This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Food hypersensitivity is the most common cause of anaphylaxis in children in western countries and more specifically in the UK . It refers to all types of overreactions against a food component and is subdivided into two types of reactions; immune-mediated (food allergy) and non-immune mediated (food intolerance).
Food allergy, a public health problem, is characterized by an abnormal or exaggerated immune response to specific food allergens resulting to disease while Food intolerance is the physiological reactions to food additives or ingested food and they are not immunologic but can be metabolic, pharmacological or toxic (Gerth et al, 2003). This is the same as lactose intolerance and it is always confused for CMPA due to the similarity of symptoms but it is different since lactose is a milk carbohydrate which cannot be digested due to absence of lactase enzyme (Bahna et al, 2002).
Food allergy prevalence is 6-8% in children less than 3 years and 2% of adults (Noimark et al, 2008). Studies have shown that children living in the cities have a higher likelihood as compared to those from the countryside (FAI, 2008) though the prevalence varies with geographical distribution (Sicherer et al, 2006). In the USA, food allergies affect dietary habits in most homes (Gerth et al, 2003). 2.5% of infants have hypersensitivity to cow's milk in the first year of life (Fiocchi et al, 2010).
Common allergies are peanuts, milk, eggs, tree nuts, seafood, shellfish, soy, wheat and sesame due to the protein in the oil and they are wholesomely known as the 'Big eight' (Fai, 2008). Rice allergy has also been observed in East Asia where rice is the staple food (Ikezawa et al, 1992). Studies have however shown that these allergens are rare due to proteolytic activity and heat processing (Yamada et al, 2005). The nutrients affected are as shown in table 1.
Table 1: Nutritional considerations affected by food allergies (Skypala et al, 2009)
Sources affected by food allergy
Egg yolk, liver, kidney, muscle, and organ meats, certain vegetables
Liver, yeast extract, green leafy vegetables, legumes, certain fruits
Beef, pork, wheat flour, maize flour, eggs, cow's milk
Chicken, beef, potatoes, oat cereals, tomato products, liver, lean meats, broccoli, whole grains
Milk, eggs, enriched cereals and grain, ice cream, liver, some lean meats, green vegetables
Unrefined grain products, white and brown flour, meat products, vegetables, dairy products, legumes, fruits and eggs
Fish, liver, kidney, pork, eggs, milk, wheat germ, brewer's yeast, brown rice, soybeans, oats, whole-wheat grains, peanuts and walnuts
Meat (particularly liver) and fish
Citrus and soft fruits, leafy green vegetables, kidney and liver
Dairy products, fortified margarine, liver, fish oils
Yellow and green vegetables
Fatty fish, fish oils, milk, eggs, fortified margarine
Plant oils (e.g. soybean oil, corn oil, olive oil), meat, poultry, dairy products
Green leafy vegetables, vegetable oils, dairy products, meat, eggs
Nuts, fruit, green vegetables
Processed meats, wholegrain products, pulses, spices
Fish, nuts, green leafy vegetables, fresh cereals
Shellfish, offal, nuts
Shellfish, marine fish, sea salt, cereals and grains, cows milk
Green vegetables, nuts, bread and other cereals, tea
Leafy vegetables, cauliflower, nuts, legumes, cereals
Fish, offal, brazil nuts, eggs, cereal
Meat, cereal products
Milk, green leafy vegetables, almonds, hazel nuts
Liver, meat, beans, nuts, dried fruits, poultry, enriched cereals, soybean flour, green leafy vegetables
Leafy vegetables, grains, nuts, dairy products, meats
Red meats, fish, dairy products, poultry, bread and other cereals
Milk, fruit, vegetables, fish, shellfish, beef, chicken, turkey, liver, salt substitutes
Grains (e.g. oats, barley, rice), beer
Table salt, cereals and cereal products e.g. bread, meat and meat products, crisps and savoury snacks
The most common food allergy is Peanut. A survey carried out by Sicherer et al, 2006 showed that Peanut and tree nut allergy affected 1.3% of adult population while 2% of adults had food allergies in USA. The results of the study were replicated by studThis was also the case in the UK according to surveys carried by Gupta et al, 2007. In Portugal, self-reported allergy in adults was 4.8% (p<0.01) while in USA 9.1% on self-reports and 5.3% on doctor diagnosis (Falcao et al, 2004). The self reports could lead inaccuracy for majority of the population do not know how to differentiate intolerance and allergy. The DRACMA guidelines, 2010 however showed that it is essential for guidance on medical services for public health planning and users between general and specialist medicine
Cow's milk protein allergy is the 2nd common since it is ubiquitous in foods as compared to other allergens (Boyano-Martinez et al, 2008). IgE reactions also known as immediate type I hypersensitivity affect 8% of patients with cow's milk being the major causative allergen (Fox et al, 2007). Children with atopic disorders e.g. eczema have increased risk of food allergy while about 6% who have asthma have food-induced wheezing (Gerth et al, 2003). About 10% of allergic children will react to beef since it contains a small amount of protein found in cow's milk (Sicherer et al, 2006). Nowak-Wegrzyn et al, 2008 showed that 75% children could tolerate heated (baked) cow's milk and this was due to the fact that high temperature alters the conformation of the protein allergen. Patients who were 3 months and could tolerate the baked milk had a small wheal size but higher IgG4 levels against casein. The sample size was large thus increase in accuracy due to more reproducibility. Confounding factors such as growth and intestinal permeability were statistically insignificant.
Egg allergy, unlike CMPA, has a prevalence of 1.6% in children and is much less in adolescents and adults (Heine et al, 2005). A 1 year cohort study carried out by Pereira et al, 2005 showed that 1% of their adolescents (649) were sensitized to egg affect. This findings were replicated by Woods et al, 2002, 1.6% of their adult population had a positive SPT to egg, although only 0.2% had illness associated with egg ingestion. The sensitivity is mainly to the egg white rather than the yolk (Sicherer, 2006). Studies have shown that children do outgrow egg allergy (Wood et al, 2002). Other studies have also shown that children who are allergic to eggs, or have asthma or both have increased risk of IgE mediated reactions to CMPA (Tariq et al, 2000).
