Buscar en
Allergologia et Immunopathologia
Toda la web
Inicio Allergologia et Immunopathologia Self-reported hypersensitivity and allergy to foods amongst Mexican adolescents:...
Journal Information
Vol. 47. Issue 3.
Pages 246-253 (May - June 2019)
Share
Share
Download PDF
More article options
Visits
2231
Vol. 47. Issue 3.
Pages 246-253 (May - June 2019)
Original Article
Full text access
Self-reported hypersensitivity and allergy to foods amongst Mexican adolescents: Prevalence and associated factors
Visits
2231
T.R. Bedolla-Pulidoa, M. Bedolla-Barajasa,
Corresponding author
drmbedbar@gmail.com

Corresponding author.
, J. Morales-Romerob, T.I. Bedolla-Pulidoc, M.V. Domínguez-Garcíad, D.D. Hernández-Colína, M.V. Flores-Merinod
a Servicio de Alergia e Inmunología Clínica, Hospital Civil de Guadalajara “Dr. Juan I. Menchaca”, Salvador de Quevedo y Zubieta No. 750, Colonia La Perla, Guadalajara, Jalisco C.P. 44340, Mexico
b Instituto de Salud Pública, Universidad Veracruzana, Av. Luis Castelazo Ayala s/n., Colonia Industrial Ánimas, Xalapa, Veracruz C.P. 91190, Mexico
c Universidad de Guadalajara, Centro Universitario de Ciencias de la Salud, Sierra Mojada No. 950, Colonia Independencia Oriente, Guadalajara, Jalisco C.P. 44340, Mexico
d Universidad Autónoma del Estado de México, Centro de Investigación en Ciencias Médicas, Estado de México, Jesús Carranza No. 205, Colonia Universidad, Toluca de Lerdo, Estado de México, C.P. 50130, Mexico
This item has received
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Tables (5)
Table 1. Frequency of self-reported clinical manifestations associated to the ingestion of foods amongst 1992 late adolescents.
Table 2. Foods most frequently associated with food hypersensitivity and probable food allergy amongst 1992 late adolescents.
Table 3. Primary foods associated with probable allergy symptoms that are regulated by IgE.
Table 4. Multivariate model of factors associated with food hypersensitivity.
Table 5. Multivariate model of factors associated with probable food allergy.
Show moreShow less
Abstract
Background

The prevalence of food allergy is on the rise on a global scale.

Objective

To determine the prevalence of food hypersensitivity (FHS) and probable food allergy (PFA), as well as the foods and factors associated with these occurrences.

Methods

A cross-sectional study was carried out among 1992 adolescents (aged 15–18 years). Each adolescent answered a structured questionnaire. A multivariate analysis was used to identify the association between the variables.

Results

The prevalence of FHS was 10.6% (the most commonly associated foods were shrimp, cow's milk and avocado) and the PFA was 7.8% (shrimp, cow's milk and pecan). The prevalences of oral allergy syndrome, food-associated urticaria and systemic reaction were 4.9%, 3.6% and 1.5%, respectively. The following factors were associated with FHS: personal history of asthma (OR 1.63; 95% CI: 1.11–2.41), allergic rhinitis (OR 2.60; 95% CI: 1.75–3.87), atopic dermatitis (OR 2.07; 95% CI: 1.25–3.43), maternal history of asthma (OR 1.80; 95% CI: 1.02–3.16), atopic dermatitis (OR 6.11; 95% CI: 2.45–15.29), and female sex (OR 1.89; 95% CI: 1.38–2.59). PFA was associated with a personal history of asthma (OR 1.65; 95% CI: 1.06–2.56), allergic rhinitis (OR 2.46; 95% CI: 1.56–3.88), atopic dermatitis (OR 2.02; 95% CI: 1.15–3.54), paternal allergic rhinitis (OR 2.52; 95% CI: 1.15–5.51), maternal atopic dermatitis (OR 7.46; 95% CI: 2.93–19.00), and female sex (OR 1.89; 95% CI: 1.31–2.72).

Conclusion

The adverse reactions associated with foods among late adolescents are a frequent occurrence, and the most commonly associated factor is atopy.

