SEARCH

SEARCH BY CITATION

Keywords:

  • allergy;
  • DBPCFC;
  • epidemiology;
  • food;
  • intolerance;
  • nutrition, pseudoallergy, oral provocation

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

Objective:  A population study was performed to identify the prevalence of all kinds of adverse reactions to food.

Methods:  In a representative cross-sectional survey performed in 1999 and 2000 in Berlin, 13 300 inhabitants of all ages were addressed by questionnaire. This questionnaire was answered by 4093 persons. All respondents mentioning any sign of food intolerance or the existence of allergic diseases (n = 2298) were followed up by telephone and, in case food intolerance could not be ruled out by patient history, were invited to attend to the clinic for personal investigation including double-blind, placebo-controlled food challenge tests (DBPCFC).

Results:  The self-reported lifetime prevalence of any adverse reaction to food in the Berlin population (mean age 41 years) was 34.9%. Eight hundred and fourteen individuals were personally investigated according to the guidelines. The point prevalence of adverse reactions to food confirmed by DBPCFC tests in the Berlin population as a mean of all age groups was 3.6% (95% confidence interval [3.0–4.2%]) and 3.7% in the adult population (18–79 years, 95% confidence interval [3.1–4.4.%]). Two and a half percent were IgE-mediated and 1.1% non-IgE-mediated, females were more frequently affected (60.6%). Based on a statistical comparison with available data of adults from the nationwide German Health Survey from 1998, adverse reactions to food in the adult population of Germany (age 18–79) were calculated with 2.6% [2.1–3.2%]).

Conclusions:  The study gives for the first time information about the point prevalence of both immunological and nonimmunological adverse reactions to food and underlines the relevance of this issue in public health. The data also show that an individualized stepwise approach including provocation tests is mandatory to confirm the diagnosis.

Adverse reactions to food have been recognized since ancient times, but only recently the many heterogeneous mechanisms have been identified. These include not only immunological IgE-mediated food allergy and nonimmunological reactions like pseudoallergy, enzyme deficiencies, toxic effects of food, but also rare reactions like food dependent exercise-induced anaphylaxis. The frequency of perceived adverse reactions to food is higher than the number of confirmed reactions (1, 2). However, the diagnosis of adverse reactions to food is very sophisticated and till date there is no large population-based study worldwide investigating all kinds of adverse reactions to food using the required selective patient-oriented approach. Investigation of nonimmunological reactions is difficult because of the lack of simple in vitro and in vivo tests, but even for allergic, IgE-mediated reactions and especially cross-reactions these tests have low predictive value (3–7). The recommended way of diagnosis, therefore, in addition to skin and blood tests, is to establish food intolerance by dietary exclusion and double-blind, placebo-controlled food challenge tests (DBPCFC) (8–10).

Due to this difficult situation and the lack of existing data, the German Ministry of Health initiated this study to investigate the prevalence of food intolerance in the German population and to identify the eliciting agents. The study was performed in the Berlin population, a single-center approach was chosen since an especially equipped allergy centre was required, to handle severe anaphylactic reactions during provocation tests and to avoid an inter-investigator bias. Data were then computed for the whole German adult population on the basis of the existing data of the nationwide German health Survey from 1998, which was performed in the age group of 18–79 years concentrating on general health issues but not on food allergy in particular.

Population

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

It was calculated that a sample size of 4000 persons would be sufficient to report an expected prevalence rate of 2% with a 95% confidence interval smaller than [1.5–2.5]. Further calculating a reply rate of 30–40%, a random sample of 15 000 persons of all districts and age groups with German citizenship was drawn from the Berlin population register. From this random sample, 13 300 persons were selected by repeated random subsampling and contacted by questionnaire.

Questionnaire

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

A questionnaire was designed with the questions adapted to the age groups 0–14 years, 14–18 years and adults and evaluated in 50 patients of the Departments of Dermatology and Pediatrics. Questions were about perceived connections between food ingestion and itching, eczema, urticaria, angioedema, rhinitis, asthma, intestinal symptoms, headache and other symptoms. In addition, questions were asked about the existence of atopic diseases, the existence of the above named symptoms without the perceived relation to food ingestion, possible risk factors such as smoking, household size, status of education. The questions regarding the status of education were taken directly from the German National Health Survey Questionnaire to enable the statistical comparison. Additional questions were related to items important for health care management, e.g. doctor shopping. For lack of space, these results are not included in this report.

