J Vet Intern Med 2001;15:7–13
Food Sensitivity in Cats with Chronic Idiopathic Gastrointestinal Problems W. Grant Guilford, Boyd R. Jones, Peter J. Markwell, Donald G. Arthur, Mark G. Collett, and John G. Harte The objectives of this study were to investigate the prevalence of food sensitivity in cats with chronic idiopathic gastrointestinal problems, to identify the food ingredients responsible, and to characterize the clinical features. Seventy cats that presented for chronic gastrointestinal signs underwent diagnostic investigation. Fifty-five cats had idiopathic problems and were entered into the study. Diagnosis of food sensitivity was made by dietary elimination-challenge studies by using commercial selected-protein diets as the elimination diet. Sixteen (29%) of the 55 cats with chronic idiopathic gastrointestinal problems were diagnosed as food sensitive. The clinical signs of another 11 cats (20%) resolved on the elimination diet but did not recur after challenge with their previous diet. The foods or food ingredients responsible for the clinical signs were dietary staples. Fifty percent of affected cats were sensitive to more than 1 food ingredient. The clinical feature most suggestive of food sensitivity was concurrent occurrence of gastrointestinal and dermatological signs. Weight loss occurred in 11 of the affected cats, and large-bowel diarrhea was more common than small-bowel diarrhea. Assay of serum antigen-specific immunoglobulin E (IgE) had limited value as a screening test, and gastroscopic food sensitivity testing was not helpful. In conclusion, adverse reactions to dietary staples were common in this population of cats, and they responded well to selected-protein diets. Diagnosis requires dietary elimination-challenge trials and cannot be made on the basis of clinical signs, routine clinicopathological data, serum antigen-specific IgE assay, gastroscopic food sensitivity testing, or gastrointestinal biopsy. Key words: Diagnosis; Diarrhea; Food allergy; Vomiting.
A
dverse reactions to food (food sensitivities) include those mediated by the immune system (food allergies) and those without an immunological basis (food intolerances).1 Clinical signs attributed to food sensitivity by veterinarians usually are dermatological or gastrointestinal. The prevalence of chronic dermatological abnormalities resulting from food sensitivity in cats has been estimated to be 5.8% in a university practice.2 Food sensitivity is thought to be the second most common cause of allergic dermatitis in cats and is considered responsible for up to 11% of cats with miliary dermatitis.3 In contrast, the frequency with which chronic gastrointestinal complaints in cats are caused by food sensitivity is unknown.4 A number of observations suggest that the prevalence of food sensitivity in cats with gastrointestinal problems may be higher than in cats with skin problems. For example, evidence from other species suggests that not only can food sensitivity produce gastrointestinal problems, but conversely that gastrointestinal diseases can lead to food sensitivity by compromising oral tolerance.5 Furthermore, food intolerances from such disorders as brush border enzyme abnormalities and ingested toxins would more commonly affect the bowel and result in gastrointestinal clinical signs rather than dermatological signs. From the Institute of Veterinary, Animal, and Biomedical Sciences, Massey University, Palmerston North, New Zealand (Guilford, Jones, Collett); Waltham Centre for Pet Nutrition, Waltham-on-the-Wolds, UK (Markwell, Harte); and the Ministry of Agriculture and Forestry, Quality Management, P.O. Box 24, Lincoln, New Zealand (Arthur). Presented in part at the 14th American College of Veterinary Internal Medicine Forum, 1996. Reprint requests: Professor W. Grant Guilford, BVSc, PhD, Dipl ACVIM, Institute of Veterinary, Animal, and Biomedical Sciences, Massey University, Private Bag 11-222, Palmerston North, New Zealand; e-mail: W.G.Guilford@massey.ac.nz. Submitted December 2, 1999; Revised May 17, 2000; Accepted August 2, 2000. Copyright q 2001 by the American College of Veterinary Internal Medicine 0891-6640/01/1501-0001/$3.00/0
It is uncertain to which food ingredients cats with gastrointestinal problems are most frequently sensitive. Limited information from studies of cats and more compelling evidence from other species suggest that the most prevalent food allergens are proteins commonly included in the diet.1,5 In contrast, the food ingredients most likely to be responsible for food intolerances are more varied and do not require prior exposure. These include disaccharides (such as lactose), food additives (such as coloring agents), pharmacologically active products (such as histamine), and food toxins.1,5,6 The gastrointestinal signs of food sensitivity in cats are poorly described. Vomiting and diarrhea usually are reported,3,5,7,8 but the specific characteristics of the pattern of vomiting and type of diarrhea are rarely mentioned. Furthermore, the presence or absence of particular clinical features suggestive of gastrointestinal food sensitivity have not been methodically investigated. Adverse reactions to food usually are suspected when an association is made between the ingestion of a certain food and the appearance of a clinical sign. The diagnosis is confirmed by dietary elimination-challenge studies. Alternative methods of diagnosis have been proposed in other species, including assay of serum antigen-specific immunoglobulin E (IgE) and gastroscopic food sensitivity testing (GFST).5,9–11 A commercial assay of cat antigen-specific IgE in serum now is available.a A technique to perform GFST recently has been developed for use in dogs11 and awaits application in cats. The primary advantage of serum antigen-specific IgE assay over elimination-challenge trials is convenience. The principal advantage of GFST testing is that the response of the gastrointestinal mucosa to several food allergens can be directly and simultaneously observed. An important disadvantage of serum antigen-specific IgE assay (and most likely GFST) is that these tests focus on the diagnosis of just 1 type of food sensitivity—immediate (type 1) hypersensitivity. Therefore, the effectiveness of these tests as screening tests for the entire range of adverse reactions to foods requires investigation.
