Ultra-processed food intake and food allergy in children and adolescents

Abstract This study aimed to describe the dietary intake of ultra-processed foods (UPF) by children and adolescents with food allergy (FA) and to verify a possible association between the UPF intake with clinical characteristics and nutritional status in this group. This cross-sectional study included 110 children and adolescents with single or multiple FA IgE and non-IgE mediated. We evaluated food intake using the NOVA classification through the three 24-h recalls. The average contribution of UPF to total energy intake (calories) ranges from 21% in the first quartile to 43% in the last quartile (mean UPF intake 33.9 ± 14.9%). After binary logistic regression, an association was verified between dietary intake of UPF (>4th quartile) as a percentage of total energy intake and having multiple food allergies (OR 4.102; 95% CI − 1.331 to 12.643; p = .014). We concluded that children and adolescents with FA consumed a higher amount of UPF.


Introduction
Food allergies (FA) affect approximately 10% of children globally and are increasingly prevalent and persistent, often triggering severe allergic reactions like anaphylaxis (Gargano et al. 2021).The onset of FA commonly occurs during early childhood, a critical period for growth and development, when major allergens are important foods in the infant diet (Mehta et al. 2014).
Eight foods account for most diagnoses of FA (peanuts, tree nuts, cow's milk, egg, soy, wheat, crustacean shellfish, and fish), with sesame also recently emerging as a common allergen in the United States (Warren et al. 2019).
The exclusion of allergenic foods remains the primary treatment for FA, potentially associated with nutritional risks (Sova et al. 2013;Sindher et al. 2022).Poorly managed elimination diets in childhood can lead to malnutrition, growth failure, and micronutrient deficiencies.Existing guidelines for FA treatment often lack emphasis on dietary quality, unlike recommendations for the general population, which advise against ultra-processed foods (UPF) (NIAID-Sponsored Expert Panel 2010; Luyt et al. 2014;Ebisawa et al. 2017;Solé et al. 2018a;Matthai et al. 2020;Halken et al. 2021).
A growing increase in UPF intake is associated with the risk of increased incidence and mortality for various types of chronic diseases (Sarmiento-Santos et al. 2022).It has been suggested that a Western diet, typically rich in UPF and consequently in advanced glycation end products (AGEs), influences allergenicity via T lymphocytes (Th2), including interleukins IL-4, IL-5, and IL-13, pro-inflammatory cytokines IL-1, IL-6, and IL-8, TNF-α, and alarmins.Other potential immunological modulatory effects of AGEs include the ability to induce new epitopes of IgE binding, increase inflammatory conditions and oxidative stress, and cause a reduction in the diversity of the intestinal microbiota, leading to increased susceptibility and severity of allergic diseases (Briceno Noriega et al. 2022;Schütte et al. 2022).
Research involving dietary food intake by children and adolescents with FA, using the NOVA classification (Monteiro et al. 2018), with emphasis on the consumption of UPF, can contribute to improving the view of nutritional intervention among this population, which is expanding and at nutritional risk.
Thus, this study aimed to describe the dietary intake of UPF by children and adolescents with FA and to verify a possible association between UPF intake with clinical characteristics and nutritional status in this group.

Study design, population, and ethical approval
This cross-sectional study included 110 children and adolescents, both sex, diagnosed with FA IgE and non-IgE mediated (convenience sample), and treated by excluding one or more foods from the diet.Data collection occurred from January 2019 to December 2020 at FA outpatient service, from low socioeconomic backgrounds, and living in urban areas.The patients were referred by primary care units.
All patients with a clinical diagnosis of FA and/or confirmed by oral food challenge who agreed to participate in the study were included.The exclusion criteria were patients who were exclusively breastfeeding or had concurrent chronic diseases, excluding allergic diseases and obesity.
The University's Research Ethics Committee approved the study, and the Informed Consent Form and Assent Form were applied to parents or guardians and adolescents, respectively.

Evaluation of food consumption according to NOVA classification
The food intake was evaluated based on three 24-h recalls (Rec 24h), collected on non-consecutive days (two weekdays and one weekend).All surveys were conducted during a personal interview by only one researcher responsible for the study (E.C.A.K.).
All foods consumed were classified according to NOVA, which categorises the food according to the degree of processing (Monteiro et al. 2018).Therefore, the foods were categorised into three groups: • "Unprocessed or minimally processed foods" including grains (cereals), legumes (pulses), fresh, dry, or frozen fruits and vegetables, nuts, milk, meat, and other foods and "processed culinary ingredients" like oils and fats, sugar, honey, and salt used to prepare, cook, or season unprocessed or minimally processed foods.(Monteiro et al. 2018).
All the ingredients used were documented for homemade recipes.When obtaining information on ingredients used in homemade recipes was impossible, a recipe manual was used for dietary calculations (Bombem et al. 2012).
All food items reported in the recalls and ingested amounts were recorded in a Microsoft Excel spread-sheet® to quantify energy intake.The homemade measurements were converted into weight units (grams) or volume units (milliliters) (Crispim 2017).For the conversion of food into energy, national tables of food composition were used (IBGE 2011; UNICAMP.2011).The label information was considered for foods not included in the tables mentioned.

