Determination of Mammalian Deoxyribonucleic Acid in Commercial Canine Treats and Supplements
ABSTRACT
Feeding an elimination diet exclusively is currently the only accurate diagnostic test for an adverse food reaction in dogs and cats. However, owner compliance has been identified as a challenge, and the inability to limit exposure to other items (including treats and supplements) is a remarkable reason for failure. The objective of the current study was to evaluate the presence of declared and undeclared mammalian deoxyribonucleic acid (DNA) in commercially available canine treats and supplements using polymerase chain reaction methodology. Eight treat products and 20 supplement products were analyzed for the DNA of 10 mammalian species (bison, cat, cow, dog, goat, horse, mouse, rat, pig, and sheep). The results showed that 88% (7/8) of treats and 40% (8/20) of supplements were found to contain at least one source of undeclared mammalian DNA. Undeclared pig and cow DNA were the most frequently identified, and there were only two instances of negative results for declared species. Because of the frequent finding of undeclared mammalian DNA in the assessed products, avoiding using treats and supplements during elimination trials is recommended.
Introduction
The term adverse food reaction (AFR) comprises both immunological and nonimmunological reactions.1 Immunological processes referred to as food allergies may include type I, III, and IV hypersensitivity reactions, which are elicited by exposure to a particular food substance (antigen), typically glycoproteins. Food intolerance describes a nonimmunological response to an ingested food substance, such as food toxicity, food poisoning, and metabolic food reactions.1,2
Ingredients most commonly associated with AFR in dogs include beef, dairy products, wheat, lamb, egg, and chicken.1,3 Factors that determine the allergenicity of foods are not completely clear but may be influenced by inherent food characteristics, processing, and chronicity of exposure. The commonality of these ingredients in pet foods is likely, in part, responsible for these components being routinely implicated rather than their greater allergenic potential per se. In fact, AFR is usually nonseasonal and often occurs after months or years of consuming the diet containing the inciting antigen(s). Patients may present with clinical signs of dermatological disease (commonly pruritus with self-trauma), gastrointestinal disease (vomiting, diarrhea, increased fecal frequency), or both. Accurate diagnosis relies on an elimination diet trial of up to 12 wk leading to the resolution of clinical signs, followed by an intentional food challenge.2
The elimination trial consists of feeding exclusively novel or hydrolyzed ingredients. Veterinary therapeutic diets for the diagnosis of AFR typically include uncommon ingredients, such as venison or rabbit, and/or contain hydrolyzed ingredients, which are manufactured using enzymatic or chemical hydrolysis to reduce peptide size and thus make them less antigenic. However, identification of what is novel for any given individual is entirely dependent on the accuracy and extent of the dietary history obtained, and all of the uncommon ingredients currently used in veterinary therapeutic diets are also available in over-the-counter foods.
One challenge in achieving an accurate diagnosis of AFR is owner compliance. In one study, suspected food-allergic dogs were prescribed a homemade diet elimination trial for 6–8 wk. Of the 28 dogs enrolled in the study, 10 did not complete the trial.4 This is not uncommon, and there are many reasons for noncompliance including cost, time of preparation, intolerance or dislike of the selected food items, inability to limit exposure to other items (including treats and supplements), exacerbation of clinical signs, poor veterinarian and client communication, or a client who doubts that food is a cause of the clinical signs.2
Another challenge is ensuring that the elimination diet only provides known ingredients. A critical feature is the inclusion of only a limited number of ingredients (one protein source and one starch source) to assist veterinarians in identifying a diet in which all ingredients are novel to the pet and to better assess tolerance of individual components. There are several studies reporting undeclared plant or animal species found in limited ingredient, hydrolyzed, vegan, and vegetarian pet foods.5–12 In addition, there is concern that exposure to unknown ingredients can occur from the use of treats, flavored medications, and supplements. One study reported the analysis of seven veterinary products (flavored medications, pill-administration treats, and supplements) using enzyme-linked immunosorbent assays (ELISA).13 The authors reported the presence of soy, pork, and/or beef, which was undeclared on the label or product insert for 4/7 products, although accurate content information was available from the respective manufacturers.13 Despite the fact that the amount of material could not bequantified, that small study confirmed the concern that even veterinary-recommended products may contain ingredients that are not listed on product labeling.13 The amount of allergen necessary to provoke clinical signs remains unknown. However, even small quantities may be relevant for sensitive patients, and polymerase chain reaction (PCR) methodology is superior to ELISA for detection of very small quantities of material.14
The objective of the current study was to use the PCR methodology to evaluate the presence of declared and undeclared mammalian deoxyribonucleic acid (DNA) in canine supplements and in treats with label claims suggesting appropriate use in dogs with AFR. The aim was to establish whether these products are appropriate for diagnosis and long-term management of patients with known or suspected AFR. Our hypothesis was that undeclared mammalian DNA would be detected in the majority of canine supplements and treats evaluated.