8
Guilford et al
The principal objective of this study was to investigate the prevalence of adverse reactions to foods in cats with chronic idiopathic gastrointestinal problems and to identify the foods or food ingredients to which the cats were sensitive. In addition, we aimed to characterize the clinical features of gastrointestinal food sensitivity in cats and to evaluate the concordance of serum antigen-specific IgE assay and GFST with the elimination-challenge tests by which the diagnosis was made.
Materials and Methods Cats Seventy cats that were presented to the Massey University Veterinary Teaching Hospital (MUVTH) with chronic gastrointestinal signs underwent thorough diagnostic investigation (see below) to determine their suitability for entry into the study. Thirty-seven cats were female, and 33 cats were male. All of the cats had diarrhea or vomiting for more than 2 weeks (36 cats with vomiting, 22 cats with diarrhea, and 12 cats with both vomiting and diarrhea). The cats showed a variety of other clinical signs, including weight loss (40 cats), anorexia (22 cats), flatulence (18 cats), and abdominal bloating (8 cats). Of the 70 cats, 15 were diagnosed with renal disease, liver disease, hyperthyroidism, parasitism, infection with feline leukemia virus (FeLV) or feline immunodeficiency virus (FIV), obstruction of the gastrointestinal tract, neoplasia, or infectious gastroenteritis and were eliminated from the study at various stages of the diagnostic workup. Fifty-five cats completed workups without identification of a specific diagnosis to explain their clinical signs. These cats were diagnosed as having ‘‘chronic idiopathic gastrointestinal problems’’ and were entered into the study. The gastrointestinal biopsy specimens of the majority of these 55 cats were subjectively assessed to have mild to moderate increases in the number of lymphocytes, plasma cells, or eosinophils in the lamina propria (ie, inflammatory bowel disease), but other cats had no evidence of increased numbers of mucosal inflammatory cells.
Gastrointestinal Workup Standardized patient history forms were completed to determine the clinical signs, the staple diet, the temporal relationship of the clinical signs to food intake (if any), and the time between bouts (if the clinical signs were episodic). A complete physical examination was performed and the results recorded on standardized physical examination forms to ensure consistency in examination. A database was collected consisting of complete blood count, serum chemistry profile, FeLV and FIV test, urinalysis, and at least 2 zinc and sugar fecal flotations for Giardia and nematode parasites, respectively. A radiopaque markerb study was performed on all vomiting cats to rule out partial obstructions of the bowel and to evaluate gastrointestinal motility. Total serum thyroxine concentration was measured in all cats older than 6 years. Cats with diarrhea additionally underwent rectal scraping for cytology and acid-fast stain of the feces to detect Cryptosporidium spp. All cats underwent gastroduodenoscopy, during which biopsy samples were obtained from the stomach and duodenum for histopathology and duodenal fluid was aspirated for quantitative aerobic and anaerobic culture. In addition, incisional biopsy was obtained from the rectal mucosa of all cats, and a colonoscopy and biopsy were performed on those cats with a history of blood or mucus in their feces. All endoscopic biopsy samples were pinch biopsy specimens, and 8 to 12 biopsy samples were collected from each region of the gastrointestinal tract examined. The biopsy specimens were viewed by the pathologist on duty at the time of the diagnostic workup. All biopsy samples subsequently were pooled and reviewed by a pathologist (D.G.A.) with a subjective grading system based on the number and type of inflammatory cells in the lamina propria. The degree of mucosal inflammation was classified as absent, mild, moderate, or severe. The patholo-
Table 1. Food additive and histamine solutions used in GFST. Solution No. 1 2 3 4 5
Additive
Percentage
Sodium nitrite Red oxide Titanium oxide Guar gum ‘‘Petset’’b Liquid caramel Histamine
0.01 0.1 0.2 0.1 0.6 0.2 0.01
Mixture 10 mg in 100 mL basea 100 mg in 100 mL base 200 mg in 100 mL base 100 mg in 100 mL base 600 mg in 100 mL base 200 mg in 100 mL base 27.5 mg histamine phosphate in 100 mL base
GFST, gastroscopic food sensitivity testing. a Base 5 sterile saline with 0.2% phenol. b Petset 5 carageenan : locus beangum : potassium chloride 40 : 40 : 20 (also called carob gum). gist was unaware of whether the cats had been diagnosed as food sensitive.
Serum IgE Assay Serum samples were taken at admission from all cats and stored at 2208C for a maximum of 6 months before analysis by antigen-specific enzyme-linked immunoadsorbent assay by a commercial laboratory.a Cats with food-specific IgE concentrations greater than 200% of control sera were considered positive.
