Does diarrhea cause allergy? Does allergy cause diarrhea? The answer to both questions is most assuredly, yes, particularly in the young child. Any acute diarrheal illness, whether viral or bacterial may be followed by a chronic diarrheal illness due to a secondary lactose intolerance (which may last for several weeks to as long as 18 months), bacterial overgrowth in the small bowel, bile acid malabsorption or gastrointestinal allergy secondary to macromolecular penetration of and sensitization by an offending food antigen. Similarly, a singular gastrointestinal allergy may cause diarrhea through induced mucosal change (subtotal villous atrophy) and release of products of mast cell degranulation. Resultant altered bowel permeability may lead to a similar process of macromolecular penetration and sensitization.
Information regarding diagnostic techniques and immunologic mechanisms have largely been the result of studies performed upon milk-allergic patients, although diarrhea and/or small bowel abnormalities have been reported after the ingestion of gluten, soy, rice, fish, chicken and eggs. The following discussion deals largely with the findings in children with cow's milk allergy although most of these may be applied equally to patients with other gastrointestinal allergies. One must first ascertain, however, that symptoms due to soy or to the grain proteins are not in fact due to a sucrose intolerance, the symptoms of which (diarrhea and abdominal distension) may mimic allergy or even celiac disease.
The diagnosis of cow's milk allergy remains a controversial one due to the absence of absolute diagnostic criteria. It is a disease confined in most part to the first 2 years of life. The manifestations are protean, involving a number of unrelated systems or organs: skin, upper and lower respiratory tracts, stomach, small and large bowel, and hemopoietic and nervous systems. l4
Although cow's milk and related allergies have received considerable attention during the last decade with regard to clinical presentation, associated immunologic phenomena and methods of diagnosis and treatment, laboratory tests currently available to the physician are indirect and often inconclusive. Abnormal immune reactions have been identified in research laboratories, ie, lymphocyte stimulation or abnormal leukocyte Chemotaxis after exposure to milk proteins, but these are costly and available to only a small population of allergic infants. 5 The diagnosis therefore lies in correlation of the symptomatology, application of indirect laboratory data, observation of a favorable clinical response to the elimination of cow's milk, reappearance of symptoms during one carefully performed challenge and disappearance of symptoms with reinstitution of the restricted diet.
Because the diagnosis is largely a clinical one, the incidence varies with the awareness of the disease. General figures range from 0.3% to 7.5% in the general population. In a prospective study of 787 infants, Gerard et al found that 7.5% developed sensitivity to cow's milk and 30% developed sensitivities to other foods. In another prospective study, 20% of infants became allergic to other food substances but none to milk. Figures from Sweden showed an incidence of cow's milk allergy in infants of 1:7,500 in 1948 which increased dramatically to 1:200 by 1979. Figures are generally higher in infants from atopic families, in whom the incidence may reach as high as 30%.1
ETIOLOGY AND PATHOGENESIS
Heredity undoubtedly plays a major role in cow's milk allergy for there is a history of atopy in 40% to 70% and infantile or adult food allergy or upper respiratory allergies in 25% of families. Although the exact pathogenesis remains unknown, two major theories have been proposed: 1) direct protein toxicity to the intestinal mucosa similar to the toxicity of the otgliadin fraction of gluten in celiac disease, and 2) altered immunologic responses (Table 1).
Cow's milk contains more than 20 antigenic proteins with four being shown as major antigens: ßlactoglobulin, ot-lactalbumtn, casein and bovine serum albumin, with ß-lactoglobulin the most active fraction causing symptoms in 62% to 100% of sensitive infants. Intrauterine sensitization as well as sensitization to cow's milk proteins transmitted through breast milk have been postulated in infants with colic or gastrointestinal bleeding. There is evidence that heating modifies cow's milk proteins and decreases or eliminates their antigenicity. However, although the whey proteins are heat-labile, casein remains stable. Indeed, in some instances heating milk may increase its anaphylactigenic properties.
FACTORS PREDISPOSING TO COW'S MILK ALLLERGY
During digestion, gastric pepsin and pancreatic proteases form peptides which are hydrolyzed by intracellular peptidases into their component amino acids. Antigenically active peptides have been prepared from β-lactoglobulin and coproantibodies have been identified to pepsin-trypsin digests of casein and β-lactoglobulin. In vivo studies have shown efficient hydrolysis of casein but slower hydrolysis of bovine alactalbumin and ß-lactoglobulin when these proteins were presented individually to infant duodenal juice.6 Hydrolysis was considerably slower when these proteins were combined together in cow's milk or cow's milk formula. Digestion was inadequate in infants with pancreatic insufficiency.