Anthropometric evaluation and pubertal development
The anthropometric data collected were weight, length (for children aged 2 and younger), and height (for children above 2 years) (Frisancho 1990;World Health Organization 1995).Anthropometric indicators, including body mass index (BMI) and height-for-age, were calculated using the WHO Anthro® (for children up to 5 years of age) and WHO Anthro Plus® (for children over 5 years of age) to determine z-score.The cut-off points adopted for the classification were those proposed by the World Health Organization (2006).For children under 2 years born preterm, corrected age was applied to assess the adequacy of anthropometric indicators up to two years of age (Silveira 2012).
The classification of the pubertal development stage, using the Tanner scale (Marshall and Tanner 1969;Marshall and Tanner 1970), by the paediatricians of the outpatient clinic where the research was carried out.

Clinical and sociodemographic data
The clinical data collected were gestational age, information on exclusive breastfeeding, and total breastfeeding.Concerning FA information was collected on the immune mechanism involved, age of first reaction, time of illness, use of adrenaline, other associated allergic diseases, and foods excluded.The socio-demographic data were maternal education.

Statistical analysis
The data were registered and consolidated in an Excel spreadsheet (Microsoft Office ®) and analysed by the IBM SPSS 28.0 statistical package.Categorical variables were presented in absolute numbers and percentages and compared by the chi-square test.The distribution of variables was evaluated using the Shapiro-Wilk test, histogram, and kurtosis values.The variables presented parametric distribution, they were presented as mean ± standard deviation and compared using the t-Student test.The variables that showed p < .20 in the bivariate analysis were selected to enter the multiple models.Binary logistic regression was used to evaluate the variables associated with > fourth quartile vs. first, second and third quartiles of ultra-processed food intake (percentage of total energy intake) by children and adolescents with FA.Ten children (9.1%) in breastfeeding were excluded of the analysis.The significance level adopted was 5%.
The detailed and stratified food intake according to the group of foods consumed (unprocessed or minimally processed, processed, and UPF) of all participants and those with cow's milk allergy (CMA) is presented in Tables 1 and 2 of supplementary materials, respectively.The main cow's milk substitutes consumed in Graph 1. distribution of ultra-processed foods (percentage of total energy intake) by children and adolescents with food allergy (n = 100).
The distribution of UPF intake (%) to total energy values (calories) ranged from 21% (first quartile) to 43% (fourth quartile).The mean of UPF intake (%) to total energy values (calories) was 33.9 ± 14.9% (Graph 1).Only one patient did not consume UPF.Among UPF, the major contributors to daily energy intake (excluding formulas and PBS) were biscuits, cakes, sweet bakery goods (4.2% of daily calories), reconstituted meat or fish products (2.7% of daily calories), soft drinks (2.3% of daily calories), and other foods such as ready-to-eat sauces, cream cheese, and breakfast cereal (2.3% of daily calories).
The UPF intake in energy percentage according to the variables studied is presented in Table 3.There was no difference in the mean percentage of energy intake of UPF about clinical variables such as gender, age group, immunological mechanism, disease severity, and anthropometric parameters.However, a trend of lower mean percentage of UPF consumption was observed for individuals with mothers with lower education (29.5 ± 15.4 vs. 35.2± 14.3; p 0.090) and children/adolescents with atopic dermatitis (AD) (28.1 ± 10.5 vs. 35.2±.15.5; p 0.071).
In binary logistic regression, we verified that the percentage of dietary intake energy of UPF (>fourth quartile) was higher in the group of individuals with FA to more than one food (OR 4.102, CI 95% 1.331 to 12.643, p = .014)(Table 4).