Gastroscopic Food Sensitivity Tests All cats underwent GFST according to a previously described technique.11 The staple diets of the cats were withheld for at least 3 days before the GFST in the hope of increasing the sensitivity of the test.11 The procedure was performed after induction of anesthesia with tiletamine-zolazepamc and maintenance with halothane. All cats were tested with 0.5-mL volumes of 5 diluted food protein extracts,d a dilute histamine solution, and 4 solutions of food additives (Table 1). Four of the food proteins (milk, beef, lamb, and chicken) were tested in all cats. Milk, beef, and lamb proteins are dietary staples for most New Zealand cats. Chicken was used as the negative control protein because it is infrequently fed as a dietary staple in New Zealand. The 5th food protein was varied according to the dietary history of the individual cat under study. Most commonly, this protein was wheat, corn, or a mixed-fish extract.d The food extracts were applied at a final concentration of 15,000 protein nitrogen units (PNUs) per milliliter with the exception of wheat and corn extracts, which were applied at concentrations of 5,000 PNU/mL. The additives tested were those included frequently in commercial cat foods in New Zealand. The additives were donated by a local pet food manufacturer and were diluted in saline to obtain the concentrations used by this manufacturer in their products. All cats were tested with the same additives. Some of the additives were tested in combination to minimize the anesthesia time required to test the additives.
Dietary Elimination-Challenge Trial to Diagnose Food Sensitivity After the diagnostic procedures were completed, the cats were fed sufficient food to satisfy their appetites with a commercially available chicken- or venison-based selected-protein diet e,f for a minimum of 4 weeks. The choice between these diets was based on the dietary history of the individual cat. Most cats were placed on the chicken-based diet, but any cats that had eaten chicken within the previous 6 months were fed the venison-based diet. The owners were given the option of
Food Sensitivity in Cats taking the cat home or leaving it in the hospital for this phase of the study. The majority of owners (90%) took their cats home. Owners were warned of the importance of feeding the elimination diet exclusively. If the owner had more than 1 cat, the owner was instructed to feed all of the cats the elimination diet. If the cat had access to the outdoors, the owners were asked to visit their neighbors to ask them not to feed the cat. During the elimination trial, the owner was asked to record on a daily basis the cat’s food intake and clinical signs on standardized logs. The clinical signs recorded by the owners included the number of times the cat vomited and the grade of the feces as compared with a photographic fecal grading chart. The 5-point chart graded the fecal consistency from firm stools (grade 5) to liquid diarrhea (grade 1). Those cats with clinical signs that did not become reduced in frequency or severity after 4 weeks were considered not to be suffering from food sensitivity. Cats with clinical signs that completely resolved on the elimination diet were returned to their staple diet for 4 to 7 days. If more than 1 food had formed part of a cat’s staple diet, the cat was challenged with each of these foods for a 4-day period followed by a 3-day washout period, during which the cat was fed the elimination diet before the next diet challenge. The 4-day challenge period was considered the minimum desirable but was chosen to allow a new food to be tested on a weekly basis. The washout period minimized the chance of misdiagnosis caused by a delayed food sensitivity reaction appearing during the subsequent week of testing. These challenge tests were performed either at home or in the hospital, depending on the owner’s preference. A cat was considered to have recrudescence of its clinical signs if fecal consistency deteriorated to the grade assigned before the elimination trial or if the vomiting increased to a frequency similar to that recorded before the elimination trial. Cats were again fed the elimination diet for 2 to 4 weeks until signs resolved. Cats were diagnosed as food-sensitive if their clinical signs resolved when they were fed the elimination diet, recrudesced when they were challenged with their staple diet, and then resolved for a 2nd time when they were returned to the elimination diet.
Dietary Elimination-Challenge Trials to Identify the Responsible Food Ingredient The owners of 12 of the cats diagnosed as food-sensitive agreed to admit their cats to the MUVTH for more detailed dietary challenge studies, with the aim of identifying the specific food ingredient to which the cat was sensitive. Dietary challenge studies were undertaken by adding a small quantity of food protein (minimum 0.6 grams of protein per kilogram body weight) or food additives to the elimination diet on a weekly basis. The food proteins chosen were those that formed the majority of the cat’s diet in the previous 3 months. Most cats were tested with lamb and beef (50 g diced fresh meat per day), wheat (25 g cooked cereal per day), canned viscera (50 g of sheep and beef lung, testicles, kidney, and liver minced on an equal weight basis), and milk (15 g of whole milk powder in a slurry). Up to 2 other protein sources were tested, depending on the owner’s willingness to accept a prolongation of the hospitalization period. The other protein sources (dependent on diet history) included tuna (50 g), whitefish (50 g), pork (50 g), corn gluten (15 g), oats or barley (25 g), or egg (1 whole egg cooked and diced into diet). The food additives chosen were the same as those used in the GFST. The additives were mixed in combination into the elimination diet at concentrations commonly found in New Zealand commercial pet foods (Table 2). If an adverse reaction to the combination of additives was observed, cats then underwent a series of challenges with the individual additives added to the elimination diet. For all challenge trials with food ingredients, a 4-day challenge period and 3-day washout period were used. Both the food proteins and the food additives were tested in a random order. Clinical signs evaluated during in-hospital challenge trials were the same as those evaluated by the owners (vomiting frequency and
9
Table 2. Food additive and histamine doses used in dietary challenge studies. Additive
Dose
Sodium nitrite Red oxide Titanium oxide Guar gum ‘‘Petset’’a Liquid caramel Histamine
12 300 600 300 1,800 600 20
mg mg mg mg mg mg mg
a Petset Carageenan : locus beangum : potassium chloride 40 : 40 : 20 (also called carob gum).
fecal grade) but with the addition of defecation frequency (which could not be assessed reliably in nonhospitalized indoor-outdoor cats).