Keywords:
Food allergy
Adolescent
Prevalence
Cross-sectional studies
Full Text
Introduction

The prevalence of food allergy is increasing worldwide,1 and can have an important effect on the quality of life, as food allergy is likely to provoke systemic reactions.2

Globally, the prevalence of self-reported hypersensitivity to food in children and adults has been estimated at 3–35%.3 In Latin America, there is a lack of population-based studies dedicated to establishing the prevalence of adverse food reactions. Nevertheless, one of the first of such studies occurred in Colombia, where the prevalence of food allergies in children aged 9–16 was 12.6%.4 Additional studies showed that preschool-aged Brazilian children had a prevalence of 23.5%,5 Chile reported a prevalence of 5.5% for school-aged children,6 and the prevalence in Mexico was 10%.7 Among early teens in Australia, the prevalence was 5.5%.8

Some of the foods that were most commonly associated with food hypersensitivity (FHS) and probable food allergy (PFA) were peanut, tree nuts, cow's milk and shellfish, among others.4–10 However, these can vary based on the geographic region, food availability, eating habits, and cultural customs.

While the factors associated with food allergy may vary, the most consistent factor is genetics.4,9,10 Furthermore, women exhibit these symptoms more frequently than men do.11–13

To our knowledge, FHS and PFA in late adolescence (the final stage of childhood) have not been previously studied. Our objectives consisted of determining the prevalence of FHS and PFA, as well as their clinical characteristics, the foods that are involved, and the factors associated with them.

MethodsDesign

In our cross-sectional study, we analyzed data that correspond to 1992 adolescents, aged 15–18 years. These students were enrolled in high schools within Guadalajara, Mexico, one of the nine cities that constitute the metropolitan area of the same name. The city of Guadalajara includes 1.5 million out of the approximate five million inhabitants throughout the entire metropolitan area; its large population is representative of the diverse socioeconomic strata in the region.

Sample

The samples were collected from April to June of 2016 through probabilistic sampling that was stratified and then clustered. To do so, we divided the city of Guadalajara into seven strata, and from each of these, we randomly selected at least one school (cluster); each stratum contributed a sub-sample of students through the proportional selection technique, by multiplying the complete sample size by the proportion of students that were enrolled in each stratum. To obtain a representative sample for each selected school, each school grade was considered a new stratum, and from each of these, we randomly selected a group. Lastly, we randomly selected students to participate in the study.

Questionnaire

The data were obtained through a structured questionnaire that was previously used and validated within the Mexican population.10 Said questionnaire includes questions regarding age, sex, personal history of atopic and non-atopic diseases of each participant, and the family history of allergic diseases of each participant's parents.

Definitions

We utilized the following question to identify self-reported FHS: Did you experience an allergic reaction after eating foods or drinking something?10 This study included all cases in which the student responded affirmatively, regardless of a prior history of allergy testing.14 To define PFA, we adopted the definition proposed by Sicherer et al.; a probable allergic reaction was considered positive if the organs and the manifested symptoms were typical allergic reactions (urticaria and angioedema, difficulty breathing, wheezing, swelling of the throat, vomiting and diarrhea) and if these occurred within two hours of having ingested the allergy-causing food.15 This definition allows us to detect up to 93% of the subjects that have specific IgE antibodies that react to the implied foods.16 If a participant had two or more organs that were affected, we considered this to be a systemic reaction.

Statistical analyses

To estimate the prevalence of FHS, we divided the frequency of affirmative answers to the question, Do you suffer from allergic reactions after ingesting food or drinking something?, by the total number of people who completed the questionnaire. The same procedure was utilized to estimate the prevalence of PFA, except in this case, we utilized the previously mentioned definition. We employed the chi-square test or Fisher's exact test, as deemed necessary, in order to compare the categorical variables. We estimated the confidence intervals at 95% (95% CI) for the proportions. The data were analyzed by using the IBM SPSS Statistics version 20.0 (IBM Co., Armonk, NY, USA) program for Windows.

