Acta Pdiatrica ISSN 08035253

VIEWPOINT ARTICLE

Gastrointestinal manifestations of cows milk protein allergy and gastrointestinal motility

Yvan Vandenplas (yvan.vandenplas@uzbrussel.be)1, Frederic Gottrand2, Gigi Veereman-Wauters1, Elisabeth De Greef1, Thierry Devreker1, Bruno Hauser1, Marc Benninga3, Hugo SA Heymans3
1.UZ Brussel, Vrije Universiteit Brussel, Brussels, Belgium 2.Department of Paediatrics, Jeanne de Flandre, University Hospital, Faculty of Medicine, University Lille2, Lille, France 3.Department of Paediatrics, Academic Medical Centre Emma Childrens Hospital, Amsterdam, The Netherlands

Keywords Constipation, Cow milk protein hydrolysate, Cows milk protein allergy, Extensive hydrolysate, Functional gastrointestinal symptoms, (Gastrointestinal) motility, Gastro-oesophageal reux, Intolerance, Prevention of cows milk allergy, Regurgitation, Treatment of cows milk allergy Correspondence Yvan Vandenplas, UZ Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium. Tel: +3224775780 | Fax: +3224775783 | Email: yvan.vandenplas@uzbrussel.be Received 10 April 2012; revised: 19 July 2012; accepted 1 August 2012. DOI:10.1111/j.1651-2227.2012.02808.x

ABSTRACT
Cows milk protein allergy (CMPA) may cause gastrointestinal motility disorders. Symptoms of both conditions overlap and diagnostic tests do not reliably differentiate between both. A decrease of symptoms with an extensive hydrolysate and relapse during challenge is not a proof of allergy, because hydrolysates enhance gastric emptying, a pathophysiologic mechanism of gastro-oesophageal reux (GER). Thickened formula reduces regurgitation, and failure to do so suggests CMPA. A thickened extensive hydrolysate may induce more rapid improvement, but does not always differentiate between CMPA and GER. Different hypotheses are discussed: is the overlap between CMPA and functional disorders coincidence, or do both entities present with identical symptoms, or does the fact that symptoms are identical indicates that there is only one entity involved? Studies on the prevention of CMPA focused on at-risk families, and resulted in a decrease of CMPA and atopic dermatitis, but did not provide data on the incidence of GER. Conclusion: As long as there are no objective diagnostic tools to separate GER from CMPA, the physician has two options: rst treat the most likely diagnosis, and switch if after 24 weeks there is no improvement, or treat both conditions with one intervention, what will not result in a diagnosis.

INTRODUCTION The overlap between gastrointestinal (GI) manifestations of cows milk protein allergy (CMPA) and frequent (functional) GI complaints such as gastro-oesophageal reux (disease) [GER(D)] and constipation is a topic of debate since many years. The debate is the logic consequence of the fact that objective diagnostic criteria for each of the entities are missing. Because not one sign or symptom is specic for CMPA, and because the same is valid for GER(D) and constipation, and because all conditions are relatively frequent, it is a given fact that some of the allergic infants will present with GER(D) and or constipation and that some of the infants with GERD or constipation will have CMPA. Many infants with CMPA present symptoms in different organ systems (1). Because GI manifestations are only part of the spectrum of symptoms caused by CMPA, this paper does not discuss CMPA in general. If more than one organ system is involved (mainly GI and cutaneous symptoms, or GI and more general symptoms or more rare GI and respiratory tract symptoms), a systemic or (auto-)immune or allergic reaction is very likely (1). The debate is focused on cows milk as this is the major food allergen in infants. While regurgitation decreases strongly between 6 and 12 months of age (2), tolerance to cows milk protein (CMP) does only develop