Seafood allergy is thought to be lifelong and has a reported prevalence, from a cohort study, of 1.3% for fish and 0.5% for shellfish in 11-15 year old in UK (Pereire et al, 2005). A cohort study carried out in 3-year old Danish children showed no allergic reactions to cod or shrimps on exposure. Their parents however had allergic reactions of 0.2% to cod and 0.3% to shrimp (Osterballe et al, 2005). Germany showed a prevalence of 1.2% to crab while Portugal 11.5% to fish and USA had 0.4% to fish and 2% to shellfish (Sicherer et al, 2006).
Fruit and vegetables have been ranked as the fourth allergen (Steinke et al, 2007). This was mainly observed in those sensitised to birch pollen. A survey carried out in UK showed that individuals with hayfever (20-25%) are affected by birch pollen and the prevalence in UK is 26% (Bauchau et al, 2004). A similar study carried out in Germany showed a prevalence of 21% to apple, 19.3% to raw potato, and 13.9% to carrot. Another study carried out in five EU countries showed that apple had 56% prevalence in Sweden and 25% in Lithunia and overall it was the second most common cause of allergy while strawberry had a prevalence of 43% of Russians and 21% of Lithunians. Kiwi and peach allergy have been observed to increase in frequency (Lucas et al, 2003).
CMPA is the most common allergy in children and a study done in Spain showed that the prevalence of AAR is 40% and could be more since only the previous year accidents were analyzed in the study. This was relatively high and their findings can also be replicated in a similar UK study. The hypothesis is that school is the main place where AAR takes place (citation from discussion) and severity of reactions is dependent on risk factors such as asthma, recurrent wheeze, Hayfever and eczema, other food allergens such as egg, wheat, fish, shellfish, soya, nuts and sesame and moreover, current SPT wheal size (mm) (Fiocchi et al, 2010).
CHAPTER 2: LITERATURE REVIEW
2.1 Cow's milk allergy
Cow's milk protein allergy (CMPA) has been the most common allergy in children. It has always been confused for lactose intolerance (LI) due to similarity of symptoms though LI is a non-immune mediated response. The allergic reactions can be IgE or non-IgE mediated (Figure 1). IgE mediated allergy accounts to 60% of cow's milk protein allergy and 35% also have other food allergies (Gerth et al, 2003). The prevalence of the milk protein allergy is 1.9% to 3.2%, while the persistence is 70-90%.
2.1.1 Milk proteins
Casein is 80% of the milk while the rest is whey proteins (Table 2) and both have been shown to have allergenic responses. Milk industrial processing (Pasteurization, ultra-high-temperature and dry blending) effect on allergenicity is absent (Nowak-Wegrzyn et al, 2009). New research has shown that persistent CMPA is due to binding epitopes. Some IgE binding regions on Î±s1-, Î²-, and k-casein were recognized only by older children but not those less than 3 years (Heine et al, 2002). Some epitopes of Î²-lactoglobulin were also associated with persistent CMPA. These have been used as potential markers to identify patients who will have CMPA in future.
Most patients who have CMPA also avoid milk from sheep, goat, buffalo, horse, donkey and camel since cross reactivity is 90% to goats and sheep milk and 4% to mare and camel milk (Saarinen et al, 2005). The Casein b-lactoglobulin has a similar structure in all these mammals because they are pyhlogenetically related. The amino acid sequence (epitope), which is similar among the mammals, is the major determinant for immunogenic activity (Restani et al, 2002).
Table 2: Proteins present in milk (Busse et al, 2002)
Cow's milk (mg/ml)
Cow's milk (mg/ml)
2.1.1 CMPA immune response
As shown in Figure 1, CMPA can be IgE, non IgE or mixed IgE and non IgE mediated. The IgE mediated response is fast due to IgE presentation of food allergen such as casein, lactoglobulin and lactalbumin in milk, to high affinity receptors of mast cells and basophils thus leading to release of highly active mediators of inflammation including histamines, proteolytic enzymes, leukotrienes and prostalglandins (Strobel, 2011). The non IgE mediated response is however delated and will take time to appear and the symptoms are mainly gastrointestinal (Mirna and Chehade, 2007).
Figure 1: Cow's milk allergy definition (FAI, 2009)
IgE mediated allergy occurs in two stages sensitization and activation. Sensitization entails food allergens presentation to T-cells of the immune system that produce cytokines (IL-4) and they also interact with B-cells important for antibody production. The IL-4 stimulates the B cell to produce IgE antibodies which circulate in the blood and bind to their receptors on the surfaces of mast cells and basophils causing activation. These activated cells undergo degranulation and release histmines and cytokines into the surrounding tissues causing several systemic effects such as nerve stimulation and this leads to rhinorrhea, itchiness and anaphylaxis. The localization of symptoms on exposure to allergen varies from one individual to individual. The most common occur in the respiratory system (rhinoconjunctivitis and asthma), and the gastrointestinal tract (nausea, vomiting, and diarrhea) (Walport et al, 2001).
Late response is collectively known as non IgE mediated allergy and it entails Type II (cytotoxic reactions), Type III (Arthus-type reactions) and Type IV (delayed T-cell reactions). This occurs due to persistence of cytokines from mast cells and migration of lymphocytes such as lymphocytes and macrophages to initial site of action (Holt and Sly, 2007). The mast cells have also been labelled as 'Yin Yang modulators of intestinal tract and airways due to their role of sensitization and down-regulation of post-inflammatory response apart from their pro-inflammatory role after IgE sensitization (Kraneveld et al, 2012). Enterocolitis and proctocolitis are frequent clinical presentations (DRACMA, 2010). Mixed IgE and non-IgE response, may present as atopic dermatitis (AD) or as one of the eosinophilic gastroenteropathies. The symptoms are mainly gastro-intestinal (50-60%), skin such as hives (50-60%) and respiratory (20-30%) (Vandenplas et al, 2007). Other symptoms are as shown in the table below (Table 3).