To identify the association between variables, we utilized two multivariate model groups in which the dependent variables were FHS and PFA, respectively. In both groups, the independent covariates were personal, paternal or maternal history of asthma, allergic rhinitis, atopic dermatitis, and the female sex. The strength of the association between variables was evaluated by estimating the odds ratio (OR) and the 95% CI through adjusted binary logistic regression with the use of the forward conditional method. The value of p<0.05 was considered to be statistically significant.

Ethics

The authors certify that this research was done under full compliance with all government policies and the Helsinki Declaration (2013). The Committee of Research Ethics at the Hospital Civil de Guadalajara approved this study. Prior to distributing our questionnaire, we explained our objective to each school district and their teachers and obtained their consent in order to be able to survey their students. If they accepted, we sought the same permission from the students, and when students agreed to participate, they were given detailed information about our study and were asked to give their verbal consent to participate.

Results

Overall, we analyzed data that corresponded to 963 men (47.0%) and 1056 women (53.0%). With regards to the personal history of allergic diseases among adolescents, we registered asthma at 12.7%, allergic rhinitis at 9.0% and atopic dermatitis at 5.2%. Where their family history was concerned, we found that the paternal history was as follows: asthma 2.6%, allergic rhinitis 2.1% and atopic dermatitis 0.7%. Maternal history corresponded to 4.5%, 3.4% and 1.1%, respectively.

The self-reported prevalence of FHS was 10.6% (95% CI: 9.4–12.1%), whereas the prevalence of PFA was 7.8% (95% CI: 6.7–9.0%). Once again, these types of reactions were predominant among women (10.1% of females vs. 5.1% of males, p<0.001).

The most common self-reported symptoms for FHS were cutaneous (hives and reddening of the skin), followed by respiratory ailments (oppression of the pharynx and difficulty breathing) and lastly intestinal discomfort (abdominal pain and heartburn) (Table 1). In PFA, cutaneous symptoms were predominant (hives and reddening of the skin), followed by respiratory ailments (oppression of the pharynx and difficulty breathing) and intestinal discomfort (abdominal pain and heartburn).

Table 1.

Frequency of self-reported clinical manifestations associated to the ingestion of foods amongst 1992 late adolescents.

  Food hypersensitivity n=212  Probable food allergy n=155
  n    n 
CutaneousCutaneous
Hives  71  33.5  Reddening of skin  70  45.2 
Reddening of skin  70  33.0  Itchiness of lips  59  38.1 
Itchiness of lips  59  27.8  Hives  59  38.1 
Edema of the mouth  51  24.1  Edema of the mouth  51  32.9 
Itchiness of the body  44  20.8  Itchiness of the body  40  25.8 
Skin rashes  42  19.8  Facial edema  37  23.9 
Facial edema  37  17.5  Skin rashes  36  23.2 
Perspiration  13  6.1  Perspiration  10  6.5 
IntestinalIntestinal
Colic  45  21.2  Heartburn  26  16.8 
Heartburn  41  19.3  Colic  25  16.1 
Diarrhea  36  17.0  Vomiting  24  15.5 
Vomiting  35  16.5  Diarrhea  21  13.5 
Abdominal distention  27  12.7  Abdominal distention  21  13.5 
Flatulence  26  12.3  Constipation  15  9.7 
Constipation  25  11.8  Flatulence  12  7.7 
RespiratoryRespiratory
Pharynx oppression  62  29.2  Pharynx oppression  62  40.0 
Difficulty breathing  51  24.1  Difficulty breathing  49  31.6 
Sneezing  49  23.1  Nasal congestion  37  23.9 
Nasal congestion  39  18.4  Sneezing  37  23.9 
Runny nose  37  17.5  Runny nose  35  22.6 
Coughing  35  16.5  Coughing  34  21.9 
Wheezing  20  9.4  Wheezing  20  12.9 

The foods that were most commonly associated with self-reported FHS were shrimp, cow's milk, avocado, pecan, and cow's milk cheese; for PHF, they were shrimp, cow's milk, pecan, avocado and almond (Table 2).

Table 2.