Results Sixteen (29%) of the 55 cats with chronic idiopathic gastrointestinal problems were diagnosed as food sensitive. In addition to the 16 food-sensitive cats, the clinical signs of another 11 cats (20%) entered into the study resolved on the selected-protein diet but did not recrudesce on challenge with their old diet. The clinical signs of the food-sensitive cats resolved quickly on the elimination diet. Vomiting stopped immediately in almost all affected cats, and diarrhea resolved in most affected cats within 2 or 3 days. Recrudescence of clinical signs in the food-sensitive cats was rapid also, occurring within 3 to 4 days of challenge with their previous diet. Commercial canned food comprised some or all of the diet of 93% of the food-sensitive cats and 94% of the non– food-sensitive cats. Commercial dry diets were fed to 60% of the food-sensitive cats and 54% of the non–food-sensitive cats. Six percent of the food-sensitive cats and none of the non–food-sensitive cats were fed exclusively dry diets. Table foods occasionally were fed to 73% of food-sensitive cats and 67% of non–food-sensitive cats. The foods or food ingredients to which the affected cats were sensitive are listed in Table 3. The most common allergens were beef, wheat, and corn gluten. One of the cats with wheat sensitivity was determined to have a transient sensitivity to wheat gluten, which resolved after approximately 6 weeks of dietary management with a selectedprotein diet. One cat vomited immediately after ingestion of the mixture of food additives but demonstrated no adverse reactions to any of the food additives when they were fed separately. The cat was rechallenged with the mixture of additives and immediately vomited a 2nd time. Of the 8 food-sensitive cats that underwent elimination-challenge testing with multiple food ingredients, 4 (50%) were found to be sensitive to more than 1 food ingredient. Nine of the food-sensitive cats were female, and 7 were male. Their ages ranged from 6 months to 14 years (median, 5 years), with only 3 cats being less than 1 year. The majority of cats were domestic short hairs. Their gender, age, and breed distributions were similar to the non–foodsensitive cats in the study. Nine of the food-sensitive cats had a history of vomiting (56%), 4 had a history of diarrhea (25%), and 3 had a
10
Guilford et al
Table 3. Foods or food ingredients responsible for food sensitivity.a Food or Ingredient
No. Cats
Commercial dry diet Commercial canned diet Beef Corn gluten Wheat Wheat gluten Barley Chicken Lamb Sardines Viscera Lactose Food additives
5b 1b 3 3 3 1 1 1 1 1 1 1 1
Fifty percent of the cats were multiply allergic. The allergen in the dry or canned foods fed to these cats was not identified either because the owners refused further testing or because further elimination-challenge testing failed to identify the allergen. a
b
history of both vomiting and diarrhea (19%). The vomiting was relatively infrequent, occurring less than once per day in most cats and up to a maximum of 3 times per day in the most severely affected cat. The nature of the vomitus and its timing after eating was variable. Some cats vomited food within minutes of eating, whereas others vomited food 2 hours to more than 12 hours after eating. In some cats, the timing of the vomiting was seemingly unrelated to eating, and the vomitus consisted primarily of bile. Four of the 7 food-sensitive cats (57%) with diarrhea had evidence of large-bowel dysfunction (mucus and fresh blood in the feces, excessive straining to defecate, or both). The frequency of vomiting (63%), diarrhea (34%), and concurrent vomiting and diarrhea (13%) in the non–foodsensitive cats was similar to that of the food-sensitive cats. Fifty-six percent of non–food-sensitive cats with diarrhea had evidence of large-bowel dysfunction. Weight loss and flatulence were observed in 11 (69%) and 6 (38%) of the food-sensitive cats and 53% and 21%, respectively, of the non–food-sensitive cats. The weight loss was classified as moderate to marked in 50% of the food-sensitive cats. Appetite was variably reported as decreased, unchanged, or increased in food-sensitive cats. Some of the food-sensitive cats with weight loss had moderate or severe histologic abnormalities (20%) of the small intestinal mucosa, but most had mild histologic abnormalities (40%) or no histologic abnormalities (40%). The demeanor of 6 (38%) of the food-sensitive cats was reported as irritable. Another 4 cats (25%) were reported as lethargic. Collectively, therefore, altered demeanor was reported in 10 (63%) food-sensitive cats. Altered demeanor was noted in 45% of the non–food-sensitive cats. Concurrent dermatological disease was observed in 4 cats (25%) with gastrointestinal problems caused by food sensitivity. These cats suffered from miliary dermatitis, pruritus, and alopecia. The dermatological signs were caused by food sensitivity in 3 of these 4 cats. In contrast, concurrent dermatological signs were reported in 15% of the non–food-sensitive cats. Of a total of 10 cats presented with