after the age of 1 year (1). There is also a pathophysiologic overlap: an allergic reaction causes inammation and secretion of substances such as histamine and serotonin. The GI tract reacts to the inammation by altering motility. As a result, the question raises if there is coincidence or if one is the logic consequence of the other. From the point of view of the gastroenterologist looking to the GI tract as an organ responsible for transport of food from mouth to anus, regurgitation and constipation are mainly regarded as functional disorders, because anatomic malformations are infrequent. Denitions have been proposed in the Rome III criteria (3). The recent changes in formula composition, such as addition of nucleotides, long-chain poly-unsaturated fatty acids (LcPUFAs), prebiotic oligosaccharides and or probiotics, are likely to interfere in this debate. All these novelties claim to induce a better immune response and to decrease allergy. They also potentially change GI tract motility. It is, for example, claimed that stools are softer with the added preand probiotics. Most of the epidemiologic data on CMPA I date from before these additions. Epidemiologic data collected after the formula changes listed above in unselected populations are missing. If the claimed theoretical benet results in a clinical benet, the incidence of frequent GI complaints should have decreased.

2012 The Author(s)/Acta Pdiatrica 2012 Foundation Acta Pdiatrica 2012 101, pp. 11051109

1105

Cows milk protein allergy and gastrointestinal motility

Vandenplas et al.

COWS MILK PROTEIN ALLERGY OR FUNCTIONAL GI SYMPTOMS? Cows milk protein allergy is a reproducible clinically abnormal reaction to cows milk protein (CMP) owing to the interaction between one or more milk proteins and one or more immune mechanisms. About 90% of the patients develop symptoms before the age of 3 months depending on the moment of CMP introduction or within 2 months after introduction. CMPA rarely develops after the age of 12 months. Improvement or disappearance of symptoms on a CMP-free diet adds substantial evidence to the diagnosis. A factor that often is not considered is the time needed for improvement of symptoms; the sooner, the more likely that allergy may be involved. If the reintroduction of CMP causes relapse of symptoms, the diagnosis seems established, because a challenge test is considered as the golden standard diagnostic test. Although false-positive diagnostic testing (specic IgE, skin prick test, patch test) does occur, the diagnosis of CMPA is likely if patients present with suggestive symptoms and (one of) these tests show(s) positive results. But, CMP can also lead to non-IgE-dependent reactions. Some symptoms are more frequently linked to non-IgE-mediated allergy. In daily routine, there is no diagnostic testing for non-IgE-mediated allergic reactions. But, immunoglobulin free light chain (Ig-fLC)-dependent allergic hypersensitivity responses have been demonstrated to occur (in mice) (4). Also, in children affected with IgE-mediated and non-IgEmediated CMPA or atopic dermatitis (AD), serum Ig-fLC concentrations were increased, implying the relevance of Ig-fLC measurements in the diagnosis of human allergic disease (4,5). These ndings suggest that in the future, parameters may become available to contribute to the diagnosis of CMPA, although the overlap between allergic and nonallergic children is important (4). The relationship between Ig-fLC and GI symptoms has not been investigated. The old term intolerance gives rise to confusion. To avoid this, intolerance is proposed to be restricted to the incapacity to fully digest carbohydrates, mainly disaccharides, of which lactose is the most important one. Primary lactose intolerance is an almost nonexisting disease at this age. Thus, if lactose intolerance occurs in infants, it is almost always secondary to another disease and the consequence of atrophy or damage of the villi. Low lactase activity may also be the consequence of a transitory slow maturation in which case the undigested lactose acts as a prebiotic. As a consequence, the term adverse reaction would be preferable to intolerance. The question arises how to make the difference between an allergic reaction of which the immunologic mechanism involved cannot be shown in routine (non-IgE-mediated allergy) and a functional symptom? Bearing this in mind, the term non-IgE-mediated CMPA will be used to designate this group of infants with adverse reactions to CMP and no increase of specic IgE or no clear positive dermatological contact tests such as skin prick or patch tests. Gastro-oesophageal reux (D) has always been mentioned as one of the presenting manifestations of CMPA. Buisseret mentioned vomiting as a typical presenting