Soybean protein is now well-recognized as having equal antigenicity with cow's milk proteins and animal studies have demonstrated enhanced uptake through an increased systemic antibody response.
In young animals and in normal infants there is a brief period of macromolecular absorption through the intestine, terminated by a "closure" phenomenon. An immune response typically occurs in infants fed cow's milk from birth.7'9 Circulating antigen-IgG antibody complexes appear at 6 days of age, and hemagglutinating antibodies between 5 to 10 days which peak at 3 months and gradually decrease thereafter. Titers in allergic infants are higher and more prolonged whereas infants who are initially breast fed have a less marked antibody response after cow's milk exposure. Antigenantibody complexes have also been reported in the sera of IgA deficient adults, stressing the role of secretory IgA in antigen penetration of the intestinal mucosa. Secretory IgA serves as a mucosal barrier in the small intestine by binding antigen. The immunoglobulin is not normally present until 6 to 10 weeks of age and its production may well be delayed in allergic infants. Its presence in breast milk exerts a role in both the prevention of allergy and enteric infection. Indeed, IgA deficiency has been reported in 10% to 40% of infants with cow's milk allergy.
Gastroenteritis itself may predispose to the development of allergy through macromolecular penetration of a damaged intestinal mucosa. Egg albumin given as a first feed after intravenous rehydration of infants with gastroenteritis is identified in postprandial serum. More direct evidence using horseradish peroxidase (having a molecular weight similar to other food antigens and generating a light- and electron-dense reaction product) shows penetration into enterocytes and lamina propria in biopsies of children with gastroenteritis. IgA-containtng cells were decreased in small bowel biopsies from children with post-enteritis allergy as compared to biopsies from normal children and those with celiac disease.
What are the types of reactions which occur after antigen has penetrated the intestinal mucosal barrier? Indeed, there are many - some which fall within the realm of normal. To review briefly, most B lymphocytes in the Beyer's patches bear surface IgA with some others bearing IgM, IgE, IgA and IgG. The cells are sensitized in the Beyer's patches, migrate to mesenteric lymph nodes where they mature and return to the intestinal mucosa to produce specific antibody. A number of our infants with cow's milk allergy have had extremely low IgG levels for their age. It has been suggested that this may be due either to immaturity of B-lymphocytes or excessive T cell suppressor function which inhibits IgG synthesis by B-cells. The T cell population of the Peyer's patches contain suppressor cells which mitigate the immune responses of the B lymphocytes and helper cells -which stimulate proliferation and maturation of the immunoglobulin-containing cells. Tolerance therefore represents an active process which is antigen-specific in which an immune response is suppressed by a subpopulation of T lymphocytes. The gradual feeding of antigen may lead to tolerance. To support these findings, Freier has demonstrated a decrease in suppressor cell activity of peripheral lymphocytes in patients with cow's milk protein allergy. Activity improves over the first 20 months of age until it reaches normal levels.5 T cells are actively involved in allergy and within several days of ingestion of a new antigen, antigen-specific T cells may damage the intestine through reagente and immuno-complex reactions and through helper T cell effects which cause direct cytotoxicity and mucosal damage from secretion of lymphokines by activated T cells.
A type III allergic reaction in which circulating antigen-antibody complexes produce tissue damage undoubtedly occurs in some infants. In support of this concept are the findings of decreased C3 complement after milk ingestion and exacerbation in some patients of steroid-responsive nephrosis by milk challenge. The association of chronic pulmonary disease in infancy (Heiner's syndrome) and circulating antibodies to cow's milk supports an immuno complex-mediated cytotoxicity with vasculitis and local inflammatory response mast-cell degranulation is an additional response to complement-derived anaphylatoxins C3a and C5a.