a combination of dermatological and gastrointestinal disease, 4 of these cats (40%) were diagnosed as food sensitive. The duration of clinical signs in the food-sensitive cats ranged from 1 to 70 months (median, 7.5 months). The clinical signs were episodic in 4 (25%) of the cats, with an interval free of clinical signs for up to 3 months. A similar percentage of non–food-sensitive cats had episodic clinical signs, but in these cats longer intervals free of clinical signs were reported (as long as 1 year). No consistent abnormalities were detected in the CBC and serum chemistry profiles of the food-sensitive cats. However, 29% of the food-sensitive cats had eosinophilia. The eosinophil counts ranged from 0.07 to 6.1 3 109 per liter (median, 0.55 3 109 per liter). A smaller percentage (10%) of non–food-sensitive cats had eosinophilia. In these cats, the eosinophil counts ranged from 0.13 to 5.4 3 109 per liter (median, 0.76 3 109 per liter). Fourteen percent of the food-sensitive cats had lymphopenia. The lymphocyte counts ranged from 1.16 to 7.0 3 109 per liter (median, 3.12 3 109 per liter). Twenty-four percent of the non–foodsensitive cats had lymphopenia. In these cats, the lymphocyte counts ranged from 0.57 to 10.68 3 109 per liter (median, 2.59 3 109 per liter). Serum antigen-specific IgE test results were available for 12 of the food-sensitive cats. Of these 12 cats, 7 cats (58%) had one or more positive test results. Two cats with adverse reactions to corn gluten and 1 cat with an adverse reaction to dry food containing corn had positive serum tests for corn-specific IgE. Two food-sensitive cats (17%) had positive serum IgE tests to antigens to which they showed no adverse reactions. Three food-sensitive cats (25%) had positive serum IgE tests to antigens to which they were not tested but to which they were unlikely to be sensitive because the antigens were novel proteins to which these cats were unlikely to have been exposed. The remaining 5 foodsensitive cats had no positive serum antigen-specific IgE tests. Serum antigen-specific IgE tests were available for 24 of the non–food-sensitive cats. Of these 24 non–food-sensitive cats, 6 (25%) had 1 or more positive tests. Eighty-three percent of these positive results were to soy (3 cats), rice (4 cats), and corn (3 cats) or some combination of these 3 foods. One cat had a positive test to peanut and another to potato. The labels of the staple diets of the non–food-sensitive cats with positive serum antigen-specific IgE tests did not list rice, soy, peanut, or potato. However, transient exposure to these food antigens could not be discounted because all of these cats had very varied diets, including table foods and canned and dry foods purchased from supermarkets and veterinary practices. Gastroscopic food sensitivity testing was not helpful. Only one suspicious reaction (swelling without erythema) was observed in all of the procedures performed. This reaction occurred after the application of a gum mixture (solution 3; Table 1) and was not confirmed by positive dietary challenge to gums. All of the food-sensitive cats had histological changes in at least 1 region of the bowel. However, the region of the gastrointestinal tract affected was not consistent, and the histological changes were nonspecific and were similar in
Food Sensitivity in Cats
frequency and severity in food-sensitive and non–food-sensitive cats. One food-sensitive cat had a subacute diffuse suppurative glossitis with focal ulceration. Gastric mucosal biopsy specimens were abnormal in two thirds of the foodsensitive cats. One cat had severe lymphocytic gastroenteropathy characterized by diffuse infiltration of lymphocytes throughout the gastric lamina propria along with clumps of plasma cells. In some areas, invasion and obliteration of glands was observed. Nine other food-sensitive cats had mild histological changes in their gastric mucosa, including small numbers of eosinophils migrating through the glandular epithelium, scattered foci of lymphocytes, patchy subepithelial edema or hemorrhage, or mild fibroplasia. Lymphoid nodules were observed in the gastric mucosa of 2 food-sensitive cats. The duodenal mucosal biopsy specimens were abnormal in 50% of the food-sensitive cats. On most occasions, the pathologic diagnosis was mild lymphocytic-plasmacytic enteritis. Two food-sensitive cats had moderately severe lymphocytic-plasmacytic enteritis, and 2 had moderate to severe eosinophilic enteritis. In the most severely affected cat with eosinophilic enteritis, the villi in all sections were blunt and often showed fusion. At the tips of the villi, epithelial cells often were vacuolated. Some of the duodenal biopsy specimens of the food-sensitive cats had a variety of other features. These included small numbers of neutrophils, occasional globular leukocytes, evidence of edema, and foci of hemorrhage, mild fibrosis, or hyperplasia of glands in the basal mucosa. The colonic or rectal mucosal biopsy specimens were abnormal in two thirds of the food-sensitive cats. The pathological diagnosis was usually mild colitis or proctitis. On most occasions, the findings included mild lymphocytic-plasmacytic infiltration. Attenuation of the epithelium occasionally was reported, as was the presence of small numbers of neutrophils, eosinophils or globular leukocytes, mild fibrosis, or mucus distension of crypts.