Foods most frequently associated with food hypersensitivity and probable food allergy amongst 1992 late adolescents.

  n 
Food hypersensitivity
Shrimp  53  2.7 
Cow's milk  37  1.9 
Avocado  19  1.0 
Pecan  20  1.0 
Cow's milk cheese  18  0.9 
Pork sausage  18  0.9 
Almond  17  0.9 
Banana  16  0.8 
Oyster  15  0.8 
Peach  13  0.7 
Apple  14  0.7 
Clam  14  0.7 
Fish  13  0.7 
Probable food allergy
Shrimp  46  2.3 
Cow's milk  20  1.0 
Pecan  18  0.9 
Avocado  17  0.9 
Almond  16  0.8 
Pork sausage  15  0.8 
Peach  13  0.7 
Apple  13  0.7 

The prevalence of diseases that are likely regulated by IgE and the main ailments that were associated with them were oral allergy syndrome 4.9% (95% CI: 4.1–6.0%), where shellfish and fish (primarily shrimp) were the most frequent triggering agents (Table 3). The prevalence of urticaria was 3.6% (95% CI: 2.9–4.5%), and once again, shrimp was the main culprit. The frequency of systemic symptoms was 1.5% (95% CI: 1.0–2.1%), and in this case, both shrimp and peach were the primary factors.

Table 3.

Primary foods associated with probable allergy symptoms that are regulated by IgE.

Oral allergy syndromeUrticariaSystemic symptoms
n=98  n  n=72  n  n=29  n 
Food categoriesFood categoriesFood categories
Shellfish & fish  37  37.8  Shellfish & fish  24  33.3  Fruit  10  34.5 
Fruit  35  35.7  Fruit  15  20.8  Shellfish & fish  24.1 
Tree nuts  17  17.3  Tree nuts  12  16.7  Tree nuts  17.2 
Dairy  13  13.3  Dairy  12.5  Vegetables  13.8 
Vegetables  10  10.2  Vegetables  9.7  Dairy  3.4 
Individual foodsIndividual foodsIndividual foods
Shrimp  36  36.7  Shrimp  18  25.0  Shrimp  27.6 
Avocado  13  13.3  Cow's milk  12.5  Peach  17.2 
Cow's milk  13  13.3  Almond  12.5  Brazilian nut  17.2 
Peach  11  11.2  Walnut  11.1  Avocado  13.8 
Apple  11  11.2  Avocado  9.7  Almond  13.8 
Almond  11  11.2  Clam  9.7  Guava  10.3 
Walnut  11  11.2  Oyster  9.7  Mango  10.3 
Banana  10  10.2  Peach  8.3  Apple  10.3 
Clam  10  10.2  Cow's milk cheese  8.3  Banana  10.3 
Fish  9.2  Peanut  8.3  Lobster  10.3 

The factors that were associated with FHS are shown in Table 4. We adjusted Model 1 to the personal history of allergic disease and sex; as a result, asthma (OR 1.75, p=0.004), allergic rhinitis (OR 2.67, p<0.001), atopic dermatitis (OR 2.39, p<0.001) and female sex (OR 1.88, p<0.001) were associated with FHS. Models 2 and 3 show that a paternal history of allergic rhinitis (OR 3.03, p=0.002) and atopic dermatitis (OR 4.69, p<0.001) was associated with FHS, as was a maternal history of asthma (OR 2.02, p=0.012), allergic rhinitis (OR 2.08, p=0.019) and atopic dermatitis (OR 7.92, p<0.001). The final model, which was based on the previously identified covariates, shows that a personal history of asthma, allergic rhinitis and atopic dermatitis, along with a maternal history of asthma or atopic dermatitis and female sex, remained associated with FHS.

Table 4.

Multivariate model of factors associated with food hypersensitivity.