symptom of CMPA in a Lancet paper in 1978 (5). Forget reported in 1985 a small series of 15 children presenting with recurrent vomiting, not responding to GER therapy, that became symptom free on a CMP-free diet (6). Difculties to demonstrate the relation between CMPA and GER are illustrated in the study by Nielsen et al. (7). The authors performed a 48-h pH monitoring in infants presenting with GERD and cow milk hypersensitivity; they could demonstrate that the pH monitoring was more abnormal in infants with GERD and allergy than in the other groups, but a challenge test did not increase reux (7). Ravelli et al. (8) used electrogastrography to show that gastric motility reacted different to a CMP challenge in control infants, infants with GERD and infants with CMPA. This abnormal motility could be related to a delayed gastric emptying (8). A cows milk challenge increases weakly acidic reux (pH > 4 and <7) in children with CMPA and GER (9). Impedance-pH monitoring may therefore be useful in identifying a subgroup of infants with cows milk proteininduced GERD (9). Unfortunately, only rough estimates can be made on the prevalence or incidence of GI manifestations of CMPA. The consequence of the denition that not every allergic reaction is IgE mediated in combination with the fact that today only IgE reactivity can be measured in daily routine is that non-IgE-mediated allergy has become a synonym for intolerance or adverse reaction. Studies using a stringent scientic approach focus mainly on IgE-mediated allergy and or a positive challenge test and report 25% as the incidence of CMPA. However, up to 10% or even 15% of the infants seem to develop some adverse reactions to cows milk. There is a quite broad consensus that in infants a challenge test can be performed open, mainly because of the complexity of a double-blind challenge (1). If no symptoms are elicited within 2 weeks of regular cows milk feeding, CMPA can be excluded. Challenge tests can be performed in an inpatient or outpatient setting to document signs and symptoms, the milk volume that provokes the symptoms and to give symptomatic treatment if needed (10). The starting dose during milk challenge should be lower than a dose that can induce a reaction and increase stepwise to 100 mL (1,10). More details on the practical aspects of a challenge test are discussed elsewhere (1). To rule out a false-positive test result owing to lactose intolerance, the challenge procedure should be performed with a lactose-free CMP-containing milk (10). Challenges should be carried out in a hospital setting in the following circumstances: a previous history of immediate type allergic reactions, unpredictable reaction such as in infants with positive IgE sensitivity who has never been given cows milk or has not been given cows milk for a long time, and severe atopic eczema (owing to difculty in assessment) (10). It is known that a double-blind challenge reduces the number of infants considered allergic after a positive open challenge with about one-third (11). Infants that react 24 h or even later to a challenge or react only to large volumes may not be picked up by a double-blind challenge. Although symptoms during the challenge are most of the time the

1106

2012 The Author(s)/Acta Pdiatrica 2012 Foundation Acta Pdiatrica 2012 101, pp. 11051109

Vandenplas et al.

Cows milk protein allergy and gastrointestinal motility

same as the original presenting symptoms, this is not a dogma. During challenge, presenting symptoms may differ from the original. As said before, infants with non-IgE-mediated CMPA tend to become tolerant to CMP more rapidly than those with IgE-mediated CMPA. The higher the IgE level, the more likely the allergy is going to persist (12). This means that if a challenge is performed some months after the initial suspected diagnosis and turns out negative, it cannot be concluded that the original diagnosis of CMPA was erroneous. A negative challenge only means that at the moment of the challenge the food is tolerated. And a positive challenge does not conrm that the immune system is involved. Bearing all the above in mind, it seems reasonable to estimate that between 5% and 10% of the formula-fed infants will develop some adverse reaction to CMP. Unfortunately, there are insufcient data on the number of supposed allergic infants that present only with GI manifestations. About 20% of all 3- to 4-month-old infants regurgitate more than four times a day, and this seems to be a threshold for mothers to seek medical help (13). It is probably reasonable to estimate the (recent) prevalence of infant constipation below 5%. Prebiotic-supplemented formula is well tolerated by full-term infants. It increases stool colony counts of bidobacteria and lactobacilli and results in stools similar to those of breastfed neonates without affecting weight gain (14,15). The administration of Lactobacillus reuteri (DSM 17938) in infants with chronic constipation had a positive effect on bowel frequency, even when there was no improvement in stool consistency and episodes of inconsolable crying episodes (16). In the light of the difculties to distinguish between allergy and adverse reactions and functional disorder, the question arises if the overall prevalence of GI complaints could not be considered as an equivalent for IgE- and nonIgE-mediated allergy? To obtain more accurate data on prevalence and incidence, proactive data collection via questionnaires in representative population samples is mandatory to minimize the impact of parental coping with these frequent symptoms. In daily primary health care, it is not easy and possibly even not clinically relevant to separate functional disorders from non-IgE-mediated allergy if therapeutic options tackling both conditions can be offered. Efcacy of certain therapeutic interventions does not help to separate non-IgEmediated allergy from functional GI manifestations. The recommended treatment of CMPA is elimination of CMP from the mothers diet in breastfed infants, amino acidbased formula (AAF) in formula-fed infants with severe manifestations of CMPA (life-threatening symptoms, failure to thrive) or extensive hydrolysates for the majority of infants (1,10). Published evidence with the new rice hydrolysates is too recent and limited to include them in guidelines (10). The NASPGHAN-ESPGHAN guidelines on the management of regurgitation and reux recommend in distressed and regurgitating infants a therapeutic approach with either thickened formula or an extensive hydrolysate (13). If a thickened regular infant formula results in a

