are discussed in the text.
Four test productsa,e,f,g tested positive for pork protein, and test productsb,c,d tested negative for pork protein. One producta indicated “pork liver meal” on the ingredient list. After contacting the manufacturer of one of the heartworm productse, it was confirmed that the “artificial beef flavor” was made using “pork liver.” One arthritis productf contained “gelatin” on the ingredient list, which the manufacturer indicated could be either from bovine or porcine origin. The other arthritis productg listed “natural flavors” on the ingredient list and after contacting the manufacturer, the flavoring was identified as a spray-dried digest derived from “pork liver.”
Two test productsb,d tested positive for beef protein and all other test products tested negative. One of the productsb indicated “beef” on the ingredient list, and the other test productd only listed therapeutic ingredients. After contacting the manufacturer of the therapeutic product, “beef” was identified as a component of the chewable heartworm preventative product. The two arthritis productsf,g contained “sodium chondroitin sulfate” on their ingredient lists that was derived from bovine tracheal cartilage; however, both products tested negative for beef. Last, one of the arthritis productsf contained “gelatin” from either porcine or bovine sources according to the manufacturer; however, the sample tested negative for beef.
Discussion
Sources of Soy, Pork, and Beef
The source of protein antigens in the ingredients “soy protein isolate” or “soy protein” would be the proteinaceous soy component prepared from dehulled soybeans after removing the majority of the nonprotein components.10 Likewise the source of protein antigens in the ingredient “soy” or “soybean meal” would be the proteinaceous soybean component. The source of protein antigens in the ingredient “soybean unsaponifiables” would be potential proteinaceous contamination from soybeans as this product is made from the unsaponifiable fractions of soybean oil. Unsaponifiable matter is defined as the ether soluble matter extractable after complete reaction with a strong alkaline product.10 Refined soybean oil should be devoid of protein; however, soybean oil that is less purified (e.g., cold pressed) may be contaminated with soy proteins.11,12 A soy protein concentration ≥10 ppm may invoke a reaction in an allergic person; however, the concentration for invoking an allergic response in dogs has not been determined.7,13
The source of protein antigens in the ingredient “pork meal” would be the rendered flesh from slaughtered porcine tissues, exclusive of any added blood, hair, hoof, horn, hide trimmings, manure, and stomach contents.10 The source of protein antigen in the ingredients “pork liver meal” or “pork liver flavor” would be the liver derived from slaughtered porcine.10 The source of porcine protein antigens in the ingredient “gelatin” would be pork skins.14
The source of protein antigens in the ingredient “beef” would be the flesh derived from slaughtered bovine limited to striated muscle or that found in the tongue, diaphragm, heart, or esophagus with and without normally accompanying fat, skin, sinew, nerve, and blood vessels.10 The source of protein antigens in the ingredient “beef tallow” would be potential proteinaceous contamination while harvesting fat from slaughtered bovines.10 The potential source of bovine antigens in the ingredient “gelatin” would be cattle bones and/or hides.14
Pill Administration Treats
Only one test productc , a pill administration treat, did not contain any of the common dietary proteins (soy, pork, or beef). That product was specifically marketed as an “allergy formula” and was manufactured with the proteins “duck” and “dried peas.” Both duck and peas could potentially be novel proteins to a patient requiring a dietary elimination trial after obtaining a thorough dietary history. Interestingly, that product bears the same company name as the other pill administration treatb; however, those two products were manufactured at different facilities based on the package labels. Only one bag of those products was tested for the presence of common dietary proteins in this study; however, the results indicated that test productc may be suitable for dogs on a dietary elimination trial using the ingredients duck and pea. Due to the availability of that OTC product and others with similar ingredients, both duck and pea may soon not be considered novel for patients requiring dietary elimination trials.
Veterinary Therapeutics: Heartworm Preventatives
The main source of discrepancy between the ingredient list and/or product insert and the ELISA test results occurred with the veterinary therapeutic products. Two test productsd,e only listed therapeutic ingredients on their product inserts. Samples of those test products were submitted for ELISA testing prior to contacting the manufacturers for additional information on the ingredients so the interpretation of the study results were not biased. After ELISA testing was complete, both of those manufacturers provided additional information on other protein ingredients in their products. One productd was positive for beef and contained 3.4 ppm of soy (as a binding agent). The other producte was positive for pork (from “pork liver” as part of the “artificial beef flavor”) but contained <2.5 ppm of soy (“soy” included as part of the “artificial beef flavor”). Most likely, the methodology used to manufacture the “artificial beef flavor” altered the structure of the soy proteins making them unrecognizable to the ELISA antibodies or the product flavoring was formulated with <2.5 mg of soy/kg of product (2.5 mg/kg = 2.5 ppm). A false negative result would be less likely as the ELISA tests were run in quadruplicate.