  OR  95% CI  p Value 
Model 1
Personal history of:
Asthma  1.75  1.19–2.56  0.004 
Allergic rhinitis  2.67  1.81–3.95  <0.001 
Atopic dermatitis  2.39  1.47–3.89  <0.001 
Sex (female)  1.88  1.37–2.57  <0.001 
Model 2
Paternal history of:
Asthma  –  –  0.158 
Allergic rhinitis  3.03  1.49–6.19  0.002 
Atopic dermatitis  4.69  1.53–14.3  <0.001 
Sex (female)  2.02  1.49–2.75  <0.001 
Model 3
Maternal history of:
Asthma  2.02  1.17–3.51  0.012 
Allergic rhinitis  2.08  1.13–3.85  0.019 
Atopic dermatitis  7.92  3.30–19.02  <0.001 
Sex (female)  2.00  1.46–2.72  <0.001 
Final model
Personal history of:
Asthma  1.63  1.11–2.41  0.013 
Allergic rhinitis  2.60  1.75–3.87  <0.001 
Atopic dermatitis  2.07  1.25–3.43  0.005 
Paternal history of:
Allergic rhinitis  –  –  0.189 
Atopic dermatitis  –  –  0.450 
Maternal history of:
Asthma  1.80  1.02–3.16  <0.001 
Allergic rhinitis  –  –  0.369 
Atopic dermatitis  6.11  2.45–15.29  <0.001 
Sex (female)  1.89  1.38–2.59  <0.001 

OR: Odds ratios obtained through logistical regression using forward conditional method.

ORs were not calculated for variables excluded from the model.

All covariates were introduced as categorical variables (with and without the characteristic of interest).

Table 5 displays the various models that were designed to identify the associated factors in the PFAs. Model 1 once again shows that the personal history of asthma, allergic rhinitis, atopic dermatitis and female sex were associated with PFA.

Table 5.

Multivariate model of factors associated with probable food allergy.

  OR  95% CI  p Value 
Model 1
Personal history of:
Asthma  1.73  1.12–2.66  0.012 
Allergic rhinitis  2.78  1.81–4.28  <0.001 
Atopic dermatitis  2.43  1.43–4.14  0.001 
Sex (female)  1.87  1.30–2.69  0.001 
Model 2
Paternal history of:
Asthma  –  –  0.108 
Allergic rhinitis  4.43  2.16–9.10  <0.001 
Atopic dermatitis  6.85  2.23–21.00  0.001 
Sex (female)  2.03  1.42–2.89  <0.001 
Model 3
Maternal history of:
Asthma  –  –  0.097 
Allergic rhinitis  2.43  1.26–4.71  0.008 
Atopic dermatitis  9.75  4.02–23.63  <0.001 
Sex (female)  2.02  1.41–2.89  <0.001 
Final model
Personal history of:
Asthma  1.65  1.06–2.56  0.025 
Allergic rhinitis  2.46  1.56–3.88  <0.001 
Atopic dermatitis  2.02  1.15–3.54  0.014 
Paternal history of:
Allergic rhinitis  2.52  1.15–5.51  0.021 
Atopic dermatitis  –  –  0.214 
Maternal history of:       
Allergic rhinitis  –  –  0.442 
Atopic dermatitis  7.46  2.93–19.00  <0.001 
Sex (female)  1.89  1.31–2.72  0.001 

OR: Odds ratios obtained through logistical regression using forward conditional method.

ORs were not calculated for variables excluded from the model.

All covariates were introduced as categorical variables (with and without the characteristic of interest).

In Models 2 and 3, the paternal and maternal history of allergic rhinitis and atopic dermatitis, along with the female sex, were associated with PFA in the surveyed individuals. In the final model, a history of asthma, allergic rhinitis, and atopic dermatitis among adolescents, as well as paternal allergic rhinitis, maternal atopic dermatitis, and the female sex, remained as significant, independently associated factors with PFA.

Discussion

Studies aimed at estimating the prevalence of FHS and PFA in Latin America, primarily among adolescents, are scarce. This is one of the first studies to analyze this matter in our region. We observed that the prevalence of FHS was slightly over 10% and that PFA was at 8%, and in both cases, shrimp and cow's milk were the most common triggering agents. Interestingly, in addition to shrimp, peaches were associated with systemic reactions. Moreover, we were able to identify that a personal and a family history of allergic disease, especially on the mother's side, as well as female sex, were the main factors associated with both FHS and PFA.