Discussion The results of this study suggest that almost one third of cats presented to a referral hospital with chronic idiopathic gastrointestinal problems have food sensitivity. It is important to emphasize that this high prevalence of food sensitivity was observed in a highly selected population of cats that had undergone extensive diagnostic effort to rule out other common causes of gastrointestinal dysfunction. The population studied included many cats that satisfied the current definition of inflammatory bowel disease and other cats that did not satisfy this definition, either because they were diagnosed as food sensitive or because they were considered to have normal numbers of mucosal inflammatory cells.12 In the present study, the diagnosis of food sensitivity was based on dietary elimination-challenge trials. Of note was the quick resolution of clinical signs in the food-sensitive cats during the dietary elimination trials. These results, together with the high prevalence of food sensitivity detected in the present study, suggest that the duration of the elimination diet in cats with gastrointestinal disease need not exceed 4 days. In contrast, veterinary dermatologists often recommend elimination diets for a minimum duration of 8
11
weeks to diagnose the dermatological manifestations of food sensitivity.3,13,14 Rapid recrudescence of signs after rechallenge of food-sensitive cats with the responsible allergen was seen in the present study. In studies of cats with dermatological problems, most cats recrudesce within 3 to 5 days, but some took up to 10 days.3,15 This observation raises the possibility that a longer period of rechallenge in the present study may have increased the number of diagnoses of food sensitivity. Countering this possibility, however, is that all of the food-sensitive cats in the present study first showed recrudescence within 3 days, and none required the full 4 to 7–day challenge period allotted. The reasons for these apparent differences in behavior between skin and gastrointestinal food sensitivities are unknown. In addition to the food-sensitive cats, the clinical signs of another 20% of cats entering the study resolved on the selected-protein diet but did not recrudesce on challenge with their staple diets. If the diagnosis of food sensitivity had not depended on observing recrudescence of signs after rechallenge with their staple diets, the number of cats diagnosed as food sensitive would have almost doubled in the present study. It is unclear why the clinical signs of so many cats resolved on the selected-protein diet and then did not recrudesce after rechallenge. The dietary histories of some of these cats may have been incomplete, and they may not have been challenged with the food to which they were sensitive. Alternatively, some of the cats may have spontaneously recovered from their gastrointestinal disorder, and the recovery of others may have been facilitated by other therapeutic aspects of the selected-protein diet, such as high digestibility. Collectively, 50% of the cats fed the selected-protein diets had resolution of their clinical signs. This observation suggests that selected-protein diets should be considered an important part of the management of cats with idiopathic gastrointestinal problems. The foods or food ingredients responsible for the clinical signs usually were dietary staples of the affected cats. Beef, wheat, and corn were common allergens. Beef also was a common allergen in another recent study of food-sensitive cats3 and frequently is incriminated in dogs.14 In other studies of food-sensitive cats, the foods most often incriminated were fish and dairy products.8,16 To our knowledge, the adverse reactions to gluten, viscera, and the mixture of additives observed in the present study are the first documented reports of these food sensitivities in cats. The high percentage of cats in the present study that were sensitive to more than 1 food ingredient contrasts with reports of animals with dermatological problems, in which reactions to multiple allergens are considered unusual.14 No clinical features allowed differentiation of food-sensitive from non–food-sensitive cats without the use of the elimination-challenge trial. The signalment was diverse, the vomiting pattern not helpful, and the characteristics of the diarrhea were variable. Of note was that a history of largebowel diarrhea was more common than a history of smallbowel diarrhea in the food-sensitive cats. Lymphocyticplasmacytic colitis in cats previously has been observed to be food responsive,17 and large-bowel diarrhea also has been described as an important feature of dogs with food sensitivity.18 This predilection for large-bowel disease contrasts with the typical clinical picture in gluten enteropathy
12
Guilford et al
of humans, in which the upper small intestine is most severely affected and suggests that the concentration of antigen is not the primary determinant of the severity of the mucosal histological changes in cat food sensitivity. The high frequency of weight loss in the food-sensitive cats was unexpected and without clear explanation in some of the cats. The appetite of some of the cats with weight loss was reduced, but most ate well. The small-bowel diarrhea and flatulence observed in some cats indicated malabsorption, but most cats with weight loss did not have these signs, and many had few histological abnormalities in their small intestinal mucosa. Weight loss is also a prominent clinical feature of mild gluten enteropathy in Irish Setters.19 Enteric protein loss has been recognized during intestinal anaphylaxis to foods in laboratory animals.5 Chronic enteric protein loss is energetically demanding and may explain the weight loss observed in the food-sensitive cats and dogs. The clinical feature most suggestive of food sensitivity was the concurrent occurrence of gastrointestinal and dermatological signs. Others also have noted an association between dermatological and gastroenteric diseases in cats with food sensitivity. Twenty-nine percent of the food-sensitive cats in a recent study had both pruritus and gastrointestinal signs.3 The concomitant appearance of cutaneous and gastrointestinal signs in food-sensitive patients also was noted by Walton.16 A high frequency of simultaneous cutaneous and gastrointestinal signs has been reported in one study of dogs with food sensitivity,18 but other studies have not described this association.14 White13 has pointed out that this discrepancy may be partially caused by failure of the owner to detect or report the gastrointestinal signs and also emphasizes the importance of a thorough medical history with respect to both gastrointestinal and dermatological signs. The concurrent appearance of cutaneous and gastrointestinal signs is not pathognomic, however, for food sensitivity. In the present study, some non–food-sensitive cats had gastrointestinal disease in association with