Table 3: Cows milkprotein-induced immune reactions (Lee et al, 2006, Moissidis et al, 2005)-skypala pg 119
Ananphylaxis- vomiting, pain and/or diarrhoea
Urticaria, angio-oedema, pruritis, morbiliform rashes andflushing
Acute rhinoconjuctivitis, wheezing, coughing and stridor
Non IgE mediated
Food-protein induced enterocolitis (proctocolitis or enteropathy syndrome)
Pulmonary haemosiderosis (Heiner's syndrome)-rare or; bleeding in the lower respiratory tract
Mixed IgE and non IgE mediated
Eosinophillic oesophagitis, colitis and or proctocolitis
Excessive mucous production; constipation, intestinal colic; Heiner's syndrome
2.1.3 CMPA prevalence
The DRACMA guidelines showed that there was inadequate information on global trends especially from Australia, Italy and North America where many referral studies come from thus inconclusive results. The most common parts were Spain, UK, Scandinavian countries and Germany. A randomized telephone survey carried out in Germany, Belgium Austria, Greece, Italy, Poland, Switzerland, Denmark, Finland and Slovenia showed that milk had the highest prevalence of 38.5% as compared to other food allergens (Steinke et al, 2007). Fruits (29.5%), eggs (19.0%) and vegetables (13.5%) followed after milk. The sample size was large thus reducing result and measurement bias and the fact that more than one country was involved in the survey increased the reproducibility of the results.
Children with non IgE mediated reactions recover earlier than those with IgE mediated. A cohort study carried out showed that children with egg allergy had increased risk of asthma and IgE mediated reaction to cow's milk (Host et al, 2002). Other adverse reaction include CMP-induced proctocolitis which is a disease of infants and presents with bloody stool though the child is well and thriving; Food protein-induced enterocolotis syndrome (FPIES) which is a severe, cell-mediated, gastrointestinal food hypersensitivity; allergic eisinophilic gastroenteropathies which presents as gut inflammation; food-sensitive eczema which is chronic inflammatory skin disorder with raised serum IgE and; CMP-induced oesophageal pathology which is a replux disease (Fox et al, 2007). A retrospective review carried out by Skripak et al, 2007 showed that 19% had resolved allergy by the age of 4; 42% by 8 years and 79% by 16 years of age. Coexisting allergy and hayfever were also significant predictors of outcome (p < 0.001). Woods et al, 2002 showed that 0.7% of adults have positive skin prick test.
Exclusively breastfed children will have lower CMPA incidence as compared to formula fed or mixed infants (Fox et al, 2007). This is due to the fact that the infants' stomach is not well developed and introduction of lactoglobulin may be viewed as an allergen by the immune system leading to the reactions. Despite difference in gut flora, breastfed children will show less severe symptoms. Infants presenting symptoms such as angio-oedema, urticaria and immediate vomiting have IgE mediated allergy (Vandenplas et al, 2007). Atopic dermatitis and failure to thrive can be observed but on occasional cases as compared to formula and mixed-fed infants (Vandenplas et al, 2007). Patients who are formula-fed can have severe respiratory reactions or anaphylaxis and thus need to be referred to an emergency department for treatment (Vandenplas et al, 2007). Early administrations of adrenaline through intramuscular route and anti-histamine have been shown to prevent anaphylaxis and to improve symptoms in children (Fox et al, 2007). Diagnosis under such conditions entails elimination of all forms of milk products and traces. This is also due to the fact that the reactions are not reproducible and they vary with food containing milk (Fox et al, 2007).
1.5 Grading of Recommendations, Assessments, Development, and Evaluation (GRADE) assessment for Diagnosis
Oral food challenge (OFC) is the reference standard for the CMPA diagnosis used to test for different allergens which are administered orally in form of a pill and the outcome (signs and symptoms) is monitored (Fox et al, 2007). The problem with OFC, it is tedious, resource intensive and patient has a high risk of anaphylaxis. Due to this in most parts of the world it is not utilized (DRACMA, 2010). In the UK, analysis is done as a follow up treatment though caution is exercised and this is after analysis of AAR and decrease in wheal size or sIgE.
The National Institutes of Health Guidelines for the Diagnosis and Management of Asthma and Guidelines for the Diagnosis and Management of Food Allergy support the use of in vivo (skin prick) or in vitro (blood) specific IgE testing. This is done along with a detailed clinical history and physical examination. The initial diagnosis is carried out by the parents or carers through observation of symptoms at home (Kwong et al, 2011). In vitro testing provides a valuable diagnostic tool for comprehensive allergy and asthma management through education, targeted allergen avoidance, pharmacotherapy, and immunotherapy (Kwong et al, 2011). Amino acid based formulas and hydrolysed formulas have been used to substitute cow's milk in allergic infants (Figure 2).
Figure 2: Cow's milk management (FAI, 2009)
A tiny board with protruding needles containing allergens is used for skin prick test (SPT). It is placed directly on the skin, for puncture and allergen entry into the body. A positive allergy test is if a hive appears (Figure 3). This test only works for IgE antibodies (Freeman, 2008).
Figure 3: Skin prick test (Ihloff, 2008)
RadioAllergoSorbent Test (RAST) is used to detect specific IgE antibodies (sIgE). The results are always compared to the standard and if higher, then one is allergic. The main advantage of the test is that more than one allergen can be tested at the same time unlike blood testing which does not work for all allergens (Freeman, 2008).
1.6 Treatment and management
Elimination diet is the main mode of treatment due to the uncertainty of allergic reactions (Fox et al, 2007). The disadvantages are that the child may have impaired growth due to lack of calcium (~115mg/100ml in cow's milk), iron, B vitamins, essential fatty acids and energy. Rice milk and other form of supplements have reduced these kinds of conditions by ensuring that the child receives the required diet. Other types of milk formula that do not contain cow's milk and Soya formula are also recommended though this will depend if the child has other food allergies and would react so this is administered and monitoring is done by the dietitians to ensure the child has no milk reactions.