In Melbourne, Australia, a recent cross-sectional study that included adolescents aged 10–14 showed that the prevalence of food allergy was 5.5% (determined with the use of a questionnaire), However, when skin-prick tests or oral challenge test were applied, there was a slight decrease to 4.5%.8 An additional study in Japan analyzed the prevalence of food allergies among female adolescents with the use of a questionnaire and reported a frequency of 8.2%.17 In Latin America, a cross-sectional questionnaire-based study proved that the prevalence of FHS among individuals aged 9–16 was 12.6%.4 In our country, there are limited efforts focused on researching the adverse reactions of food ingestion, and there are none that focus on adolescents. However, there was a study of young university students that identified adverse reactions to foods by utilizing the questionnaire that we have employed for this study. The university-based study recorded the prevalence of FHS at 30.1%. Interestingly, the frequency of PFA was quite different, at only 6%. An additional study analyzed the prevalence of FHS in young adults with the use of a questionnaire; their reported prevalence was 17%.9 Notably, this makes our study the first to report the prevalence of food allergies among late adolescents in Mexico.

For both FHS and PFA, the most commonly associated adverse foods were shrimp and cow's milk. It seems that this could be a constant among the general population of our country, as there have been two prior studies in two different geographic regions, in which these foods were the most frequently reported triggering agents.9,10 In a study carried out among early adolescents, the most commonly associated foods were peanuts and tree nuts.8 In a study that included children, adolescents and adults from various European countries, hazelnuts and apples were the foods that caused the most hypersensitivity; interestingly, neither shrimp nor cow's milk were significant factors.18

Something similar was observed in Canada, where the most commonly associated foods among children and adults were peanuts and tree nuts.19 A Mexican diet, which is primarily based on corn tortillas, beans and the low consumption of refined grains, usually does not include the regular consumption of shrimp and cow's milk. Consequently, we suggest that this diet is a risk factor and increases the probability of developing an allergy to these foods within our population, since in comparison to other populations, our time of exposure to such foods is considerably shorter.20 Foods such as peanuts, which in other parts of the world are a major agent of allergic reaction,2,8,9,21 are not a major culprit in our population, although the pecan is.

Although there were differences in age groups, the prevalence of OAS, urticaria and systemic reactions associated with foods, these were similar to the findings among adults in our city.9 In a population of late adolescents, OAS was the most common clinical entity, and notably, it was linked to the ingestion of shellfish (mainly shrimp). However, it is likely that we are only observing the natural history of the disease among these patients, in which the persistence of a stimulated allergen begins with oral symptoms, followed by urticaria and the triggering of anaphylaxis. Up to 1.5% of late adolescent symptoms were compatible with anaphylaxis after ingesting foods, and shrimp and peaches were among the foods most related to this event. Previous epidemiological studies have shown similar findings.4,18,22–24 Notably, none of these late adolescents had been informed by a physician that these symptoms were characteristics of an anaphylactic reaction.

In our study, various factors were significantly associated with FHS and FPA. One such factor was a personal history of an allergic disease. Notably, previous studies have had similar results to our own.4,9,10 Moreover, a maternal history of atopic dermatitis acts as a conducive agent, which we had previously observed among university students in our country,10 although we have yet to find an explanation for this occurrence.

Lastly, the predominance of FHS and FPA among women, which is reported in our study, is similar to that observed for asthma, allergic rhinitis, or atopic dermatitis, among others.11–13 It seems that within our adolescent population, these reactions are triggered during the final stage of adolescence and not during adulthood. Some of the factors that might help explain the prevalence of adverse reactions to foods among women are: hormonal (primarily estrogens), psychological sensitivity, the use of medications (such as antacids), food preferences, and the processing of information, among others.25,26 While various hypotheses have emerged to explain the causes of food allergies, some of the most reoccurring are the hypothesis of fatty diets and allergies, the antioxidant hypothesis, the vitamin D hypothesis, the hygiene hypothesis, the exposure to food allergens, and recently the dual exposure to allergen hypothesis.27,28

An important limitation of studies based on self-reported symptoms that are compatible with adverse reactions and are associated with food ingestion is that they tend to overestimate or underestimate the prevalence, because most of these occurrences lack objective tests (clinical history, specific identification of IgE, or oral challenge tests). In this study, as we only included adolescents that were from an industrialized city, the generalization of our results is limited when considering rural regions, or places that do not have access to the same types of foods. Two additional limitations of our study involve the lack of data related to the triggers of allergic rhinitis symptoms in students and the inability of the questionnaire to identify a student with symptoms related to the consumption of foods containing baked or cooked milk. On the other hand, given that this is the first epidemiological study of its kind based on probabilistic sampling, our results can be extrapolated among the population of late adolescents in our city.