flea allergic or idiopathic pruritus. The routine laboratory work did not allow differentiation of food-sensitive and non–food-sensitive cats. Peripheral eosinophilia was present in fewer than one third of the food-sensitive cats and also was observed in non–food-sensitive cats. The diagnostic value of in vitro tests for food-specific IgE antibodies varies widely in studies of humans10 and has been disappointing in dogs.14 Several observations in the present study call into question the value of the serum antigen-specific IgE assay as a screening test for food sensitivity in cats. These include the observations that as many as 25% of the non–food-sensitive cats had positive tests, only 58% of the food-sensitive cats had positive tests, and only 25% of the food-sensitive cats showed concordance between a positive antigen-specific IgE and a positive oral challenge test. The lack of value of antigen-specific IgE as a screening test for gastrointestinal food sensitivity is not surprising. It is unreasonable to expect measurement of serum antigen-specific IgE concentration to be effective for this purpose when there is strong evidence that many gastrointestinal food sensitivities are food intolerances that are not mediated by IgE.5,20,21 It is perhaps more appropriate for clinicians to use serum antigen-specific IgE tests to assist
them in determining whether animals suffering from adverse reactions to foods are affected by type I hypersensitivity reactions. The results of the present study suggest that type I hypersensitivities account for only 25% of gastrointestinal food sensitivities in cats. Caution must be exercised, however, in interpreting the serum IgE results of the present study because, to the authors’ knowledge, validation of this commercial assay has not been reported. An earlier publication, comparing the results of intradermal skin tests and serum antigen-specific IgE in cats showed a poor correlation.22 Other than the observation that the positive serum antigen-specific IgE tests were twice as common in the food-sensitive cats than the non–food-sensitive cats, there was little evidence to support the diagnostic accuracy of the assay. The observation that 17% of the food-sensitive cats and 25% of the non–foodsensitive cats had positive serum IgE tests to antigens to which they showed no clinically evident adverse reactions indicates the presence of subclinical food sensitivity, asymptomatic sensitization, or false-positive test results.10 Gastroscopic food sensitivity testing has been successfully used to detect food sensitivity in dogs11,23 and humans9 but was not helpful in detecting food sensitivity in cats in this study. The reason for this apparent species difference is unclear. Possibilities include a difference in gastric mucosal permeability to macromolecules, differences in the behavior of mucosal mast cells, or a difference in the pathogenesis of food sensitivity among species. The modification of GFST proposed by Ermel et al,24 in which the food allergen is injected into the mucosa rather than dripped onto it, may provide better results in cats. Unfortunately, however, in one author’s experience this method in dogs creates difficulties in separating the acute phase reaction to mucosal injury from that induced by hypersensitivity (Guilford, personal communication). No histological features distinguished food-sensitive from non–food-sensitive cats in the present study. The stomach and rectum were the gastrointestinal sites that were most frequently considered abnormal in the food-sensitive cats but, as with humans and laboratory animals,20,25–27 all levels of the gastrointestinal tract can be affected by food sensitivity. Eosinophilic infiltrates were observed in the mucosa of some food-sensitive cats, but the majority of cats did not have excessive mucosal eosinophilic inflammation. The severity of the mucosal changes was also not helpful in differentiating food-sensitive from non–food-sensitive cats. Some food-sensitive cats had no detectable mucosal abnormalities, and others had severe changes. Idiopathic chronic gastritis, lymphocytic-plasmacytic enteritis, and chronic colitis currently cannot be differentiated from food sensitivity on the basis of histological examination of a single set of biopsy specimens. At times, the lymphocytic infiltrate in the intestinal mucosa of food-sensitive cats can be so intense as to mimic lymphosarcoma.28 The diverse histological findings in the gastrointestinal tract of food-sensitive cats observed in the present study have been described previously. Eosinophils have been found in rectal scrapings from food-sensitive cats.3 Lymphocytic-plasmacytic colitis has been attributed previously to food sensitivity in cats.17 The small intestine of a cat allergic to milk showed congestion, edema, villous degen-
Food Sensitivity in Cats
eration, hemorrhage, and an increase in the number of plasma cells after 4 days of milk challenge.29 The acute phase reactions observed in the stomachs of food-sensitive dogs are characterized by congestion and edema.24 In food-sensitive humans and laboratory animals, the mucosa can be microscopically normal or can show mucosal edema, separation of the epithelium from the lamina propria, fat accumulation in the epithelium, sloughing of villus tip epithelium, increased mucus secretion, increased intraepithelial lymphocytes, infiltration of the lamina propria with lymphocytes and plasma cells, and villus atrophy.20,25–27,30,31 The histological features of food protein–induced enteropathies usually are patchy and the villus atrophy partial rather than complete.26 In conclusion, adverse reactions to dietary staples are common in cats with chronic gastrointestinal problems and can be successfully managed by feeding selected-protein diets. Diagnosis requires dietary elimination-challenge trials and cannot be made on the basis of clinical signs, routine laboratory work, serum antigen-specific IgE assay, gastroscopic food sensitivity testing, or histological examination of a single set of gastrointestinal biopsy specimens.
Footnotes Bio-Medical Services, Austin, TX BIPS, Med ID, Grand Rapids, MI c Zoletil, TechVet, Auckland, New Zealand d Greer Laboratories, Lenoir, NC e WHISKAS Feline Selected Protein Diet (Waltham), Masterfoods, Bruck, Austria (chicken and rice) f WALTHAM Veterinary Diet Feline Selected Protein Diet, Effem Foods, Bolton, Canada (venison and rice) a
b
Acknowledgments The authors would like to thank the many veterinarians who referred cats for the study, in particular Drs Pru Galloway and Stuart Burroughs. We would also like to acknowledge the contribution of Jo Wills of the Waltham Centre, pathologists Drs Maurice Alley and Michelle Cooke, and the technical assistance of the clinical staff of the Massey University Veterinary Teaching Hospital. The work was performed at Massey University and supported by a grant from the Waltham Centre for Pet Nutrition, Waltham-on-the-Wolds, UK.