Materials and Methods
Sample Selection
Inclusion criteria stipulated that products had to be commercially available from an online source, Chewy (chewy.com), at one time point in January 2020. For treats, the label had to include one or more of the following terms in the product names or marketing language: limited ingredient (specifying the use of a single protein source), hydrolyzed, food sensitivities, pure, or real. Supplements were selected based on popularity (based on the website’s “bestselling” sort tool) and either did not declare any specific protein source or had at least one declared animal protein source. The selected supplements did not necessarily have to contain 1 of the 10 species detected by the PCR panel available at the Veterinary Genetics Laboratory, University of California, Davis (Meat ID test) because these were able to be analyzed for undeclared mammalian species. A total of 8 canine treatsa,b,c,d,e,f,g,h and 20 canine supplementsi,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,aa,bb were purchased.
Analysis Procedures
Label information from each product was recorded, and labels were photographed on all sides. Only new, unopened products were used.
To ensure even distribution of the contents, all products were mixed before opening. For sample collection, only one item was opened at a time. The workspace was cleaned between each sampling procedure, and new nitrile gloves were used to collect each sample. For treats and for supplements in the form of tablets and chews, a small amount of sample was placed into individual unused resealable plastic bags. Using a hammer, samples were macerated for 30 s. The hammer was washed between samples. Macerated samples, powdered products, and capsules were placed into sterile polypropylene centrifuge tubes. Both the capsule and its contents were included. Enough sample was collected to fill to the indicator of 5 mL on the tube. Each tube was labeled with a letter and number code correlating to the specific product to ensure that the laboratory was masked to sample identities. The samples were stored in a controlled laboratory environment at room temperature until they were processed and submitted on the same day to the laboratory for processing and analysis.
Extraction of DNA was accomplished by digesting 0.3 g of material in 1.0 mL of 200 mmol/L NaOH at 95° C for 10 min followed by neutralization with an equal volume of 200 mmol/L Tris-HCl (pH 8.5). PCR amplification was performed in 25 μLreactions on a thermal cyclercc using 1 μL DNA extract, 0.2 μL Taq polymerase,dd 2.5 μL Taq polymerase PCR Buffer,ee 2.5 μL2.0mmol/Ldeoxyribose nucleotide triphosphate (dNTPs),ff 7 μL primer mix, and molecular grade watergg to the final volume. The PCR began with a 1 min activation step at 95° C followed by 31 cycles of 30 s at 95° C, 30 s at 60° C, 1 min at 72° C, and a final extension at 72° C for 30 min. Aquantity of 1 μL of PCR product was then serially diluted out to 1:100 into deionized water, and 1 μL of each dilution was further diluted into 10 μL highly deionized formamidehh and 0.0625 μL size standardii. Fragment separation was carried out on a DNA analyzerjj using the GeneMapper36_POP7 run module and a 10 s injection. Fragment analysis was performed using a computer-based software.kk
The primer mix incorporated three fluorescently labeled universal primers that anneal to highly conserved regions of the mitochondrial cytochrome b gene.15 Reverse species- or genus-specific primers, two for each target species, were included in the primer mix with the dye-labeled primers to create a multiplex capable of detecting DNA of bison (Bison bison), feline (Felis catus), bovine (Bos taurus), canine (Canis sp.), caprine (Capra hircus), equine (Equus sp.), murine (mouse: Mus musculus and rat: Rattus norvegicus), porcine (Sus scrofa), and ovine (Ovis sp.) species. The primers produce fragments of 247 and 257 base pairs (bp) for bison, 98 and 183 bp for cat, 92 and 286 bp for cow, 172 and 305 bp for dog, 275 and 316 bp for goat, 212 and 334 bp for horse, 188 and 364 bp for mouse, 193 and 312 bp for rat, 202 and 220 bp for pig, and 101 and 338 bp for sheep.
The positive control was made up of known DNA from the species listed above. This was run on the same plate as the samples, separated by at least six lanes. The positive control passed if both signals were detected for each species in the test. The negative control was run on the same plate as the samples but separated by at least six lanes. It consisted of water and all reagents that were present with the samples. The negative passed if there were no peaks present.
