Review Article

The Diagnosis and Management of Food Allergy and Food Intolerances

J. L. Turnbull; H. N. Adams; D. A. Gorard


Aliment Pharmacol Ther. 2015;41(1):3-25. 

In This Article

Mixed IgE- and Non-IgE-mediated Food Allergy

Conditions associated with food allergy and involving both IgE- and non-IgE-mediated mechanisms include cow's milk protein allergy, eosinophilic oesophagitis (EO) and eosinophilic gastroenteritis (EG). In the eosinophil-rich inflammatory disorders EO and EG, there is increasing evidence that food allergy is involved in their pathogenesis, although ingestion of culprit food allergens is not closely associated with the onset of symptoms.

Cow's Milk Protein Allergy

Cow's milk protein allergy (CMPA) is very common in childhood. 2–7% of babies under 1 year are affected but most become tolerant to milk in the long-term.[99,100] CMPA presents in varied ways, as there multiple overlapping mechanisms for the allergy. It is mediated through both IgE and non-IgE mechanisms, but the non-IgE reactions are less readily recognised due to a less distinct temporal relationship between exposure and symptoms. The protein content of cow's milk comprises approximately 80% casein proteins and the remainder as α- and β-lactoglobulin. Newborn babies do not possess sIgE against cow's milk and sensitisation occurs after birth. Exposure to cow's milk proteins is either directly through infant formulas, or indirectly through maternal breast milk. Lactoglobins can be detected in the breast milk of 95% of lactating women, though the level present (0.9–15 g/L) is usually not significant for most infants with CMPA.[101] However, the prevailing opinion is that early exposure to milk does not predispose to CMPA.

Cow's milk protein allergy can present like other IgE-mediated allergies with features of immediate hypersensitivity, and anaphylaxis in up to 15% of cases. However, an equal proportion present with symptoms of non-IgE mediated allergy such as treatment-resistant gastro-oesophageal reflux disease, eczema, persistent crying, diarrhoea and sometimes constipation. It can be challenging to identify these non-IgE reactions as there is considerable overlap between these symptoms and other common complaints of infancy. CMPA should be considered in children with atopy, where standard therapies for eczema or gastro-oesophageal reflux have been ineffective, where there is a clear milk dose-dependent effect, and if there are symptoms in more than one system.[102]

Cow's milk protein allergy may be responsible for up to 40% of childhood gastro-oesophageal reflux disease. The milk protein induces mast cell degranulation, disturbing stomach peristalsis and altering lower oesophageal sphincter tone leading to gastro-oesophageal reflux.[103,104] CMPA can also lead to disordered colonic motility and constipation.

Cow's milk protein allergy is very common and can be managed in primary care.[21,102] However, since anaphylaxis and even death have resulted from CMPA, the baby should be referred to a paediatric allergy specialist if there has been a severe reaction. Similarly faltering growth, atopic comorbidity, multiple food allergies, incomplete resolution after milk exclusion or diagnostic uncertainty should prompt a specialist referral.[99] Suspected IgE allergies should be investigated with skin prick tests or sIgE levels, but if the symptoms suggest a non-IgE mechanism, an elimination diet and re-challenge will suffice for diagnosis.

Cow's Milk Protein-free Diet. Cow's milk protein allergy is managed by exclusion of cow's milk. A response may be seen in 2–6 weeks. Ideally, there should then be a brief re-challenge to prove that symptoms recur. Breast-feeding can continue if the mother agrees to a milk-free diet, supplemented with calcium and vitamin D. Formula-fed infants should be prescribed an extensively hydrolysed formula (Table 8). Extensively hydrolysed formulas contain peptides with molecular weight less than 3000 Da. This is adequate in 90% of formula-fed infants with CMPA, since peptides usually need to have a molecular weight of 10 000–70 000 Da to be allergenic. If symptoms do not improve after 2 weeks of extensively hydrolysed formula, an amino acid based formula may be required. An amino acid based formula should be used from the onset if there are features of anaphylaxis, severe GI bleeding, faltering growth, severe eczema or symptoms that developed whilst exclusively breast feeding.[105]

Increasingly some parents are choosing to use other animal or soy-based formula milks, with the intention of avoiding milk allergy. Soy formulas do not protect against eczema development or allergy in general, and should not be used as a replacement in CMPA as there is cross-reactivity in up to 14% of IgE-mediated and up to 60% of non-IgE-mediated allergies.[106] Soy is not recommended in the first 6 months of life due to the presence of phytoestrogens. However, soy can be trialled after 6 months and soy products can be used in weaning. There is significant cross-reactivity between cow's and other animal milks (e.g. goat's), and these are not recommended.