References 1. Anderson JA, Sogu DD, eds. Adverse Reactions to Foods. National Institute for Health Publication No. 84-2442, July 1984. 2. Denis S, Paradis M. Food allergies in dogs and cats. Part 2: Retrospective study. Med Vet Quebec 1994;24:15–20. 3. Guaguere E. Food intolerance in cats with cutaneous manifestations: A review of 17 cases. Eur J Companion Anim Pract 1995;5: 27–35. 4. Hall EJ. Gastrointestinal aspects of food allergy: A review. J Small Anim Pract 1994;35:145–152.
13
5. Guilford WG. Adverse reactions to foods. In: Strombeck DR, Guilford WG, eds. Small Animal Gastroenterology, 3rd ed. Philadelphia, PA: WB Saunders; 1996:436–450. 6. Anderson JA. Non-immunologically-mediated food sensitivity. Nutr Rev 1984;42:109–116. 7. Stogdale L, Bomzon L, van den Berg PB. Food allergy in cats. J Am Anim Hosp Assoc 1982;18:188–194. 8. White SD, Sequoia D. Food hypersensitivity in cats: 14 cases (1982–1987). J Am Vet Med Assoc 1989;194:692–695. 9. Reimann HJ, Ring J, Ultsch B, Wendt P. Intragastral provocation under endoscopic control (IPEC) in food allergy: Mast cell and histamine changes in gastric mucosa. Clin Allergy 1985;15:195–202. 10. Ownby DR. In vitro assays for the evaluation of immunologic reactions to foods. Immunol Allergy Clin North Am 1991;11:851–862. 11. Guilford WG, Strombeck DR, Rogers Q, et al. Development of gastroscopic food sensitivity testing in dogs. J Vet Intern Med 1994; 8:414–422. 12. Guilford WG. Idiopathic inflammatory bowel diseases. In: Strombeck DR, Guilford WG, eds. Small Animal Gastroenterology, 3rd ed. Philadelphia, PA: WB Saunders; 1996:454–455. 13. White SD. Food allergy in dogs. Compend Contin Educ Pract Vet 1998;20:261–268. 14. Rosser EJ. Diagnosis of food allergy in dogs. J Am Vet Med Assoc 1993;203:259–262. 15. Sousa CA. Exudative, crusting and scaling dermatoses. Vet Clin North Am 1995;25:813–831. 16. Walton GS. Skin responses in the dog and cat to ingested allergens: Observations on one hundred confirmed cases. Vet Rec 1967; 81:709–713. 17. Nelson RW, Dimperio ME, Long GG. Lymphocytic-plasmacytic colitis in the cat. J Am Vet Med Assoc 1984;184:1133–1135. 18. Paterson S. Food sensitivity in 20 dogs with skin and gastrointestinal signs. J Small Anim Pract 1995;36:529–534. 19. Batt RM, Hall EJ. Chronic enteropathies in the dog. J Small Anim Pract 1989;30:3–12. 20. Gryboski JD. Gastrointestinal aspects of cow’s milk protein intolerance and allergy. Immunol Allergy Clin North Am 1991;11:733– 797. 21. Heyman MB. Food sensitivity and eosinophilic gastroenteropathies. In: Sleisinger MH, Fordtran JS, eds. Gastrointestinal Disease, 4th ed. Philadelphia, PA: WB Saunders; 1989:1113–1134. 22. Foster AP, O’Dair H. Allergy testing for skin disease in the cat in vivo versus in vitro tests. Vet Dermatol 1993;4:111–115. 23. Elwood CM, Rutgers HC, Batt RM. Gastroscopic food sensitivity testing in 17 dogs. J Small Anim Pract 1994;35:199–203. 24. Ermel RW, Kock M, Griffey SM, et al. The atopic dog: A model of food allergy. Lab Anim Sci 1997;47:40–49. 25. Proujansky R, Winter HS, Walker WA. Gastrointestinal syndromes associated with food sensitivity. Adv Pediatr 1988;35:219– 238. 26. Patrick MK, Gall DG. Protein intolerance and immunocyte and enterocyte interaction. Pediatr Clin North Am 1988;35:17–34. 27. Sampson HA. Immunologic mechanisms in adverse reactions to foods. Immunol Allergy Clin North Am 1991;11:701–716. 28. Wasmer ML, Willard MD, Helman RG, Edwards JF. Food intolerance mimicking alimentary lymphosarcoma. J Am Anim Hosp Assoc 1995;31:463–466. 29. Walton GS, Parish WE, Coombs RAA. Spontaneous allergic dermatitis and enteritis in a cat. Vet Rec 1968;83:35–41. 30. Sampson HA, Buckley RH, Metcalfe DD. Food allergy. J Am Med Assoc 1987;258:2886–2890. 31. Curtis GH, Patrick MK, Catto-Smith AG, Gall DG. Intestinal anaphylaxis in the rat. Effect of chronic antigen exposure. Gastroenterology 1990;98:1558–1566.