signicant reduction of episodes of regurgitation, allergy seems unlikely because protein structure did not change. But what if an extensive hydrolysate results in a signicant reduction of regurgitation? As hydrolysates have a much faster gastric emptying than native protein, and as delayed gastric emptying is a pathophysiologic mechanism causing GER, improvement of regurgitation with a hydrolysate cannot be considered as a proof of allergy. Partial hydrolysates are not indicated in the treatment of CMPA (1,10). Partial hydrolysates may be tolerated in about half of the infants with IgE-mediated CMPA (17). Partial hydrolysates are also considered to be more easily digested, although scientic evidence for this statement is missing. Several infant formula companies commercialized thickened partial hydrolysates. As a consequence, a thickened extensive hydrolysate may be considered as a treatment option in infants presenting with troublesome regurgitation and in whom allergy would be a plausible diagnosis on clinical grounds. This attitude would mean that a number of nonallergic infants are given an extensive hydrolysate for no reason and that a number of allergic infants are given a thickened formula for no reason. But, up to now there is no convincing literature regarding nutritional adverse events of a thickened formula or an extensive hydrolysate, if cost is not considered. In difcult cases, AAF may contribute to separate infants with CMPA from those with reux. Indeed, about 5% of the CMPA infants may still react to an extensive hydrolysate (1). If that is the case, the infant will improve with AAF. Mainly because AAF is very expensive, and because reimbursement systems differ from country to country, there is still debate whether AAF should be used rst line (in all infants) or second line (only in infants not improving on eHF). A long-term elimination diet and feeding restrictions may lead to intake problems during diversication. In clear-cut CMPA, the CM-free diet is recommended at least during 69 months and up to the age of 1 year (1). The difculties to diagnose CMPA strengthen the importance of prevention. Infant formula is derived from cows milk for the simple reason that there are many cows and because cows provide more milk. There is broad consensus that almost every constituent of cows milk needs to be adapted according to better full the nutritional needs of infants: the total amount and proportion of protein, the amount of minerals, the addition of nucleotides, LcPUFAs, etc. So, why not accept that it would also be preferable to change the structure of the proteins and feed every infant a (partial) hydrolysate? Unfortunately, data showing benet of doing so are limited to at-risk populations. There are very limited data showing a reduction of CMPA in the general population. This is merely the consequence of methodological problems regarding the size of the needed study populations. There are in absolute numbers more allergic children in the non-at-risk group than in the at-risk group, for the simple reason that the non-at-risk group is much larger. Regarding constipation, a similar criticism can be hypothesized. Several studies mentioned constipation as a (frequent) symptom of CMPA. But milk, especially casein, is known to constipate. Hydrolysates are known to cause soft,

2012 The Author(s)/Acta Pdiatrica 2012 Foundation Acta Pdiatrica 2012 101, pp. 11051109

1107

Cows milk protein allergy and gastrointestinal motility

Vandenplas et al.