Ethic committee consent was granted for all stages of the study as well as the consent of the Berlin Data Protection Commissioner.

Study design

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

The study was performed during the years 1999 and 2000 and 4093 questionnaires were returned and evaluated. All persons were contacted by telephone, who either claimed to have perceived adverse reactions to food or mentioned diseases possibly associated with food, e.g. recurrent diarrhoea, allergic rhinitis or eczema. During the telephone call, a decision was made following an algorithm as to whether food intolerance could probably be ruled out or not. A clear decision to rule out food intolerance was made when food items had given rise to symptoms once only and were afterwards again repeatedly ingested without any complications. In all cases where any suspicions remained for adverse reactions to food, persons were invited for a personal investigation at the clinic. During this investigation from all attending persons (see also flow chart, Fig. 1), a detailed standardized personal history was taken, blood was drawn for total IgE, inhalant allergen mix SX1, food allergen mix Fx5 (Pharmacia CAP Systems, Freiburg, Germany). A skin-prick test was performed including inhalant allergens (grass pollen, birch, cat, mugwort, dog, derm. pt., alternaria, cladosporium, latex by ALK and Bencard, Germany) and food (apple, raw potato, carrot, celery, barley, wheat and rye flour, oat meal, pork meat, hen's egg, cow's milk, soy, herring, mackerel, mussels, crab, hazelnut, walnut, peanut, sesame, poppy, guar gum, carob, carrageen). The food items were tested using the prick-to-prick technique with native food (except for soy bean and fish). Native food items were chosen since their sensitivity is higher compared to commercial extracts (11). Prick tests were regarded as positive when wheel diameter was at least 3 mm and is equal to 60% of that induced by histamine. In case suspicion of adverse reactions to food could not be ruled out after these investigations, persons were asked to participate in controlled challenge tests. They received detailed instructions by a nutritionist how to omit also hidden sources of suspicious food items from the diet for a specified period of time and how to keep a diary, in order to certify that the diet was followed properly. Duration on the diet depended on the clinical symptoms and was 3 days for typical immediate type reactions, like gastrointestinal symptoms, and 7 days for eczema.

image

Figure 1. Flow sheet of the study and investigations performed.

Download figure to PowerPoint

Elimination diet was followed by DBPCFC at the clinic. Only in case adverse reactions to food appeared to be unlikely in patient's history, this was preceded by an open challenge test. If this open challenge test was negative, food intolerance was excluded. If it was positive, DBPCFC tests were performed as well. For DBPCFC tests food items were blinded by the nutritionist using Sinlac® (Nestle, München, Germany; ingredients: rice flour, carob, saccharose, natriumcitrate, potassiumchloride, natriumphosphate, vitamines, zincsulfate), orange flavor (SHS International, Liverpool, UK), carotine, cereal flakes and/or pure cacao powder. Blinding was confirmed by tasting panels. At one time, a range of food additives known to cause non-allergic intolerance reactions were given in altogether 13 capsules (Table 1). On another day, the same number of capsules filled with mannit and silicium dioxide were given as a placebo. All provocation tests were performed at the clinic, with subsequent surveillance of 2–12 h, depending on patient's history.

Table 1.  Pseudoallergens used for provocation tests in gelatine capsules
PseudoallergensNameE-numberDosage (mg)
Coloring agents
Azo-dyesTartrazineE10250
Sunset yellowE1105
AzorubineE1225
AmaranteE1235
PonceauE1245
Brillant black BNE1515
Other synthetic dyesQuinoline yellowE1045
ErythrosineE1275
Patent blueE1315
IndigotineE1325
Natural colorsIron(III) oxideE1725
Red cochinealE1205
PreservativesSorbic acidE2001000
SodiumbenzoateE2111000
p-HydroxybenzoateE214–2191000
Sodium metabisulfiteE22350
Sodium nitrateE251100
AntioxidantsButylhydroxyanisol (BHA)E32050
Butylhydroxytoluol (BHT)E32150
PropylgallateE31050
TocopherolE306–30950
Taste enhancerMonosodiumglutamateE621200
Naturally occurring substancesSalicylic acid 100

In addition, 50 persons of the population sample with no history of adverse reactions to food were asked to participate in an open food challenge test in the form of a three-course meal. This included the following food allergens: wheat, rye, oat, egg, soy milk, crab, codfish, milk, apple, carrot, celery, peanut, hazelnut, guar gum, carrageen, carob as well as the mixture of food additives which were given in capsules.