In 1975, it was proposed that cow's milk sensitivity was mediated by reagente IgE antibody. IgE-mediated reactions occur quickly after the ingestion of antigen. Such reactions occur in the skin of normals, sensitized with sera from allergic patients and challenged with antigen as well as in the gastrointestinal tract where a wheal can be produced by rectal instillation of antigen. The mechanism involved is binding of IgE or antigen-specific IgE antibody to the membrane receptors of tissue mast cells and basophils. Through ensuing events, degranulation occurs with release or generation of mediators involved in the allergic reaction, ie, histamine, serotonin, chemotactic factors and slow reactive substances. After challenge of the sensitized intestine, there is increased macromolecular uptake which may lead to sensitization to other food antigens (Table 2).
In the small bowel mucosa there are increased IgE immunofluorescence, mast cell degranulation and infiltration by eosinophils and polymorphonuclear leukocytes after milk challenge. Although IgM and IgG cells increase somewhat, IgA cells do not increase until 24 hours or more after challenge. There is evidence of villus injury in many infants. The enteropathy is patchy with areas of normal small bowel mucosa adjacent to others showing severe subtotal villus atrophy. Quantitative morphometry shows a mucosa thinner than that of patients with celiac disease. Lactase levels vary with the degree of histologic abnormality and revert to normal with restoration of a normal mucosa.
Gastrointestinal lesions of cow's milk allergy may extend from the stomach through the colon and symptoms may be acute or delayed.
Upper gastrointestinal bleeding occurs most often in infants 8 to 15 weeks of age with symptoms of vomiting and signs of occult blood loss in the stools. ,0 Some present with upper gastrointestinal hemorrhage. Endoscopy reveals edematous, red and friable gastric and proximal duodenal mucosa with superficial erosions. The lesions regress on a milk-free diet, but will recur in 8 to 12 hours after milk challenge. We have seen a similar course in an infant sensitive to milk and gluten. While on a milk-free diet, he experienced upper gastrointestinal bleeding 8 to 12 hours after gluten ingestion.
Lower gastrointestinal bleeding will present as blood-streaking of the stools or massive hemorrhage. Infants with this type of presentation are 2 days to 4 months old and have bleeding associated with diarrhea several days to weeks after exposure to cow's milk or soy formula or after change from heat-processed formula to whole cow's milk. Abdominal distension and emesis are frequent accompaniments and rarely hyperkalemia and shock, suggest Addison's disease. Sigmoidoscopy and biopsy are entirely compatible with a diagnosis of ulcerative colitis (red, friable mucosa and histologic evidence of increased cellular infiltrate and crypt abscesses) and hence this form of the disease has been termed milk-induced colitis. Necrotizing enterocolitis and even pneumatosis coli have been associated with this type of presentation. The diarrhea and bleeding resolve rapidly after removal of dietary milk or soy products and the rectal mucosa reverts to normal within 72 hours.
IMMUNOLOGIC FACTORS IN COW'S MILK ALLERGY
Occult blood in excess of 1.8 ml/day and enteric protein loss may occur in otherwise normal infants taking whole cow's milk. The restriction of cow's milk or reversion to a heat-processed formula will reverse the process.
Colic may be a major form of presentation of allergy with diarrhea being of minor import. Indeed, studies have shown a decrease in the incidence of colic in breastfed infants if the mothers refrain from ingesting large quantities of cow's milk.
Most often, however, one sees the infant who between 4 and 6 weeks of age (extremes 2 days to 6 months) develops a chronic and often mucoid diarrhea. If the infant has been initially breastfed, the onset is delayed. The diarrhea may be accompanied by gastroesophageal reflux and colic. With mild symptoms, growth is not disturbed, but if diarrhea is severe there is impaired growth, anemia, hypoproteinemia and generalized edema. The disease may be confused with gluten-sensitive enteropathy when it is associated with abdominal distension, muscle wasting and loss of subcutaneous fat. The stools are at times explosive, occasionally are blood-streaked and a wheal-like erythema may develop about the anus. A similar pattern is observed in some infants fed soy from birth or in milk-allergic infants given soy formula. In the latter group it is not unusual for symptoms to resolve for several days or weeks after formula change but only to recur.
In a few babies, respiratory symptoms are the major cause for concern and these range from chronic rhinitis to severe pulmonary disease. Wheezing, tachypnea and migratory pulmonary infiltrates in the young infant should always arouse a suspicion of cow's milk allergy. Heiner's syndrome of chronic pulmonary disease is associated with serum precipitins of the IgA and IgE types to milk and is assumed to represent deposition of antigen-antibody complexes within the lung. Ten percent of those affected have pulmonary hemosiderosis and hemosiderin-laden macrophages may be found in the gastric aspirate. Again, but less frequent is the appearance of this syndrome in infants taking soy.