In conclusion, the adverse reactions associated with foods among late adolescents in Guadalajara, Mexico are a common occurrence. The prevalence of FHS was calculated at 10.6%, while FPA was estimated at 7.8%. Shrimp and cow's milk were the primary allergens in this age group. Notably, the prevalence of compatible symptoms with a systemic reaction linked to a single food was calculated at 1.5%. The most important risk factors associated with FHS and FPA were a prior personal history, a relative with an allergic disease, and the female sex.

Funding source

None declared.

Conflict of interest

The authors have no conflict of interest to declare.

References
[1]
L. Verrill, R. Bruns, S. Luccioli.
Prevalence of self-reported food allergy in U.S. adults: 2001, 2006, and 2010.
Allergy Asthma Proc, 36 (2015), pp. 458-467
[2]
D. Antolín-Amérigo, L. Manso, M. Caminati, et al.
Quality of life in patients with food allergy.
Clin Mol Allergy, 14 (2016), pp. 1-10
[3]
R.J. Rona, T. Keil, C. Summers, D. Gislason, et al.
The prevalence of food allergy: a meta-analysis.
J Allergy Clin Immunol, 120 (2007), pp. 638-646
[4]
J. Marrugo, L. Hernández, V. Villalba.
Prevalence of self-reported food allergy in Cartagena (Colombia) population.
Allergol Immunopathol (Madrid), 36 (2008), pp. 320-324
[5]
L.C. Gonçalves, T.C. Guimarães, R.M. Silva, et al.
Prevalence of food allergy in infants and pre-schoolers in Brazil.
Allergol Immunopathol (Madrid), 44 (2016), pp. 497-503
[6]
R. Hoyos-Bachiloglu, D. Ivanovic-Zuvic, J. Álvarez, et al.
Prevalence of parent-reported immediate hypersensitivity food allergy in Chilean school-aged children.
Allergol Immunopathol (Madrid), 42 (2014), pp. 527-532
[7]
N. Ontiveros, E.E. Valdez-Meza, M.J. Vergara-Jiménez, A. Canizalez-Román, A. Borzutzky, F. Cabrera-Chávez.
Parent-reported prevalence of food allergy in Mexican schoolchildren: a population-based study.
Allergol Immunopathol (Madrid), 44 (2016), pp. 563-570
[8]
M. Sasaki, J.J. Koplin, S.C. Dharmage, et al.
Prevalence of clinic-defined food allergy in early adolescence: the Schoolnuts study.
J Allergy Clin Immunol, 141 (2018),
391–398.e4
[9]
M. Bedolla-Barajas, T.R. Bedolla-Pulido, A.S. Camacho-Peña, E. González-García, J. Morales-Romero.
Food hypersensitivity in mexican adults at 18 to 50 years of age: a questionnaire survey.
Allergy Asthma Immunol Res, 6 (2014), pp. 511-516
[10]
C. Puente-Fernández, R.L. Maya-Hernández, M.V. Flores-Merino, M. Romero-Figueroa, S. del, M. Bedolla-Barajas, et al.
Self-reported prevalence and risk factors associated with food hypersensitivity in Mexican young adults.
Ann Allergy Asthma Immunol, 116 (2016), pp. 523-527
[11]
A. Akcay, Z. Tamay, A. Ergin, N. Guler.
Prevalence and risk factors of atopic eczema in Turkish adolescents.
Pediatr Dermatol, 31 (2014), pp. 319-325
[12]
A.H. Ziyab.
Prevalence and risk factors of asthma, rhinitis, and eczema and their multimorbidity among young adults in kuwait: a cross-sectional study.
Biomed Res Int, (2017), pp. 1-10
2184193
[13]
L.M. Chu, P. Pahwa.
Prevalence and associated factors for self-reported asthma in a Canadian population: the Canadian Community Health Survey, 2014.