Statistics

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

Prevalence rates with 95% confidence intervals are reported. Assessment and correction for response bias was made by comparison of the study data with the Berlin Micro Census data of 1999, weighting of the study data was done with regard to age, sex and educational status.

Projections from the Berlin study data to the adult population of Germany (age 18–79 years) were computed by comparison with the German Health Survey Data from 1998 (12).

Weighting of the study data was done with regard to age, sex, education, atopic dermatitis, rhinoconjunctivitis and asthma.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

Questionnaire replies were received from 4093 persons, i.e. the response rate was 31%. Figure 2 shows the unweighted and the weighted data for sex, age and education. The lifetime prevalence of self-perceived adverse reactions against food was frequent (Fig. 3). Self-reported allergic symptoms were highest for allergic rhinitis with 23.4% lifetime prevalence and 19.7% in the last 12 months, followed by urticaria (9.4% lifetime prevalence, 5.2 in the last 12 months) and allergic asthma (9.4% lifetime prevalence, 6.7% in the last 12 months. In 2289 persons, answers in the questionnaire had given rise to possible suspicions for food intolerance. Of these, 1981 persons could be contacted by phone (86.2%). Suspicion of food intolerance could be excluded by patient's history in 547 persons, 814 of the remaining persons followed the invitation to a personal investigation. Of these 814 patients, 804 consented to prick testing (weighted 765 persons). The most frequent sensitizations were observed to pollen and pollen-related foods (Fig. 4). Of the 804 persons undergoing prick testing, 493 were women, 311 were men, in contrast at least one positive skin-prick test was observed more frequently in men tested than in women tested (61 vs 40.8%, P < 0.01).

image

Figure 2. Epidemiological data of persons responding to the questionnaire.

Download figure to PowerPoint

image

Figure 3. Self-reported prevalence of adverse reactions to food.

Download figure to PowerPoint

image

Figure 4. Prevalence of positive reactions to food and inhalant allergens in the skin-prick test.

Download figure to PowerPoint

Of the total group of 814, 794 persons consented to drawing blood for determination of IgE antibodies. Specific IgE was positive in 53.9% for at least one inhalant allergen and in 11.2% for at least one of the food allergens wheat, milk, egg, soy or fish. In 39.9% of the participants, an increase of the total serum IgE could be observed.

Following history, personal investigation, blood and skin-prick tests, food challenge tests were performed where the suspicion of food intolerance remained.

Open challenge was performed in 141 individuals with little suspicion of adverse reactions to food, followed by DBPCFC in case of a positive reaction. Altogether in 216 individuals a DBPCFC was performed. This was clearly positive in 116 persons (weighted data: 104 persons with IgE-mediated reactions and partly overlapping 44 persons with non-IgE-mediated reactions). In addition 50 individuals with negative history for food intolerance and negative skin-prick tests also underwent open challenge tests after 3 days diet free of typical food allergens and pseudoallergens. Of these, only one reported a mild unspecific itching sensation in the mouth.

Based on the data of the 116 persons with reactions proven by DBPCFC, the overall prevalence calculated for the Berlin population in all age groups was 3.6% (95% confidence interval [3.0–4.2%]). Of these 2.5% were IgE-mediated reactions against food and 1.1% were non-IgE-mediated hypersensitivity reactions to food. Calculated for the German adult population (18–70 years) the point prevalence of food intolerance is 2.6% (95% confidence interval [2.1–3.2%]).

The robustness of the reported prevalence rates with regard to incomplete data, i.e. persons who had returned the questionnaire but refused to answer phone calls or personal clinical investigations, was analyzed by a comparison of two approaches. First, prevalence rates were calculated by including persons with no missing information. Second, a simple imputation algorithm based on a multinominal model was used to predict the health status for persons with incomplete data from the available questionnaire data. Both approaches resulted in nearly identical prevalence rates.