Cross reactivity has also been observed in some CMPA children. Few also react to soy though there are formulas in which the soy and milk soy proteins are degraded so the immune system does not recognize the allergen. Hypoallergenic infant formulas are composed of less antigenic partially pre-digested proteins. Other formulas based on free amino acid are also less antigenic and provide wholesome nutrition support in severe CMPA (Table 4).
Table 4: Different types of formulas used by allergic children (Skypala et al, 2009)
Type of formula
Extensively hydrolyzed whey
Whey empties the stomach faster and thus beneficial for patients with gastrointestinal symptoms. It has higher peptide percentage (>1.5kD) and the lactose is important to increase taste
Pepti (Cow & Gate)- contains lactose
Pepti junior (Cow & Gate)- has triglycerides as fat source
Extensively hydrolyzed casein
Mostly protein (<1.5kD). lactose free and well tolerated by IgE mediated allergic children
Nutramigen (<6 months)-1 and (>6 months)-2
Soy formula (non-hydrolyzed protein)
Usually given to those who refuse hydrolyzed formulas and suitable for vegans. It is however not administered to children < 6 months of age and it is not the first choice
Farley's soy formula (Heinz)
Non-milk based extensively hydrolyzed formula
It is suitable for malabsorption disorders and not for vegetarians and vegans due to its meat derivatives. More so, not for CMPA treatment due to poor taste
Pepdite (SHS)-<1 year and 1+ (SHS)->1 year
Elemental amino-based formula
Suitable for multiple food allergies and when other formulas are unsuitable; faltering growth, severe eczema, enterocolitis and proctitis
Neocate (SHS)-<1 year, active (SHS)- 1-10 years- has a better taste, more calcium and iron and advance (SHS)-1-10 years
People who are extremely sensitive should avoid contact with the allergen, including touching or inhaling allergen, as well as touching any surface that may have been exposed. Epinephrine, antihistamines and steroids have been used as emergency treatment when AARs occur. Their administration depends on severity of reaction and doctor's prescription. The mode of action is shown in Table 5.
Table 5: Treatments and their mode of action
Mode of action
It is used to treat severe allergic reactions such as anaphylaxis whereby it tightens blood vessels leading to increased heart beat and circulation hence lower allergic reaction. It is prescribed as an Injectable known as Epipen.
Figure 4: Epipen injector (Sean William, 2006)
They alter histamine action of blood vessels dilation leading to leakage of plasma proteins thus itchiness due to action at the sensory nerve terminals. The most commonly used is Benedryl. They however, do not completely improve the breathing symptoms of anaphylaxis (Sicherer, 2006)
They are usually in the form of a nasal spray and act by calming down immune cells attacked by the products of an allergic reaction. They are not used for anaphylaxis since symptoms are relieved only in the area of steroid contact if injected and they also take long to reduce inflammation. Steroids taken orally or through injection, reach every part of the body though still take longer (Sicherer, 2006).
1.6.1 Allergen immunotherapy
Immunotherapy entails gradual increase of exposure to allergen extract doses to patient with the aim of reducing an immune response. It is also known as allergy vaccination or desensitization and can be subcutaneously or sublingually administered. There are two types; injection and oral immunotherapy (Table 6). Studies have shown that it reduces Hayfever and seasonal asthma for it alters how the immune system reacts to allergen such as grass or tree pollen and the benefits continue 3 years (Buchanan et al, 2006).
In animal studies, tolerance has been shown through apoptosis after exposure to allergen though in humans it is still risky due to the uncertainty of outcome thus not a common mode of treatment ( DRACMA, 2010). Passalacqua et al, 2012 showed in a reveiew that oral immunotherapy is an attractive form of treatment for CMPA since some trials showed high percentage remission in children and <20% discontinued to severe counter effects. Brozek et al, 2012 however showed through a systematic review and meta-analysis that the 5 observational studies replicated the findings of the 5 randomized control trials included in the study though the evidence was of low quality leading to inconclusive results on the potential of oral immunotherapy on CMPA.
Table 6: Desensitization and how it is carried out (Fiocchi et al, 2010).
How it is carried out
Giving a series of injections to desensitize the patient. It has been shown to be most effective for severe cases of hayfever. It is however, not suitable for asthmatics due to severity of reaction
This is given without injections but orally either as tablets or drops. This however is rarely done and the dose treatment is practised with caution in hospitals. It can be used in asthmatics so long as the asthma is well controlled
The remission rate is high with increase in age though the symptoms disappear after the child undergoes an elimination diet and have other formulas without the allergens (Table 7). Prolonged elimination can affect the growth and development of the child due to nutrition aspects such as calcium essential for bones and other minerals such as riboflavin thus the need for regular assessment of tolerance using the diagnosis methods mentioned above (SPT, OFC and sIgE). OFC should not be carried out due to reduction of SPT or sIgE but on the length of elimination diet and reaction to accidental ingestion of CMP in foods.
Vanto et al, 2004 carried out 4-year follow up study on children with immediate and delayed CMPA and noted that those with <5mm wheal size had developed tolerance to milk and those who IgE > 2kUA/L were still persistant. Vassilopoulou et al, 2008 also showed that those who had <4mm wheal size (sIgE <3.94kUA/L) would have a negative challenge while >7.5mm (sIgE >25.4kUA/L) would have a positive challenge. In this study, >3mm wheal size was utilized as an inclusion criteria as it is used in most literature (DRACMA).