Older children (6 months to 4 years) may develop cow's milk or other allergies after an episode of viral or bacterial gastroenteritis.11·12 Other affected family members recover while the child continues for weeks to months with diarrhea.
A chronic allergic gastroenteropathy has been described in children 2 months to 2Vi years of age. The major symptoms are hypoprotememia, peripheral eosinophilia, anemia, major allergies such as eczema or asthma, and growth retardation. The gastrointestinal symptoms consist of intermittent diarrhea, abdominal pain or vomiting after the ingestion of certain foods. All are intolerant of milk. There are enteric protein and occult blood loss and the serum contains precipitins to cow's milk. Radiologic examination of the small bowel shows mucosal edema and histologic examination is essentially normal with some infiltration of the lamina propria by eosinophils. The process is reversed by elimination of milk from the diet and/or steroid therapy.
Other diseases which have been associated with cow's milk protein allergy are: otitis media, sudden infant death syndrome, nephrotic syndrome, conjunctival injection and eyelid edema, eczema, angioneurotic edema, necrotizing enterocolitis and behavior disorders varying from weakness and lethargy to irritability and hyperactivity.
No laboratory test is available to confirm or refute the diagnosis of cow's milk allergy. The early criteria required three clinical challenges with symptoms subsiding after the elimination of milk from the diet and similar ones recurring within 48 hours after challenge. Challenge however, may be associated with anaphylaxis, cardiovascular collapse or lower gastrointestinal hemorrhage. It is now accepted that one positive challenge, carefully performed, is adequate for establishing the diagnosis.2 Challenge should be delayed for 6 months after the onset of treatment of the sensitive child and for 12 months in the infant who presented with rectal bleeding or severe enterocolitis. The major differential lies in distinguishing allergy from lactose intolerance. If mucosal damage is significant, the two may coexist. The most logical approach is to eliminate milk from the diet for 4 to 6 months and then perform a lactose tolerance or hydrogen breath test. If this is normal, then a cow's milk challenge is undertaken at 6 months.
A number of indirect confirmatory studies are used but their value is questionable. In the young infant without atopy, the serum IgE is usually normal. Older children may or may not have an elevated serum IgE. Specific allergoimmunoabsorbent tests (RAST) are not consistently positive although the enzyme-linked immunosorbent (ELISA) tests seem to be more sensitive. Similarly, skin tests may be positive or negative, although they are more often positive in atopic individuals. While coproantibodies are occasionally noted in the stools of infants and are felt by some to be nonspecific, their initial presence and their absence after elimination of cow's milk from the diet have proven reliable diagnostic measures in our experience.
Eosinophilie greater than 10% is rare, but anemia is present in 20% to 70%, vitamin K deficiency in 35% to 50%, aminoaciduria in 42% and elevated alkaline phosphatase in 10% of infants studied. Laboratory investigations have identified lymphocyte stimulation by milk and depressed leukocyte Chemotaxis in sensitized patients.
The role of pre- and post-challenge biopsy is controversial for the rapidity with which a reproducible lesion develops is variable. An abnormal intestinal mucosa may persist for months after the patient is asymptomatic and on a restricted diet.
The secondary manifestations of disaccharide intolerance, malabsorption, occult blood loss and enteric protein loss are not diagnostic for they may occur with other severe enteropathies.
There is considerable evidence to show that restriction of cow's milk in the susceptible infant for the first 6 months of life will decrease by half the incidence of gastrointestinal cow's milk allergy, but perhaps not other manifestations. This may be achieved by breast feeding or the use of soy or other hypoallergenic formulas. Since soy protein is also extremely antigenic, some infants may show an exacerbation of symptoms within days to weeks of introduction of soy into their diet, and may require more hypoallergenic formulas.
Most infants will respond to the elimination of cow's milk protein from their diet within 48 to 72 hours. Infants with rectal bleeding or severe malnutrition are given a hydrolyzed protein formula such as Nutramigen or Pregestamil in the early recovery phase since these have been shown to produce no antibody response. l3 Later, to reduce formula costs, a trial of soy formula may be considered. Wheat sensitivity is not unusual and cereals are initially restricted to those prepared from rice. Other allergies have been described in children with cow's milk allergy. Rice, fish, chicken and eggs have not only provoked clinical symptoms but have caused changes in the intestinal mucosa resembling those caused by milk.