[14]
M. Ben-Shoshan, D.W. Harrington, L. Soller, et al.
A population-based study on peanut, tree nut, fish, shellfish, and sesame allergy prevalence in Canada.
J Allergy Clin Immunol, 125 (2010), pp. 1327-1335
[15]
S.H. Sicherer, A. Munoz-Furlong, A.W. Burks, H.A. Sampson.
Prevalence of peanut and tree nut allergy in the US determined by a random digit dial telephone survey.
J Allergy Clin Immunol, 103 (1999), pp. 559-562
[16]
S.H. Sicherer, A.W. Burks, H.A. Sampson.
Clinical features of acute allergic reactions to peanut and tree nuts in children.
Pediatrics, 102 (1998), pp. 1-6
[17]
A. Fujimori, T. Yamashita, M. Kubota, H. Saito, N. Takamatsu, M. Nambu.
Comparison of the prevalence and characteristics of food hypersensitivity among adolescent and older women.
Asia Pac J Clin Nutr, 25 (2016), pp. 858-862
[18]
N.E. Eriksson, C. Möller, S. Werner, J. Magnusson, U. Bengtsson, M. Zolubas.
Self-reported food hypersensitivity in Sweden, Denmark, Estonia, Lithuania, and Russia.
J Investig Allergol Clin Immunol, 14 (2004), pp. 70-79
[19]
L. Soller, M. Ben-Shoshan, D.W. Harrington, et al.
Overall prevalence of self-reported food allergy in Canada.
J Allergy Clin Immunol, 130 (2012), pp. 986-988
[20]
M. Bedolla-Barajas, T.R. Bedolla-Pulido, N. Macriz-Romero, J. Morales-Romero, M. Robles-Figueroa.
Prevalence of peanut, tree nut, sesame, and seafood allergy in Mexican adults.
Rev Invest Clin, 67 (2015), pp. 379-386
[21]
B.I. Nwaru, L. Hickstein, S.S. Panesar, EAACI Food Allergy and Anaphylaxis Guidelines Group, et al.
Prevalence of common food allergies in Europe: a systematic review and meta-analysis.
Allergy, 69 (2014), pp. 992-1007
[22]
M. Osterballe, T.K. Hansen, C.G. Mortz, A. Høst, C. Bindslev-Jensen.
The prevalence of food hypersensitivity in an unselected population of children and adults.
Pediatr Allergy Immunol, 16 (2005), pp. 567-573
[23]
G. Kanny, D.A. Moneret-Vautrin, J. Flabbee, E. Beaudouin, M. Morisset, F. Thevenin.
Population study of food allergy in France.
J Allergy Clin Immunol, 108 (2001), pp. 133-140
[24]
V.L. McWilliam, J.J. Koplin, M.J. Field, et al.
Self-reported adverse food reactions and anaphylaxis in the SchoolNuts study: a population-based study of adolescents.
J Allergy Clin Immunol, 141 (2018), pp. 982-990
[25]
S.M. Afify, I. Pali-Schöll.
Adverse reactions to food: the female dominance – a secondary publication and update.
World Allergy Organ J, 10 (2017), pp. 1-8
[26]
E. Ridolo, C. Incorvaia, I. Martignago, M. Caminati, G.W. Canonica, G. Senna.
Sex in respiratory and skin allergies.
Clin Rev Allergy Immunol, (2018), pp. 1-11
[27]
G. Lack.
Epidemiologic risks for food allergy.
J Allergy Clin Immunol, 121 (2008), pp. 1331-1336
[28]
S.H. Sicherer, H.A. Sampson.
Food allergy: a review and update on epidemiology, pathogenesis, diagnosis, prevention, and management.
J Allergy Clin Immunol, 141 (2018), pp. 41-58
Copyright © 2018. SEICAP
Article options
Tools
es en pt

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?

Você é um profissional de saúde habilitado a prescrever ou dispensar medicamentos