A study involving 12 dogs with confirmed food allergies were provocation tested with one of those test productse.15 An increased clinical score (i.e., severity of skin and otic disease) was observed in 10 of the 12 dogs after feeding (i.e., provocation testing) the product, which was a significant increase (P < .05) when compared with clinical scores prior to the provocation test. The protein responsible for the return of the clinical signs was not identified (pork versus soy). Those findings are interesting because this small clinical trial revealed that feeding the test producte could invoke an allergic response in dogs with confirmed food allergies, and the authors of the current study suggest that there is potential for other flavored OTC products and medications to produce similar results based on the results of ELISA testing.
The lack of information on other protein ingredients in the product inserts for therapeutic medications leads to confusion on the part of pet owners who cannot easily determine what common proteins are present in those products. Many owners are aware of the importance of monthly heartworm preventatives, such as the two productsd,e included in the current study and could easily overlook the fact that those therapeutic products contain common proteins that will likely confound their dog’s dietary elimination trial. It is the responsibility of the veterinarian prescribing the dietary elimination trial to discontinue flavored oral medications and prescribe either nonflavored oral medications or topical therapy where appropriate during the dietary elimination trial.
Veterinary Therapeutics: Arthritis Supplements
Two test productsf,g provided full ingredient lists; however, there were several ingredients that could have acted as points of confusion for owners and veterinarians. One of those productsf contained “gelatin,” which could be from either porcine or bovine sources according to the manufacturer. That batch of product only tested positive for pork protein on ELISA.
The most probable explanation for the negative beef protein test is that the particular batch of gelatin capsules was composed of only porcine materials because the majority of gelatin produced in North America is porcine in origin.14 Additionally, the ELISA tests used in this study were designed to detect striated beef and pork muscle. Bovine gelatin is manufactured from bones and hides. Lastly, the processing involved in manufacturing bovine gelatin from bones and hides could alter the structure of the beef proteins, making them unrecognizable to the ELISA test.14 Pork skins used to make porcine gelatin undergo less processing, which may explain the positive pork but negative beef results in these capsulated products.14
Ultimately, gelatin capsules must be considered another common confounder for dietary elimination trials. Owners should be instructed to open gelatin capsules and pour the unflavored, powdered medications/supplements onto their dog’s food during a dietary elimination trial and discard the gelatin capsule. Powdered medications/supplements are often less palatable than those in capsule form, presenting a challenge for owners administering these products.
Additionally, both arthritis productsf,g contained sodium chondroitin sulfate (derived from bovine cartilage), and neither product tested positive for beef protein on ELISA testing. According to the manufacturer, the technology used to extract the sodium chondroitin sulfate involves a chemical extraction method to purify the chondroitin sulfate from bovine tracheal cartilage leaving no remaining bovine proteins; however, the specific methodology is proprietary per the manufacturer (J. Parr, oral communication, February 2012). Owners and veterinarians may find this confusing when selecting arthritis supplements while pets are on a dietary elimination trial and may assume that the sodium chondroitin sulfate is a source of beef proteins.
Natural and Artificial Flavors
Another point of confusion for owners and veterinarians would be the ingredient “natural flavor” included in some of the test productsb,c,g. “Natural” is defined as an ingredient derived from plant, animal, or mined sources, either unprocessed or having undergone processing (physical, heat, rendering, purification, extraction, hydrolysis, enzymolysis, or fermentation) but that has not undergone chemical/synthetic processing.14 The manufacturers indicated that the “natural flavors” in one pill administration treatb was “beef broth” and was “duck broth” in the other pill administration treatc (J. Parr, oral communication, May 2012). The “natural flavors” in the arthritis supplementg was “pork liver” in a spray-dried form.
Further confusion could arise from “artificial flavor” such as “artificial beef flavor” manufactured from pork liver and soy as in the case of one test productb. Thus, manufacturers should be contacted directly regarding both “natural flavors” and “artificial flavors” before using products containing these ingredients during a dietary elimination trial.
Conclusion
All OTC test products (the vitamin/mineral supplement and pill administration treats) produced ELISA results in agreement with their stated ingredient lists. As a result, the duck and pea formula pill administration treat may be appropriate for dogs on a duck and pea elimination diet trial. The results of ELISA testing of veterinary therapeutic products did not agree with their ingredient lists/product inserts because other ingredients were included that were not listed. After directly contacting the manufacturers of the veterinary therapeutics, they readily provided the protein components of those other ingredients. Thus, veterinarians should contact manufacturers of oral therapeutics prior to prescribing them during a dietary elimination trial to determine the other ingredients in those products. Likewise, manufacturers should be contacted regarding “natural flavors” and “artificial flavors” before feeding products containing those ingredients. Lastly, gelatin capsules may contain either beef or pork proteins; therefore, gelatin capsules should not be fed during a dietary elimination trial. Thus, the study authors’ hypothesis was rejected due to the fact that flavored medications contained the common antigens soy, pork, and beef, although they were not listed on the ingredient list or product insert.
Contributor Notes
J. Parr’s present affiliation is Department of Clinical Studies, Ontario Veterinary College, University of Guelph, Guelph, Ontario, Canada, and On Parr Nutrition, Inc., Guelph, Ontario, Canada.
R. Remillard’s present affiliation is Veterinary Nutritional Consultations, Inc., Hollister, NC.