Table 7: Remission rate with increase in age
1.7 PROJECT JUSTIFICATION
Cow's milk is the 2nd major food allergy experienced in children after Peanut allergy (Skypala and Venter, 2012). Elimination diet is the only mode of treatment in order to achieve remission. This is however difficult since traces of milk can be found in most food commodities such as biscuits, sauce mixes and processed meats thus the increased risk of encountering them. Accidental allergic reaction to cow's milk can occur anywhere and are brought about by contamination, skin contact, misread labels or new recipes in products they are accustomed to and non-packaged foods. The symptoms can be severe, mild and moderate. A cross-sectional study carried out in Spain by Boyano-Martinez et al, 2009, showed that AAR frequency was 40% which was high since only reactions in the past year had been evaluated. The frequency was higher than the peanut AAR frequency of 14.3% from Canada (Yu et al, 2006) and egg allergy AAR frequency of 21% in Spain (Boyano-Martinez et al, 2012). Their sample sizes were large thus increased reproducibility and accuracy. Since the products were different and the locations were also different, there was need to analyze AAR frequency of a similar product in a different country in order to assess the reproducibility of results depending on geographical distribution.
Pereira et al, 2005 showed that 3% of 11 year-old and 4% of 15 year-olds had adverse reactions to milk ingestion, most of which were immediate and severity was affected by history of asthma and recurrent wheeze. The sample size was large thus increased reproducibility (n=757 and n=775 for 11 and 15 year-olds respectively), there was no recall bias since it was a follow-up study though there may be response bias due to lack of information from the nonrespondents. The study results replicated the findings from a prospective survey carried out by Macdougall et al, 2002, who showed that coexisting asthma was associated to AAR severity (p-value <0.01) and CMPA anaphylaxis had 10% rate for paediatric hospital admissions and in the previous 10 years prior to the study, 8 children had died 4 of which had CMPA. This led to the objective of determining AAR severity and the risk factors associated with it.
A recent 3-year follow-up study carried out in the 5 USA states (New York, Denver, Durham, Little Rock, Baltimore), on 512 children showed that the annual rate was 81% for all the allergies (Peanut, egg, milk and other). Milk had the highest AAR annual frequency (71%) of 34% which was relatively similar to the Spain CMPA study frequency of 40% (Boyano-Martinez et al, 2009). They showed that a third was due to foods provided by parents and the rest by caregivers (grandparents, teachers and babysitters). 90% of the reactions were due to label-errors (16%), cross-contamination (15%), reduced caregiver supervision and forgetfulness (65%) and mistakes in food preparation (4%). The parents had received advice on food avoidance and given advice on emergency treatment using Epipen. On follow-up, 10% of the reactions were severe but only a third were treated with epinephrine. They speculated that this may be due to the fear of using Epipen injections. They also showed that 1 in 9 of the AAR were intentional and Dr. Fleischer explained in ABC news that this may have been done to test if they had outgrown their allergy though follow-up would be done to verify the reason (Albin, 2012). The sample size was large, the methodology was explicit; a 3 year follow-up of children who were 3-15 months with documented milk or egg allergy. The inclusion of 5 states increased the reproducibility of the results which were well interpreted. The limitations were however measurement error especially with the intentional AARs though follow-up would be done to replicate the discovery. The study was long and expensive since it involved more than one state thus possibilities of study drift. This led to the objective of the place where AAR occurred, type of food exposed to and mode of treatment thus aid in health care management. Our inclusion criteria was however different since the children had to be more than 18 months. The sample selection was carried out in the Allergy Paediatrics clinic at St. Thomas hospital, London.
To establish the frequency of the accidental allergic reactions in children with cow's milk protein allergy in the past year.
To establish the circumstances (food type, place) and severity of the reactions and treatments used.
To identify risk factors that have potential to cause severe reactions.
To determine if the age at which they had their atopic diseases had an influence on AAR severity.
CHAPTER 3: METHODOLOGY
This was project No: 2807, St. Thomas Hospital, London.
The hypothesis stated that the AAR severity was dependent on the type and number of atopic diseases and most reactions occurred mainly in school (Sicherer et al, 2001). The chief questions were AAR frequency, the type of food that caused most of the AARs and the place or circumstance so as get to a solution on how the healthcare team can tackle the problem and lower AAR risk.
This prospective study was carried out as an audit and the methodology is an account of the rationale of subject selection for the study, ethical consideration and data collection through semi-structured interviews. This was in accordance to the cross-sectional study carried out in Spain by Boyano-Martinez et al, 2008.
3.2 Ethic consideration
This entails the morals, rules and forms of conduct which should be applied when conducting research. The Research and development department at St. Thomas hospital cleared the project as an audit and thus said there was no need for ethics application. I however, had to apply for a Criminal Records Bureau (CRB) certificate as a requirement to carry out the study. The rules and regulations on collecting and handling data were adhered to strictly and the patient confidentiality was maintained (no patient information left the clinic).
3.3 Subject selection
The survey was carried out at the children allergy unit at St. Thomas hospital, London. The patients had to be 18 months and over with a clear CMPA history (this was done by explaining the type of symptoms they experience and how long they took to appear) and avoided all form of cow's milk (baked, traces in foods) (Table 8).
Table 8: The inclusion and exclusion criteria
Patient inclusion criteria
Attending Paediatric Allergy clinic
Over 18 months of age (therefore weaned over 1 year previously)
Diagnosis over IgE mediated cow's milk protein allergy
Positive SPT (> 3mm wheal size)and/or Positive Specific IgE
Clear History for cow's milk allergy (see correspondence section for clinic letter)
Avoids all cow's milk and cow's milk containing foods in the diet
Infants < 18 months of age
Non IgE mediated cow's milk protein allergy
Patients without a clear history of cow's milk allergy
Patients tolerating some cow's in their diet
Patients with history of eczema and positive SPT but no other clear history of cow's milk allergy
The patient selection took place from 25th May till 14th July 2012. The age preferred was >18 months because they had more than one year of weaning and introduction of different kinds of food thus increased risk of allergen intake.