It is our policy in sensitive infants to reintroduce gluten products after 4 months, soy at 5 to 6 months and milk products at I year after initiation of the diet. Allergic infants seem particularly prone to develop diarrhea with upper respiratory infections and teething and family or physicians may tend to restrict the diet even further, particularly for foods which they see passing through in the stool (ie, carrots, beans, corn). This is usually unnecessary.
The role of medication is limited and is reserved for children with multiple allergies. Success has been reported with disodium cromoglycate and cyproheptadine (Periactin) administered before feedings. These drugs function by stabilizing mast cell membranes and block degranulation and release of mast cell mediators. Their use is best reserved for infants with multiple allergies and best results are obtained if the drug is started after stabilization on a very restricted diet. Foods may then be reintroduced.
Although tolerance of cow's milk proteins is slowly acquired, most infants can tolerate milk-containing products and even whole milk by 1 to 2 years of age. A small minority may remain sensitive until 5 or 6 years, and other gastrointestinal allergies may be acquired. Fifteen percent to 25% of milk-allergic infants will by several years of age develop upper respiratory allergies or asthma.
1. Salatiti E, Kuitenen P, Visakorpi JK: Cow'sMilk Allergy, in Lebenthal E (ed): Textbook of Gastroenterology and Nutrition in Infancy. New York, Raven Press, 1981, pp689-708.
2. Bahrta SL. Heiner DC: Allergies to Milk. New York, Grune and Stratton. 1980, pp 23-40.
3. Visakorpi JK: Milk and soybean protein allergy^ I ftdiarr Gastroenterol Nutr 1983; 2(suppl l):293-297.
4. Lake A, Gryboski J: Inherited and metabolic disorders of absorption, in Gryboski J, Walker WA (eds): Gastrointestinal Problems in the Infant. Philadelphia, WB Saunders, 1983, pp 612-618.
5. Frier S, Kuperman O: Immune regulation in the gastrointestinal tract. ) ftdiarr Gastroenterol Nutr 1983; 2(suppl. I):310-314.
6. Jakobsson I, Borulf S, Lindberg T, et al: Partial hydrolysis of cow's milk proteins by human trypsins and elastases in vitro. I ftdiarr Gastroenterol Nutr 1983; 2:613-616.
7. Udall J, Walker WA: The physiologic and pathologic basis for the transport of macromolecules across the intestinal tract. J Pediatr Gastroenterol Nutr 1982; 1:295-301.
8. Ferguson A, Mowat A McI, Strobel S: Influence of lymphocytes and of cell-mediated immunity on the epithelial cell kinetics of the intestine, in Krerchmer N, Minkowski A (eds): Nutrition Adaptation of die Gastromtestmal Tract of the Neu/fwm. New York, Nestle, Vevey/Raven Press, 1983, pp 59-72.
9. Udail S, Kuperman O, Infcld D, et al: Nonspecific suppressor cell activity and lymphocyte response to β-lactoglobulin in cow's milk protein hypersensitivity. ] Pediatr Gastroenterol Nutr 1982; 1:389-393.
10. Coello-Ramirez P, Larossa-Haro A: Gastrointestinal occult hemorrhage and gastroduodenitis in cow's milk protein intolerance. J ftdiarr Gastroenterol Nutr 1983; 3:215-218.
11. lynkaran N, Abdin Z, Davis K, et al: Acquired carbohydrate intolerance and cow's milk protein-sensitive enteropathy in young infants. ) Pediatr 1979; 95:373, 1-12.
12. Walker-Smith JA: Cow's milk intolerance as a cause of post-gastroenteritis diarrhea. J Pediatr Gastroenterol Nutr 1982; 1:162-174.
13. Eastham E, Lichauco T, Eng K, et al: Antigenicity of infant formulas and the induction of systemic immunological tolerance by oral feeding cow's milk versus soy milk. J ftdiarr Gastroenterol Nurr 1982; 1:23-28.
FACTORS PREDISPOSING TO COW'S MILK ALLLERGY
IMMUNOLOGIC FACTORS IN COW'S MILK ALLERGY