Statistics
Descriptive statistics (median and range) were calculated using computer-based software.ll
Results
Treats
Labels of the treat products contained the terms limited ingredient (2/8 treats), hydrolyzed (2/8 treats), food sensitivities (5/8 treats), and pure or real (3/8 treats). Five products (5/8; 63%) declared one main mammalian protein source on the ingredient list, one of which also included a nonmammalian protein source (Table 1). These were beef (n = 2), lamb (n = 2), and bison and chicken (n = 1). Of those, all were positive for the DNA of the declared mammalian species, and all were also positive for at least one other species (Table 1).
The remaining three products (3/8; 37%) did not declare animal sourced ingredients (n = 2) or only declared a nonmammalian ingredient (n = 1). Of those, only one was negative for all tested mammalian DNA.
In total, almost all treat products (7/8; 88%) were found to have the presence of DNA from at least one undeclared mammalian species. The product names or marketing terms on the labels of all seven treats included one or more of the terms: limited ingredient (n = 2), hydrolyzed (n = 2), food sensitivities (n = 4), pure (n = 2), or real (n = 1). The median number of undeclared animal protein sources per treat sample was 1.5 (range 0–2), with a total of 11 instances of undeclared DNA found in seven treats. Pig was the most frequently undeclared species (n = 6), with cow the second most common (n = 3). Sheep (n = 1) and bison (n = 1) were found as undeclared species in one treat each. No products were positive for only the declared species. No treat samples were positive for the DNA of cat, dog, goat, horse, mouse, or rat species.
Supplements
Some supplements (9/20; 45%) did not declare any animal-sourced ingredients. Of those, most (6/9; 67%) were negative for the presence of all tested DNA. When present, declared animal-sourced ingredients were variable in specificity and number (range 1–4 sources). Some products declared “chicken” or “beef” whereas others declared “milk,” “egg,” or “liver.” Of the 18 total declared animal sourced ingredients, only 7 specified a unique species (Table 1).
In total, only five supplements (5/20; 25%) declared specific mammalian sources detectable by the PCR panel: two beef only, one beef and poultry, one pig and shellfish, and one pig, shellfish, cheese, and liver. Of those five, two were positive for only the declared species (both cow), and two were negative for all species including those that were declared (one pig and one cow).
In total, eight supplements (8/20; 40%) were found to contain DNA from mammalian species not declared on the ingredient list. The median number of undeclared animal protein sources per supplement was 0 (range 0–2). Cow (n = 5) and pig (n = 5) were equally frequently undeclared and were the only species identified in the supplements. No supplement samples were positive for the DNA of bison, cat, dog, goat, horse, mouse, rat, or sheep species.
Discussion
The present study was the first to use PCR technology to assess for the presence of DNA from mammalian species in canine supplements and in treats with label claims suggesting appropriate use in dogs with AFR. Our hypothesis was partly supported by our findings, in that undeclared mammalian DNA was detected in 7/8 treats and 8/20 supplements. These results cast doubt on the ability to confidently use treats and supplements during elimination trials, regardless of labeling. Feeding an elimination diet exclusively is currently the only accurate diagnostic test for AFR, because laboratory tests performed with blood, saliva, and hair are not reliable.2,16 However, strict compliance to the diet has been identified as a challenge, and the use of treats and supplements is sometimes to blame.2,4 Owners may base purchasing decisions on ingredient lists and label claims. The treat products evaluated in the current study could be perceived as having implied or explicit claims for use in animals with presumptive or definitive AFR. In fact, 5/8 treats, comprising both veterinary and over-the-counter products, carried a label statement of intended use for dogs with food sensitivities. Depending on the exact wording, any of the label terms recorded in the current study could be considered drug claims under the US Food, Drug, and Cosmetic Act, which requires that products intended to diagnose, prevent, or treat diseases undergo a preapproval process by the US FDA. Although dietary supplements for people are regulated according to the Dietary Supplement Health and Education Act passed by Congress in 1994, the FDA has determined that this rule does not apply to products for animal use.17 Thus, veterinary supplements are regulated as food, and there is no requirement for them to be reviewed for effectiveness, safety, purity, or quality before marketing.17 The overarching concern is that the expectations of both pet owners and veterinarians for treat and supplement products may inappropriately rely on ingredient and marketing claims.