greenish stools. One of the studies on CMPA and reux used soy formula as treatment, thus avoiding the stool-softening effect of hydrolysates (18). However, this study can be criticized because it was reported that soy formula was effective in 100% of the CMPA allergic infants, while there is consensus that at least some (around 10%) CMPA infants develop also allergy to soy. Moreover, paediatric gastroenterologists know (although epidemiologic data are missing) that constipation as a functional disorder on soy formula exists, probably with the same frequency as functional constipation with regular infant formula. Again, whether this is functional or non-IgE-mediated allergy can be debated. Is constipation really a symptom of CMPA? According to data from Italy, the incidence of constipation is not different in atopic and nonatopic children (19). Therefore, constipation as single manifestation of CMPA seems to be relatively rare. Anti-constipation formulas have been commercialized. Although some of these formulas have been evaluated in small trials (20,21), scientic evidence is very limited. As extensive hydrolysates are known to cause greenish semisolid stools, their use could be considered, although there are no data on the efcacy of extensive hydrolysate in infant constipation. An interaction between GER regurgitation and constipation has been advocated as well. Frequent regurgitation will result in an excessive loss of water, and may therefore be considered as a risk factor for constipation.

treatment of functional constipation in infants younger than 6 months is difcult to recommend because of the poverty of data. The efcacy of anti-constipation formula is not documented, but neither is the efcacy of extensive hydrolysates.

CONFLICT OF INTEREST YVDP is the speaker for Abbott and Mead Johnson Nutritionals, and he is a consultant and speaker of Biocodex, Danone (Nutricia, SHS), Nestle Nutrition and United Pharmaceuticals. FG is a consultant and speaker of Nutri , Movetis and Astra Zeneca. cia, Danone, Nestle

References
1. Vandenplas Y, Koletzko S, Isolauri E, Hill D, Oranje AP, Brueton M, et al. Guidelines for the diagnosis and management of cows milk protein allergy in infants. Arch Dis Child 2007; 92: 9028. 2. Hegar B, Dewanti NR, Kadim M, Alatas S, Firmansyah A, Vandenplas Y. Natural evolution of regurgitation in healthy infants. Acta Paediatr 2009; 98: 118993. 3. Hyman PE, Milla PJ, Benninga MA, Davidson GP, Fleisher DF, Taminiau J. Childhood functional gastrointestinal disorders: neonate toddler. Gastroenterology 2006; 130: 151926. 4. Schouten B, van Esch BC, van Thuijl AO, Blokhuis BR, Groot Kormelink T, Hofman GA, et al. Contribution of IgE and immunoglobulin free light chain in the allergic reaction to cows milk proteins. J Allergy Clin Immunol 2010; 125: 1308 14. 5. Buisseret PD. Common manifestations of CMA in children. Lancet 1978; 8059: 3045. 6. Forget P, Arends JW. Cows milk protein allergy and gastrooesophageal reux. Eur J Pediatr 1985; 144: 298300. 7. Nielsen RG, Bindslev-Jensen C, Kruse-Andersen S, Husby S. Severe gastro-esophageal reux disease and cow milk hypersensitivity in infants and children: disease association and evaluation of a new challenge procedure. J Pediatr Gastroenterol Nutr 2004; 39: 38391. 8. Ravelli AM, Tobanelli P, Volpi S, Ugazio AG. Vomiting and gastric motility in infants with cows milk allergy. J Pediatr Gastroenterol Nutr 2001; 32: 5964. 9. Borrelli O, Mancini V, Thapar N, Giorgio V, Elawad M, Hill S, et al. Cows milk challenge increases weakly acidic reux in children with cows milk allergy and gastroesophageal reux. J Pediatr 2012 (in press). 10. Koletzko S, Niggemann B, Arato A, Dias JA, Heuschkel R, Husby S, et al. Diagnostic approach and management of cows milk protein allergy in infants and children: a practical guideline of the GI-committee of ESPGHAN. J Pediatr Gastroenterol Nutr 2012; 55: 2219. 11. Kneepkens CM, Meijer Y. Clinical practice. Diagnosis and treatment of cows milk allergy. Eur J Pediatr 2009; 168: 8916. 12. Skripak JM, Matsui EC, Mudd K, Wood RA. The natural history of IgE-mediated cows milk allergy. J Allergy Clin Immunol 2007; 120: 11727. 13. Vandenplas Y, Rudolph CD, Di Lorenzo C, Hassall E, Liptak G, Mazur L, et al. Pediatric gastroesophageal reux clinical practice guidelines: joint recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr 2009; 49: 498547.