In CMPA, the milk-free diet should be continued until at least 1 year of age, or 6 months after diagnosis. A milk challenge can then normally be conducted at home, but the hospital setting is advised where there is suspicion of FPIES or anaphylaxis or current atopic eczema with positive milk skin prick test. Under the advice of a dietitian the family can follow a 'milk ladder' – an escalating regimen of theoretically more allergenic preparations and quantities of milk.[107] Around 55% of children with IgE-mediated CMPA become tolerant to milk by 5 years of age.[108] With non-IgE mediated CMPA, most children are milk tolerant by 2.5 years.[109]

Eosinophilic Oesophagitis

Eosinophilic oesophagitis (EO) is an increasingly recognised chronic inflammatory condition of the oesophagus affecting both children and adults.[110,111] There is a high concentration of intraepithelial eosinophils, but it is a tansmural disease.

Both food allergens and airbourne allergens seem to be relevant in the pathogenesis of EO.[112,113] There is a strong association between EO and atopic illnesses such as asthma, allergic rhinitis and eczema.[112] Peripheral eosinophilia is common. Patients with EO often have a history of food allergy, and often have positive skin prick tests and sIgE to foods,[9,113] although these links to allergy are weaker in adults. The most common food allergens testing positive in EO are cow's milk, soy, eggs and wheat.[114]

Children suffering from EO are usually diagnosed within the first 3 years of life and can present with abdominal pain, reflux, vomiting, dysphagia, airways symptoms, cough and chest pain.[115] Adolescents and adults with EO often present with dysphagia and oesophageal food impaction. In EO there are some characteristic endoscopic findings and histological features of eosinophilic inflammation (≥15 eosinophils per high power field).[110,112]

Although treatment with topical corticosteroids and endoscopic dilatation of oesophageal strictures are effective therapies in EO,[110] dietary therapy is also important. Avoidance of putative culprit food allergens is an appealing strategy. Dietary strategies may involve an elemental diet, a six-food elimination diet (6FED) and a targeted-elimination diet.

Elemental Diet. An amino acid-based elemental diet improves symptoms and histological features in children with EO.[116,117] Elemental diet is less useful in adults. It improves histological features but not symptoms of EO.[118] Elemental diets are difficult to adhere to and impractical for lengthy periods. However these observations have provided the first evidence for food-triggered oesophageal eosinophilia.

Six-food Elimination Diet. A six-food exclusion diet, devoid of cow's milk protein, soy, wheat, egg, peanut, and seafood has been shown to induce remission and improve histology in a majority of children with EO.[119] Serial single-food reintroductions following induction of histological remission with the 6FED can lead to the identification of specific causal food antigen(s) in EO.[120] Cow's milk is the most common single trigger identified after reintroductions. The same 6FED has also been found to be effective in adults, improving symptoms and reducing endoscopic and histopathological features of EO.[121] Skin prick tests were not useful in identifying relevant food allergens in these adults.

Targeted-elimination Diet. Tailoring an elimination diet to the results of allergy testing should allow fewer foods to be removed from the diet. In vitro allergy testing for sIgE is not very useful in EO, and this may be in part due to the fact that some of the mechanisms of allergy in EO are not IgE-mediated. However, testing for food allergens and for environmental aeroallergens using skin prick tests should be considered.[122] In a recent study, only the foods to which children with EO had positive results on skin prick tests and atopy patch tests, were excluded.[114] This led to a similar resolution of oesophageal eosinophilia as the 6FED but allowed a more liberal diet.

More work is required in children and in adults to identify incriminating food allergens in EO to develop acceptable dietary therapy. However in children at least, an empiric 6FED or a targeted elimination diet both seem effective.[114,123] Motivated adult patients can be offered similar dietary therapies as an alternative to long-term topical steroid therapy. Although an empiric dietary approach of eliminating six foods can be instigated with skilled dietetic input, a targeted approach using allergy testing will involve the expertise of an allergist. Gastroenterologists managing patients with EO should ideally have links with allergists as well as dieticians.[124]

Eosinophilic Gastroenteritis

Eosinophilic gastroenteritis includes eosinophilic gastritis, eosinophilic enteropathy and eosinophilic colitis, and can manifest at any age with a male predominance.[125] Symptoms may mimic those of IBS with chronic abdominal pain, nausea, vomiting, diarrhoea, but if there is more extensive small bowel involvement then weight loss, anaemia and malabsorption features are evident.[9,125] The constellation of symptoms involved in EG varies depending on the site of pathological eosinophil infiltration, depth of involvement and if it is localised or wide-spread.[9] The diagnosis of EG is not always straightforward but there are some characteristic endoscopic and histopathological features.

Evidence for a link between EG and food allergy is based on an association with atopy and the response to elimination diets. Most patients with EG have atopy,[126,127] a peripheral eosinophilia and elevated total serum IgE. Eosinophilic gastroenteritis is rare and most cases are managed with corticosteroid drugs. There is no controlled trial evidence of dietary therapy in EG, but retrospective case reports and series suggest that improvement can be achieved with elimination diets. Eliminating the foods implicated by skin prick tests or sIgE is unpredictable. However successful resolution of symptoms has been reported in a series of children developing eosinophilic colitis before the age of 2, by empirically eliminating foods starting with cow's milk and eggs.[128] In that series exclusion diets were tailored and maintained according to the results of food challenges. All foods were successfully reintroduced by 5 years.[128]