For IgE-mediated food allergy the highest frequency with 4.3% is seen in the age group of 20–39 years (Fig. 5). Females constituted 60.6% of this case and the prevalence was higher in persons with a higher education (3.8% in persons with ‘Abitur’/high school, 12 years school education vs 1.5% in persons with ‘Hauptschule’, 9 years school education). The eliciting foods of IgE-mediated reactions were mostly pollen related (Fig. 6), hazelnuts and apple being the most frequent items (67.6% each) in the birch pollen related fruit and nuts group (Fig. 7). Other fruits were mainly pineapple, berries, kiwi and bananas.

image

Figure 5. Calculated prevalence of DBPCFC proven IgE-mediated food allergy depending on age.

Download figure to PowerPoint

image

Figure 6. Frequency of eliciting factors in DBPCFC proven IgE-mediated food allergy (given in percent of total number of persons with reactions, weighted n = 104).

Download figure to PowerPoint

image

Figure 7. Frequency of eliciting factors in DBPCFC proven food allergy in the group of nuts, pipfruit and stone fruit (given in percent of total number of persons with reactions, weighted n = 104).

Download figure to PowerPoint

For non-IgE-mediated food intolerance reactions the prevalence was highest in the age group of 60 years and older (Fig. 8). Similar to type-I allergies women were more frequently effected (59.1%), but in contrast to IgE-mediated food allergy there was no association with the level of education. Figure 9 shows the food items identified as eliciting factors in these patients.

image

Figure 8. Calculated prevalence of DBPCFC proven food allergy depending on age.

Download figure to PowerPoint

image

Figure 9. Frequency of DBPCFC proven eliciting foods in non-IgE-mediated adverse reactions to food (given in percent of total number of persons with reactions, weighted n = 44).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

This is the first representative cross-sectional population study addressing the prevalence of adverse reactions to food, taking into consideration all possible mechanisms of adverse reactions. Previous studies have either involved selective populations, e.g. children, or were limited to a selection of certain food items in provocation tests. In addition, our study is the largest of its type with the inclusion of 814 participants in the personal investigation.

The results reveal information not only about the prevalence rates of adverse reactions to food in the population of Berlin, which is the largest city in Germany and comparable to other large European cities in many aspects. But by statistical comparison, we could also provide a good estimation of the nationwide prevalence in Germany. This calculation is very robust, since it is based on data of the German Health Survey, which is a nationwide cross-sectional study of health issues in the German population performed by a touring medical team, performing personal medical examinations in randomized samples all over Germany. The limitations are, however, that these data are only available for the age group of 18–79 years. Yet comparing the small difference between the prevalence rate of all groups of the Berlin population (3.6%) and the prevalence of the Berlin population aged 18–79 years (3.7%), it can be expected that the number of adverse reactions to food in the adult German population (2.6%) would be very similar if all age groups had been included. The fact that adverse reactions to food are more frequent in the population of the city of Berlin than in the whole German population is not surprising, since other epidemiological data also point at higher rate of allergies in urban centers compared to the countryside.

The weighting of the data according to the Berlin Micro Census resulted only in a minor change of the Berlin population prevalence rate, which confirmed that the study could be regarded as representative.

The largest population study to date was performed by Young et al. (1) in randomly selected English households but only 93 persons were actually tested for food allergy against a limited range of items (cow's milk, hen's egg, wheat, orange, prawn, nuts and chocolate). They reported a prevalence of 1.4–1.8% but they did not include pollen-associated food allergy and nonimmunological types of adverse reactions to food. The calculated prevalence in the British population of 1.4–1.8%, however, is similar to our results regarding the fact that only a limited number of food items were investigated by Young et al. (1).

In the Netherlands, Jansen et al. (13) handed out a questionnaire to 1483 adults. Of these, 198 persons reported adverse reactions to food, 144 were investigated, 37 put on an elimination diet and only 19 underwent DBPCFC. From these results a prevalence of 2.4% for adverse reactions to food was calculated for the adult population of the Netherlands. In a Swedish study in the age group of 20–44 years in 1812 persons, a self-reported food intolerance of 25% was stated, but only 6% proved to be positive for one of the mentioned food items in the serum IgE (14). Kanny et al. (15) estimate the prevalence of food allergy in France at 3.24% based on questionnaire sent to a subsample of 1129 persons of a representative sample of the French population. However, provocation tests to verify the diagnosis were not performed in both studies.