3.4 Identifying suitable patients
As the student in charge of the study, I had to arrive early and look through the files' notes for patients attending that day's clinic who qualified to be part of the study. Then record their current age and SPT at the time of diagnosis which was found in the first clinical letter. A clear history of CMPA which entailed reaction on milk ingestion and avoidance of all traces of milk in products were also recorded. The time at which they arrived and the doctor to be seen was noted. When they arrived, I introduced myself and the study and; ensured that the parents were content to participate in the study and complete the questionnaire (Appendix 1) through a semi-structured interview.
3.4 Study design
A semi-structured interview using a close- ended standardized questionnaire on accidental allergic reactions experienced in the past year was carried out by me when the parents would bring their children to the clinic for follow-up. The questionnaire was completed by the student. The questions were as follows:
The type of food that caused the reaction, this was classified as milk in products, direct contact (taking milk) or contamination for example pizza cut with knife used to cut cheese. Type of exposure (skin contact, inhaled, food eaten- trace, mouthful, full portion), place of reaction (home, friend's house, nursery/childcare, school, restaurant/cafe), symptoms of reactions, time taken for onset (<15, 15-30, 30-60 and 60-120 minutes), treatment required (none, ant-histamine e.g. piriton, adrenaline autoinjector e.g. inhaler) and progression of disease (spontaneous without/with treatment, within a few hours with treatment-<1, 1-6, 7-24 hours; A&E, 999, G.P.). The symptoms were IgE mediated thus immediate hypersensitivity (<24 hours). The severity of the allergic reactions was based on the symptoms as mild, moderate and severe (Table 9).
All the interviews were carried out face-face and the method of interview utilized allowed alteration of questions. To avoid bias, the questions in the questionnaire had closed questions and some answers were obtained directly from the history documents and no leading questions were asked.
Table 9: Symptoms of accidental allergic reactions
Urticaria/Angioedema (hives or swelling of skin), Rash , Conjunctivitis
Vomiting, Abdominal pain, Diarrhoea
Wheezing, Cough, Loss of voice, Rhinitis
Loss of consciousness, Weakness, Dizziness
The aim was to establish the accidental allergic reactions and the severity of symptoms brought about by other types of food allergies or atopic diseases and would help in health management and how to create awareness to lower the accidental incidences which in some children can lead to death.
The reactions, symptoms, drug utilized and prognosis were coded (Table 10). This would make it easier for parent to answer though the student took them through in order to ensure that all questions were answered and explanation of questions not understood. The time of onset was also coded to prevent entry of symptoms which took more than 120 minutes since they would be non-IgE mediated leading to inaccurate results.
Table 10a: The codes used to identify type of exposure, place of reaction, time of AAR reaction onset, treatment, prognosis and the symptom (below).
0= Not applicable
1= Skin contact
3= Food eaten
6= Full portion
Place of reaction
2= Friends house
0= Not applicable
1= Spontaneous without treatment
2= Spontaneous with treatment
3= with treatment <1hr
4= with treatment 1-6 hrs
5= with treatment 7-24 hrs
6= without treatment <1hr
7= without treatment 1-6 hrs
8= without treatment 7-24 hrs
0= Not applicable
1= No treatment required
2= Anti-histamine e.g. Piriton
3= Adrenaline autoinjector e.g. Epipen
Time from intake to onset
1= <15 minutes
2= 15-30 mins
3= 30-60 mins
Table 10b: Symptoms
1= Urticaria/Angioedema (hives or swelling of skin)
5= Abdominal pain
9= Loss of consciousness
11= Loss of voice
3.5 Arrangements to ensure anonymity and confidentiality
The patients were identified according to numbers and codes used to identify their names. This was done in order to prevent reassessing the patients if they had more than one appointment within the month of data collection and in case of data omission. All the collected information that had patients' names was left in the hospital and access was restricted to the supervisor and me. The questionnaires would later be destroyed after publication. The data obtained was inputted into Ms. Excel and later transferred to SPSS for result analysis. The patients were recorded as numbers thus to prevent patient information getting out of hospital and more so, to prevent bias while analyzing results and increase accuracy
3.6 Skin prick test determination
The SPT wheal size for both previous (first diagnosis) and current test carried out during follow up when the patient came to the clinic for their appointments, were recorded. The wheal size at the time of diagnosis was sued as the inclusion criteria
3.7 Statistical analysis
The statistical calculations were carried out using SPSS 18.0. The Descriptive statistics were utilized to obtain the frequency of the type of food, place of reaction, treatment used and prognosis. The mean and standard deviation of all the continuous variables were also obtained through Descriptive statistics. Chi-square (X2) and Fisher's exact test were used to analyze the relationship between the qualitative variables and AAR severity. Median values were used instead of the mean; for age, current wheal size, total food allergies and atopic diseases due to abnormal distribution.
Figure 4: Chart showing the recruitment and assessment of AAR
Total number of patients (n= 73)
DNAs (Did Not Attend) n=11
Skin prick test (n= 69)
Positive SPT but could have baked milk (n= 3)
Current wheal size <3mm (n=4)
Food specific IgE (n=12)
Accidental exposures within the past year (n= 32)
Type of food; direct (n= 8), In products (n=18), And contamination (n=6)
Place of reaction (home=22, nursery and school=10, restaurant=4 and friends house=1)
Symptoms (none= 20), (mild=13), (moderate = 7) and (severe= 12)
Treatment; antihistamine (n=26), adrenaline (n= 2), inhaler (n= 3)
Prognosis (spontaneous with treatment=3; without 3, with treatment <10, 1-6 hours=14, 7-24 hours 2 and hospital=3
Other food allergies and risk factors such as asthma, hayfever, recurrent wheeze, eczema
CHAPTER 4: RESULTS
The total number of study participants was 62 who attended St. Thomas allergy Paediatric clinic from 25th May to 14th July 2012. 35 (74Â± 44.9) were male and 27 (69Â± 40.9) female. The mean age was 71Â± 42.9 and the median age was 66.5 (range, 18-225 months)  . There were a total of 73 patients who qualified to be in the study but they 10 did not attend their clinical appointments and language barrier for one patient. 13 did not fit into the inclusion criteria due to 3 could have baked milk, 4 had a wheal size <3mm during follow up, 4 had non IgE CMPA, 2 were less than 18 months. The mean age for the 11 nonparticipants' was 43Â± 15.2 (males 41Â± 12.8; females 49Â± 22.5). Majority of the children were in school and a small number were in childcare the rest were still at home. Egg allergy was quite high for the participants (79%) followed by eczema (66%). All the children had other food allergies (Table 11).