SUMMARY It is extremely difcult, virtually still impossible today, to separate non-IgE-mediated allergy from functional disorders in infants presenting with GI adverse reactions to cows milk-based formula. If more than one organ system is involved, the diagnosis of allergy is much more likely than functional disorder. In distressed infants presenting with troublesome regurgitation and or vomiting, two options for dietetic treatment seem reasonable. If the goal is symptom relief as soon as possible, a thickened extensive hydrolysate is an option. If symptom relief is obtained, it will not allow to distinguish allergy from functional disorders. If the goal is to separate functional symptoms from allergy, again two choices are possible. If a functional disorder is the most likely diagnosis (regurgitation and vomiting as single manifestation, negative family history of atopic disease), a thickened formula with unmodied protein should be the choice. If allergy is the more likely diagnosis (different organ systems involved, positive family history of atopic disease), an extensive hydrolysate should be the preferred option. A re-evaluation after 2 weeks is recommended and other diagnostic options should be considered according to the clinical evolution. As in many cases the diagnosis is not clearly established, duration of treatment should be limited. Regarding constipation, the situation is even more nebulous. First, true constipation in formula-fed infants below 6 months seems rare, both as functional disorder and as single manifestation of CMPA. As a consequence,

1108

2012 The Author(s)/Acta Pdiatrica 2012 Foundation Acta Pdiatrica 2012 101, pp. 11051109

Vandenplas et al.

Cows milk protein allergy and gastrointestinal motility

14. Ziegler E, Vanderhoof JA, Petschow B, Mitmesser SH, Stolz SI, Harris CL, et al. Term infants fed formula supplemented with selected blends of prebiotics grow normally and have soft stools similar to those reported for breast-fed infants. J Pediatr Gastroenterol Nutr 2007; 44: 35964. 15. Rao S, Srinivasjois R, Patole S. Prebiotic supplementation in full-term neonates: a systematic review of randomized controlled trials. Arch Pediatr Adolesc Med 2009; 163: 75564. 16. Coccorullo P, Strisciuglio C, Martinelli M, Miele E, Greco L, Staiano A. Lactobacillus reuteri (DSM 17938) in infants with functional chronic constipation: a double-blind, randomized, placebo-controlled study. J Pediatr 2010; 157: 598602. 17. Giampietro PG, Kjellman NI, Oldaeus G, Wouters-Wesseling W, Businco L. Hypoallergenicity of an extensively hydrolyzed whey formula. Pediatr Allergy Immunol 2001; 12: 836.

18. Iacono G, Carroccio A, Cavataio F, Montalto G, Kazmierska I, Lorello D, et al. Gastroesophageal reux and cows milk allergy in infants: a prospective study. J Allergy Clin Immunol 1996; 97: 8227. 19. Simeone D, Miele E, Boccia G, Marino A, Troncone R, Staiano A. Prevalence of atopy in children with chronic constipation. Arch Dis Child 2008; 93: 10447. 20. Chao HC, Vandenplas Y. Therapeutic effect of Novalac-IT in infants with constipation. Nutrition 2007; 23: 46973. 21. Bongers ME, de Lorijn F, Reitsma JB, Groeneweg M, Taminiau JA, Benninga MA. The clinical effect of a new infant formula in term infants with constipation: a double-blind, randomized cross-over trial. Nutr J 2007; 6: 8.

2012 The Author(s)/Acta Pdiatrica 2012 Foundation Acta Pdiatrica 2012 101, pp. 11051109

1109

Copyright of Acta Paediatrica is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.