In the most recent German population-based study, Schäfer et al. (16) investigated the frequency of food allergy and intolerance in a nested case–control study in the population of southern German town. Of the studied persons, 20.8% reported food related symptoms. Double-blind, placebo-controlled food challenge tests were also not undertaken.

Regarding pseudoallergic reactions to food, existing data are even more sparse. Only one large population-based study concentrating on food intolerance to additives has been performed previously by Young et al. (17), sending questionnaires to more than 11 000 British households. They calculated a prevalence of only 0.026% but based this on provocation tests with a number of food additives tested in only 81 persons. This is less than the 0.18% of reactions against food additives observed by us, but it needs to be noted that the range of additives used by us was broader, the dosages higher and we included not only patients who themselves had suspected to react to additives but also all patients with diseases like urticaria where food intolerance has frequently been described.

No study worldwide so far has looked into the prevalence of pseudoallergy against naturally occurring food ingredients, which have been described to be more important at least in urticaria than artificial additives (18, 19). This is supported by our study, where additives contributed only to a minority of pseudoallergic reactions detected by DBPCFC.

Nonallergic reactions against alcoholic beverages can be induced by additives like sulfite but also a considerable number of reactions appear to be induced by ethanol itself (20).

Our results are also in line with previous studies showing that the self-perceived food intolerance is higher than the point prevalence which can be detected only by appropriate testing. Thus, for true results in epidemiological studies the use of DBPCFC is inevitable for objective assessment and regarded as gold standard. However, two points need to be remembered: first, DBPCFC tests can also be falsely negative since co-factors, e.g. infections, alcoholic beverages, physical exercise, could be required (21). Second, symptoms were only transient and no longer present at the time of investigation. This holds especially for IgE-mediated food allergy in early childhood and pseudoallergic adverse reactions to food. For pseudoallergic reactions, data are available from acetylsalicylic acid intolerance, confirmed in DBPCFC, which can no longer be reconfirmed in 50% after 1 year and 90% after 10 years (22).

Thus, the reported number of positive DBPCFC in our study only reflects the point prevalence and probably underestimates the problem of adverse reactions to food in the population. On the other hand, the prevalence of self-reported adverse reactions in patient's history (lifetime prevalence 34.9% at the mean age of 41) is surely an overestimation. However, although the exact lifetime prevalence cannot be clearly calculated, it must be stated that in view of the point prevalence rate with more than one out of 50 truly reacting to food and the additional high number of individuals worried by symptoms which are perceived to be food related, adverse reactions to food are a major health issue. Further research efforts are, therefore, needed at a European level which includes monitoring for relevant allergens which should be included in food labeling (23).