57 (92%) participants had other food allergies (egg 49; wheat 15; fish 19; shellfish 11; soya 12; peanuts 43; treenuts 40 and sesame 21). The average number of food allergies per individual was 4Â±2. 54 (87%) children had clinical histories (asthma 27; recurrent wheeze 4; hayfever 20; and eczema 41). The average history per individual was 2 Â±1. All the nonparticipants had other food allergies (egg 10; wheat 4; fish 6; shellfish 5; soya 4; peanuts 7; treenuts 7 and sesame 3) and history of atopy (asthma 4; recurrent wheeze 2; hayfever 1; and eczema 6). The average food allergy was 4Â±2 while for history was 1Â±1 (Table 11).
Table 11: The mean age (months), food allergies, history of atopy, egg, wheat, fish, shellfish, soya, peanuts, tree nuts, Sesame. Asthma, Recurrent wheeze, Hayfever and Eczema for the participants (n=62) and nonparticipants (n=11) 
Non participants (n=11)
Non participants (n=11)
Mean Age (months)
All the children had had AARs before diagnosis at the time of diagnosis thus advised to follow an elimination diet for the particular food allergens. Some patients had some AARs despite the strict exclusion diet but not in the previous year. 35 (56.5%) children reported AARs in the past year. 23 (65.7%) children had 1 AAR, 10 (28.6%) had 2 AARs and 2 (5.7%) had 4 AARs. This led to a total of 51 allergies in the past year (Table 12).
Table 12: Cross tabulation for the number of accidents in relation to AAR severity
Number of accidents
4.1 Type of food and exposure causing AAR
33(89%) AARs were caused by food (mouthful 21, trace 8 and full portion 8), 3 (8.1%) were due to skin contact and 1 (2.7) inhalation.
Figure 5: 19 (51.4%) children had products containing milk as an ingredient were the major causatives of AAR. 12 (32.4%) had took cow's milk and 6 (16.2%) had cross-contamination in restaurants and at home
4.2 Place of reaction
Figure 6: 22 (59.5%) children had their AARs at home under normal daily routines, 5 (13.5%) in the nursery/childcare, 5 (13.5%) school, 4 (10.8%) in a restaurant/cafe and 1 (1.6%) in a friend's house
15 children below the median age and 7 children above the median had their AAR at home; 1 under the median in a friend's house, 3 under the median and 2 above the median in a nursery, 4 above the median in school and; 2 below the median and 2 above the median in a restaurant (Table 13). There was no statistical significance between the age and place of reaction showing that the place of reaction was independent of age (x 2= 7.312, df= 4, p-value= 0.12).
Table 13: Cross tabulation for place of reaction to median age (months)
Median Age (months)
Place of reaction
4.3 AAR Symptoms and their time of onset
The cutaneous symptoms observed were 29 (22 hives and rash 5). The respiratory symptoms were 10 (loss of voice 5, cough 3, wheeze 2) while gastrointestinal were 7 (vomiting 4, diarrhoea 2, abdominal pain 1). Since most children had more than one symptom it was easier to categorize them into mild moderate and severe (Table 10). 14 (40%) children had mild symptoms, 7(20%) had moderate and 14 (40%) had severe reactions.
9 children below the median age and 17 above the median had their symptoms in less than 15 minutes while 21 below the median and 15 above the median had their reactions in more than 15 minutes (Table 14). The time of onset was presumed to depend on age. There was however no statistical significance thus no difference between those who were below or above the median age in relation to the time of onset (p-value >0.05).
Table 14: Crosstabulation for time taken for symptom onset for children above and below the median age
Median Age (months)
Figure 7: Most children took a mouthful of the milk allergen in the food and also through direct contact thus the high number and most had mild reactions and this may be due to protein breakdown in the stomach thus less antigen presentation although some patients still had severe symptoms. Some of the patients, who took traces of the milk allergen, had severe reactions showing that their memory cells are quite high (x 2= 11.17, df= 1, p-value= 0.004).
4.4 Treatment and prognosis
Treatment included all the children (AAR and no AAR) and 49 (79%) children used antihistamines Piriton as their emergency treatment, 8 (13%) used adrenaline (Epipen) and 5 (8%) used inhaler (Table 14). 59 (95%) children were treated at home and only 3 (5%) were rushed to hospital due to anaphylaxis. 3 (9%) children who had AARs did not need treatment. 3 (9%) subsided spontaneously with treatment, 10 (29%) within an hour, 14 (40%) in 1-6 hours and 2 (6%) within 7-24 hours.
Table 15: AAR treatment and prognosis
Spontaneous without treat
Spontaneous with treat
Few hrs with treat <1
Few hrs with treat 1-6
Few hrs with treat 7-24
4.5 Relationship between AAR severity and risk factors
For the current wheal size, the frequency of moderate to severe was 48% for those above the median  and it was twice those who were below the median (21%). There was statistical significance between those with no AAR and mild to those who had moderate and severe symptoms in relation to the current wheal size taken during follow up (x 2= 5.047, df= 1, p-value= 0.024). This was also the case for Hayfever where there was statistical significance between the no AAR and mild symptoms and; moderate to severe in relation to those who had Hayfever and those who didn't (x 2= 5.885, df= 1, p-value= 0.015). The time of onset also had a statistical significance (x 2= 10.316, df= 1, p-value= 0.001) showing that those who reacted in less than 15 minutes (62%) were more likely to have severe symptoms than those who reacted after 15 minutes (38%) (Table 16)-put table.