The findings of the current study are similar to previous reports documenting the common occurrence of undeclared species found using PCR in various categories of pet foods.5,7,8,11,18 Using species-specific primers for PCR is effective for detection of minute quantities, as low as 0.1%, of DNA in highly processed meat products.19,20 With the PCR method, DNA of target species can be recovered and amplified from a few copies to easily detectable quantities, even from a complex background of highly degraded samples, such as some pet food products. Even though this method helps identify the presence of DNA with high sensitivity, the clinical significance to patients with AFR is unknown. Further research is necessary to characterize antigen quantities that evoke reactions in animals with confirmed AFR. Additionally, studies regarding the clinical effect of feeding treats, supplements, or pet foods with undeclared ingredients are needed.
Although the current study confirmed the presence of undeclared ingredients in many of the canine treats and supplements, it was not possible to discern a reason for their presence or to quantify the amount of material. Cross-contamination is considered more likely than intentional adulteration, given typical manufacturing conditions. Cross-contamination could have occurred at two main levels: (a) during the production of the raw materials or (b) during the processing of the end products, especially if equipment was not thoroughly cleaned between production lines.10 Cross-contamination by the authors during sample preparation and analysis was unlikely in this study because of the strict management of the samples throughout the process. Based on comparison of the sample preparation order and the PCR findings, evidence of cross-contamination was not apparent.
Intentional substitution with alternative ingredients for economic gain is possible; however, this may be less likely given that apparent complete ingredient substitutions in pet products appear to be uncommon. In a study that tested 52 commercial pet products (dry, wet, and treats) using real-time PCR, 20 products were found to be mislabeled, and of those, 7 were negative for one or more declared ingredients, which may indicate omission or substitution for lower cost ingredients.7 Evidence for substitution was not apparent in the current study, with only two instances of a product being negative for the declared species, but both were also negative for all other tested species. Our findings were similar to a previous study that evaluated 21 commercial canine diets using PCR and found only 1 diet negative for the declared species (bison) but also negative for the remainder of the 10 mammalian species analyzed.11
False negative results are possible because of the inherent sample factors such as the presence of compounds that are inhibitory to PCR such as numerous plant-based ingredients, undetectable quantity of DNA, or degraded DNA.21 Methodological factors may also influence the false negative rate, including insufficiently selective primers, suboptimal temperatures, or too many amplification cycles used in the PCR.22 Such occurrences related to the analyses in the current study are less likely, given the implemented quality controls. The Veterinary Genetics Laboratory- Forensics Laboratory is the only US laboratory accredited for domestic animal forensic DNA casework. Techniques and experience gained from examining and testing evidence from the most serious of crimes were applied to the testing of these research samples ensuring the reliability of the test. Although technically under the administration umbrella of the Veterinary Genetics Laboratory–Service Laboratory section of the unit, the Meat ID PCR test is run in the Forensics Laboratory and subject to all of the contamination controls and procedures used there. The Veterinary Genetics Laboratory- Service Laboratory is compliant with ISO 17025 certification requirements23 and has applied for accreditation; assessment visits are pending.
The PCR panel used in the present study includes some species that are closely related; however, this was not expected to result in testing errors and misidentification. The analysis distinguishes among water buffalo (Bubalus bubalis), American bison (Bbison), and beef cattle (B taurus) species. Consequently, cross-reactivity was not considered as a potential cause of the negative result for the sample that declared beef. Additionally, mitochondrial DNA has a high mutation rate (10 times greater than nuclear DNA), allowing for the differentiation between closely related species.24 For this reason, the treat product that was positive for both cow and bison (Table 1) but declared only beef on the label likely represents a true contaminant.
Pig and cow were the most frequent mammalian DNA contaminants in this study. Pork, poultry, and beef are most commonly used in the United States for human consumption, and it is expected that their by-products are frequently used for the manufacturing of pet products.25 The use of gelatin capsules has been reported to influence the presence of ingredients that are not specified on the label.13 In this study, there were two products (2/20; 10%) that included capsules. One declared beef and was positive for cow, and the second was positive for cow but only declared an unspecified source of milk. As milk should not contain cow DNA, it is presumed that the gelatin capsule was the source in this case. In the United States, the principal materials used in gelatin production are pork skin, cattle bones, and cattle hides.26 In any case, supplement products with gelatin capsules should not be recommended during the diagnosis and treatment of canine AFR. If medications only available in capsules are necessary, assuming product stability, owners should be instructed to only use the contents and not the gelatin capsules during elimination trials.