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References

This study was supported by the German Ministry of Health. We thank Pharmacia, ALK Scherax and Bencard for providing allergy test material free of charge. We also like to thank our staff, especially Martin Balke, Meral Esen, Karin Forschner, Steven Goetze and our medical nurses.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Population
  5. Questionnaire
  6. Study design
  7. Statistics
  8. Results
  9. Discussion
  10. Acknowledgments
  11. References
  • 1
    Young E, Stoneham MD, Petruckevitch A, Barton J, Rona R. A population study of food intolerance. Lancet 1994;334: 11271130.
  • 2
    Emmett SE, Angus FJ, Fry JS, Lee PN. Perceived prevalence of peanut allergy in Great Britain and its association with other atopic conditions and with peanut allergy in other household members. Allergy 1999;54: 380385.
  • 3
    Ortolani C, Ispano M, Pastorello EA, Ansaloni R, Magri GC. Comparison of results of skin-prick tests (with fresh foods and commercila food extracts) and RAST in 100 patients with oral allergy syndrome. J Allergy Clin Immunol 1989;83: 683690.
  • 4
    Niggemann B, Reibel S, Roehr CC, Felger D, Ziegert M, Sommerfeld C, et al. Predictors of positive food challenge outcome in non-IgE-mediated reactions to food in children with atopic dermatitis. J Allergy Clin Immunol 2001;108: 10531058.
  • 5
    Roehr CC, Reibel S, Ziegert M, Sommerfeld C, Wahn U, Niggemann B. Atopy patch test together with level of specific IgE reduces the need for oral food challenges in children with atopic dermatitis. J Allergy Clin Immunol 2001;107: 548553.
  • 6
    Ortolani C, Bruijnzeel-Koomen C, Bengtsson U, Bindslev-Jensen C, Bjorksten B, Host A, et al. Controversial aspects of adverse reactions to food. European Academy of Allergology and Clinical Immunology (EAACI) reactions to food subcommittee. Allergy 1999;54: 2745.
  • 7
    Ortolani C, Ispano M, Ansaloni R, Rotondo F, Incorvaia C, Pastorello EA. Diagnostic problems due to cross-reactions in food allergy. Allergy 1998;53: 5861.
  • 8
    Bock SA, Sampson HA, Atkins FM. Double-blind, placebo controlled food challenge (DBPCFC) as an official procedure: a manual. J Allergy Clin Immunol 1988;82: 986997.
  • 9
    Sampson HA. Immunologically mediated food allergy: the importance of food challenge procedures. Ann Allergy 1988;60: 262269.
  • 10
    Niggemann B, Wahn U, Sampson HA. Proposals for standardization of oral food challenge tests in infants and children. Pediatr Allergy Immunol 1994;5: 1113.
  • 11
    Romano A, Di Fonso M, Giuffreda F, Papa G, Artesani MC, Viola M, et al. Food-dependent exercise-induced anaphylaxis: clinical and laboratory findings in 54 subjects. Int Arch Allergy Immunol 2001;125: 264272.
  • 12
    Stolzenberg H. German National Health Survey: public use file. Berlin: Robert Koch Institute, 2000.
  • 13
    Jansen JJN, Kardinaal AFM, Huijbers G, Vlieg-Boerstra BJ, Martens BPM, Ockhuizen T. Prevalence of food allergy and intolerance in the adult Dutch population. J Allergy Clin Immunol 1994;93: 446456.
  • 14
    Björnssen E, Janson C, Plaschke P, Norrman E, Sjöberg O. Prevalence of sensitization to food allergens in adult Swedes. Ann Allergy Asthma Immunol 1996;77: 327332.
  • 15
    Kanny G, Moneret-Vautrin DA, Flabbee J, Beaudouin E, Morisset M, Thevenin F. Population study of food allergy in France. J Allergy Clin Immunol 2001;108: 133140.
  • 16
    Schäfer T, Böhler E, Ruhdorfer S, Weigl L, Wessner D, Heinrich J, et al. Epidemiology of food allergy/food intolerance in adults: associations with other manifestations of atopy. Allergy 2001;56: 11721179.
  • 17
    Young E, Patel S, Stoneham RR, Wilkonson JD. The prevalence of reactions to food additives in a survey population. J R Coll Physicians Lond 1987;21: 241271.
  • 18
    Pigatto PD, Valsecchi RH. Chronic urticaria: a mystery. Allergy 2000;55: 306308.
  • 19
    Zuberbier T, Pfrommer C, Specht K, Vieths S, Bastl-Borrmann R, Worm M, et al. Aromatic components of food as novel eliciting factors of pseudoallergic reactions in chronic urticaria. J Allergy Clin Immunol 2002;109: 348349.
  • 20
    Ehlers I, Hipler CU, Zuberbier T, Worm M. Ethanol as a cause of hypersensitivity reactions to alcoholic beverages. Clin Exp Allergy 2002;32: 12311235.
  • 21
    Hanakawa Y, Tohyama M, Shirakata Y, Murakami S, Hashimaoto K. Food-dependent exercise-induced anaphylaxis: a case related to the amount of food allergen ingested. Br J Dermatol 1998;138: 898900.
  • 22
    Paul E, Geipel M, Moller R. Catamnestic studies of patients with chronic urticaria and aspirin intolerance. Hautarzt 1994;45: 1214.
  • 23
    Bousquet J, Bjorksten B, Bruijnzeel-Koomen CA, Huggett A, Ortolani C, Warner JO, et al. Scientific criteria and the selection of allergenic foods for product labelling. Allergy 1998;53: 321.