Table 16: The relationship between AAR severity and risk factors
No AAR (A)
Moderate or severe (B2)
(A+ B1vs B2)
Time for symptom onset
<15 minutes (n=)
>15 minutes (n=)
Current wheal size (mm)
Atopic diseases (total)
a=x2 test b= Fisher exact test
There was however no statistical significance for Gender, Age, Asthma, Recurrent wheeze and Eczema. This showed that there was no difference between the categories in relation to the severity of the symptoms and thus they were not risk factors to AAR severity. There was statistical significance between those with no AAR and mild to those who had moderate and severe symptoms in relation to those above and below the median(x 2= 5.047, df= 1, p-value= 0.024). The frequency of those above the median was 48% and those below was 21% (Table 16). There was no statistical difference between the age at which they got the various atopic diseases and AAR severity.
Table 17: The relationship between AAR severity and risk factors
No AAR (A)
Moderate or severe (B2)
(A+ B1vs B2)
Food allergies (total)
a=x2 test b= Fisher exact test
The frequency of moderate to severe was 42% for those who had Peanut allergy while those who didn't was 16%. There was statistical significance between those with no AAR and mild to those who had moderate and severe symptoms in relation to those who had and those who didn't have Peanut allergy (x 2= 3.999, df= 1, p-value= 0.04). There was however no statistical significance for Egg, Wheat, Fish, Shellfish, Soya, Treenuts, Sesame and total number of food allergies (p-value>0.05). In all the categories, there was statistical significance between AAR and no AAR. This showed that there was no difference between the categories in relation to the severity of the symptoms and thus they were not risk factors to AAR severity (Table 17).
Figure 8: (a) in the 1-4 years age group, 8 children had no AAR, 10 had one, 4 had 2 and 2 had 4 AARs. In the 5-7 years age group, 7 had no AAR, 8 had one and 3 had 2. And in the 7 and above age group, 10 had no AAR, 7 had only one and 3 had 2 AARs. There was however no statistical significance to number of accidents in relation to age (x 2= 4.182, df= 6, p-value= 0.652). (b) 9 children below the median had no reaction; 14 had one, 5 had 2 and 2 had 4 while those above the median; 16 had no AAR, 11 had one and 5 had 2 AARs. There was no statistical significance between AAR numbers and age, thus no difference between those above and below the median (x 2= 4.26, df= 3, p-value= 0.235).
The AAR frequency was 59.7%. This outcome was higher than a similar CMPA AAR cross sectional study carried out in Spain by Boyano-Martinez et al, 2009 frequency of 40%. The sample size for this study was smaller (n= 62) as compared to Boyano-Martinez et al study (n= 88) though the milk consumption in the UK may be higher than Spain. The advantages of a cross-sectional study are that it is fast and cheap. The limitations are however that it did not provide evidence of temporal association and their tables was hard to interpret though the explanation was explicit. Other studies showing different food allergies have given lower values thus giving more light to the high milk consumption in foods and the different types of dairy products may have led to the high frequency. Boyano-martinez et al, 2012, also carried out a similar study on the egg allergy AAR frequency and it was 21%. The sample size (n=92) was large thus increased accuracy (n=92). Yu et al, 2006 also showed that peanut allergy AAR frequency was 14.3%. This was lower than our results because the annual rate was obtained from calculation of accidental exposures since birth to their current age. This was low due to safer environments, strict food legislation and school awareness campaigns (Yu et al, 2006). The sample size was large thus the reduced response bias. In North America and Europe the prevalence was approximately 50%. The rates are believed to have gone down in Europe due to the implementation of EU food legislative regulations (Directive 2001/95/EC and Regulation 2002/178/EC). The high rates were due to high levels of peanut traces and hazelnuts in cookies and chocolates and; lack of proper food labelling (Pele et al, 2007).
We excluded symptoms that took more than 4 hours to appear since that was a non-IgE mediated response. Some patients reported that they would also react to meat and this is due to presence of some of the protein allergens present in milk also being found in dairy meat. Piriton (79%) was the main mode of treatment though in anaphylaxis it has been shown not to be effective in improving breathing symptoms (Sicherer et al, 2006). Only 3 patients had anaphylaxis and they were rushed to hospital for analysis where Epipen was administered. 13% were treated using Epipen at home and 5% using inhaler. The egg allergy had no hospitalizations (Boyano-Martinez et al, 2012 while the peanut allergy study had 52% who sought medical attention in the participants and 32% in the non participants (Yu et al, 2006). 52% were admitted to hospital due to anaphylaxis in the Germany study (Mehl et al, 2005). This was high due to the fact the study was mainly about anaphylaxis reactions.
The most common route of exposure was oral (91%). Only 2 patients had skin contact and 1 who's AAR was caused by inhalation. Mild AARs were 40% of the symptoms observed and this was similar to the severe symptoms (40%) while moderate was 20% of the total. This was lower than the spain CMPA AAR study (53%) and this may be due to the fact that more children with a wheal size above the median had severe symptoms. An egg allergy study carried out in Spain showed that the mild AARs were 42% while the moderate AARs were more frequent (50%). Yu et al, 2006, showed that in Canada the peanut allergy mild AAR prevalence was 31% the results were similar to a study carried out in Germany (21%). In both the severe symptoms were high and this may be due to the nature of the allergens (Mehl et al, 2005) and classification of AAR severity could also vary from one study to another (Boyano-Martinez et al, 2012).
The main place of action was home (35.5%) (Figure 6). This led to rejection of the alternative hypothesis that most reactions occurred in school and replicated the findings from a nationwide survey carried out in Germany by Mehl et al, 2005 who showed that majority of the anaphylactic reactions occurred at home (58%) though was against most reports which stated that the frequency was more in school (Sicherer et al, 2001, Fuhlong et al, 2001).This may be due to siblings at home who can have milk and the children have access to the fridge as compared to school since 71% of the ch