Because the provision of treats is often an important component of the pet–owner bond, alternative options during the diagnosis and treatment of AFR is a frequent topic of discussion between veterinarians and pet owners.27 Owners usually express affection for their animals through the provision of food.28 In one survey study, 96% of 268 interviewed dog owners reported using treats, and a considerable number of these (n = 192) reported feeding commercial treats on a daily basis.29 Although a small number of treats were assessed in the current study, the findings reported here demonstrate that 88% of the treats had at least one undeclared mammalian species, despite labeling terminology such as limited ingredient, hydrolyzed, food sensitivities, pure, or real. Veterinarians should consider using some portion of the main diet or the alternative form of the diet for treats (i.e., if feeding wet, give dry as treats and vice versa). This can also provide a way to give any needed medications without compromising the elimination trial.
Forty percent (8/20) of the supplements assessed in the current study were found to contain DNA from mammalian species not declared on the ingredient list. In 2006, 10% of healthy dog owners reported the use of supplements through a telephone survey performed in the United States and Australia.30 This number was found to be higher in a study performed through a global online survey published in 2020, showing that about 50% of healthy dog owners use dietary supplements.31 However, it is unclear how many of them consulted their veterinarians for the purchase of supplements. The use of these products is certainly a concern during the diagnosis and treatment of AFR. In the majority, their efficacy is not well proven or studied, and their safety and marketing is not well regulated. If the patient is eating a complete and balanced commercial diet, supplements are not recommended unless there is a specific indication to address particular needs. In that case, veterinarians should evaluate the benefit of a specific supplement versus the risk of confounding an elimination trial or treating an AFR. When a home-cooked diet is used, consulting a veterinary nutritionist for the use of appropriate supplements is recommended. A related concern was unspecified declared animal sources. Of the 18 total declared animal sources in the 20 assessed supplements, only 7 (39%) were specific with regard to naming a unique species; the remainder listed ingredients such as poultry, milk, egg, or liver. This lack of information leads to confusion for owners and veterinarians on selecting products and, thus, could result in diagnostic or management failure of canine AFR.
An important limitation of this study is that nonmammalian species such as poultry and fish were not available for testing; these were not uncommonly claimed in the assessed products. In addition, these species have previously been identified as undeclared components in pet foods. One study detected by microarray analysis the presence of 19 animal species (including livestock, poultry, and fish) in 40 pet foods used for elimination trials and available in Europe; pork, chicken, and turkey were the most frequently identified undeclared species.9 In contrast, other studies have found undeclared mammalian DNA to be more common despite also including analysis for DNA of specific poultry species. For example, one study that used real-time PCR to test 52 pet foods and treats for eight different species (bovine, caprine, ovine, chicken, goose, turkey, porcine, and equine) found both declared and undeclared chicken DNA in all but one sample.7 However, beef was the most common undeclared species.7 Similarly, another study that tested 12 commercial canine elimination diets for five different species (chicken, turkey, beef, mutton, and pork) found chicken DNA in four products (two declared and two undeclared).18 However, DNA from undeclared beef was identified in eight products, six of which were also positive for undeclared pork.18 It is unknown if testing nonmammalian species would increase the rate of contaminants found in the treats and supplements assessed in the present study, and further investigation is warranted. Another limitation is that only one sample of each product could be assessed, and a limited number of treats and supplements were included. Ideally, the range and number of products would be expanded, and different batches of each treat and supplement could be analyzed for repeatability and to increase the predictive value through longitudinal study. Furthermore, the use of complementary techniques to enable quantification of any undeclared material such as quantitative ELISA or real-time PCR would be ideal. However, in the absence of knowledge regarding the amount of allergen needed to evoke AFR, quantification may be more important in studies aiming to distinguish intentional adulteration from cross-contamination.
Conclusion
In this study of 8 treats and 20 supplements labeled for dogs, a significant percentage of products contained undeclared mammalian DNA. Although none of the products assessed in this study specified an indication for use in dogs with suspected or confirmed AFR, treats were selected because they included verbiage that may imply appropriateness for this use (limited ingredient, hydrolyzed, food sensitivities, pure, or real). Based on our findings and the potential to confuse interpretation of animal response to elimination diet trials, the use of commercial treats and supplements should be carefully considered during the diagnosis and treatment of canine AFR. Veterinarians should discuss with dog owners the potential confusion caused by feeding these products during an elimination trial. Efforts to reduce manufacturing cross-contamination are necessary to ensure accurate ingredient declarations, because many pet owners and veterinarians purchase or recommend products based on label information. Additionally, testing for possible contamination as part of good quality control practices should be considered for products intended for use in pets with AFR.
Contributor Notes