Discussion
There are no published studies to date evaluating UPF intake in children and adolescents with FA.Our findings indicate a UPF intake percentage was higher than observed in healthy individuals aged 10 to 18 years in Brazil (26.7% according to the Household Budget Survey -POF 2017-2018) (IBGE 2003), like a study among the North American population aged 2 to 19 years, that UPFs accounted for 67% of total energy consumption (Wang et al. 2021).
The increase in UPF intake over the last three decades among adults has been linked to a rise in mortality from chronic non-communicable diseases.UPF affects human health by complex mechanisms involving the synergistic effects of nutrients and food additives.Its consumption is mainly observed in Western countries and provides higher exposure to AGEs.AGEs, products of Maillard reaction, interact with receptors (RAGE), potentially triggering inflammatory responses associated with diabetes, cardiovascular issues, certain cancers, neurodegenerative conditions, and, recently, the development of food allergies (Briceno Noriega et al. 2022).
AGEs might induce new IgE-binding epitopes, oxidative stress, and reduce gut microbiota diversity, heightening susceptibility to allergies and delaying tolerance acquisition in individuals with FA (Briceno Noriega et al. 2022).Another hypothesis links UPF intake to a possible delay in acquiring food tolerance through interference with the gut microbiota.Butyrate-producing bacteria are crucial, promoting the development of T regulator cells (Treg), essential for oral tolerance (Furusawa et al. 2013;Fujimura and Lynch 2015).
In our study, 28% of participants with FA also had Atopic Dermatitis (AD).Although we did not establish a direct association between UPF intake and AD, we observed a trend suggesting lower UPF consumption among individuals with FA and AD.Globally, the prevalence of AD is on the rise, estimated at 7.1% to 12.5% among Brazilian adolescents (Antunes et al. 2017;Bylund et al. 2020).Studies highlight a correlation between severe AD phenotypes and higher rates of FA, ranging from 33% to 39% (Papapostolou et al. 2022).
The lower consumption of UPFs among individuals with FA and AD might stem from parental concerns about exacerbating the condition due to additives in processed foods.Studies on food additives' impact on skin reactions are inconclusive and affected by methodological gaps (Wilson and Bahna 2005).A recent study population-based cohort study was conducted to evaluate the dietary intake of 10-year-old children and analyses were performed with a focus on AD and FA, allergic diseases known to be affected by foods.Dietary intake was assessed by semi-quantitative food frequency questionnaires.Based on these data, individual dietary intake, as well as children's Dietary Inflammatory Index scores, were calculated.Logistic regression analysis with adjustment for confounding factors showed that having atopic outcomes was associated with a pro-inflammatory dietary pattern at ten years for children with AD [OR = 2.22 (95% CI: 1.14-4.31)]and children with FA in the first ten years of life [OR = 3.82 (95% CI: 1.47-9.93)](Schütte et al. 2022).
The independent association between multiple food allergies and increased UPF intake is notable in our study.The nutritional inadequacy of UPFs, characterised by high salt, added sugar, unhealthy fats, and low fibre and micronutrients, alongside their convenience and widespread availability (Machado et al. 2020), contributes to their overconsumption among families of children with FA.Furthermore, the increase of UPFs catering to various food restrictions, predominantly excluding milk and egg (common allergens in our sample), complicates efforts to limit their intake.Guidelines sometimes recommend UPF substitutes for FA patients (such as infant formulas and PBS), which contributes to increased consumption of UPFs (NIAID-Sponsored Expert Panel 2010; Luyt et al. 2014;Ebisawa et al. 2017;Matthai et al. 2020;Halken et al. 2021).It has been observed in our clinical practice that children with multiple food allergies also consume more formulas and PBS.
The alarming proportion of energy from UPFs among our participants emphasises a fundamental consideration that nutritional interventions for children and adolescents with FA should prioritise dietary quality.Notably, the changing landscape of FA prevalence and persistence warrants attention (Savage et al. 2016;Solé et al. 2018b).Studies also suggest a relationship between UPF consumption and exacerbated allergic diseases (Teodorowicz et al. 2017;Briceno Noriega et al. 2022;Kong et al. 2022).
Table 4. odds ratio of variables associated with greater consumption of ultra-processed foods (Q4 vs. Q1-3) calculated as a percentage of total energy intake by children and adolescents with food allergies (n = 100).dependent variable: ultraprocessed foods dietary intake % of energy ≥ 4 th vs. 1st, 2 nd , and 3 rd quartiles. 1odds ratio adjusted for the presence of atopic dermatitis (yes) and maternal education (years).
Therefore, adherence to nutritional guidelines such as the Food Guide for the Brazilian Population (Brasil 2014) and the Food Guide for Brazilian Children under 2 years (Brasil 2019) becomes imperative, especially for individuals with FA.These guides provide crucial principles for a healthy diet and practical tools to implement them, including strategies to reduce UPF consumption (Mendonça et al. 2021).
This study is the first to examine UPF intake in children and adolescents with FA.However, limitations include a cross-sectional design, single-centre involvement, and the absence of a control group without FA.

Conclusion
This study, involving children and adolescents with FA in treatment at a multidisciplinary centre, showed elevated UPF consumption.There was a correlation between the intake of UPFs and the prevalence of multiple food allergies.Notably, individuals with allergies to more than one food demonstrated a remarkably higher dietary intake of UPFs.These findings reinforce the importance of a multidisciplinary approach to food allergies, highlighting the role of dietitians.Their involvement is key to facilitating the safe and healthy substitution of foods implicated in food allergies.

Figure 1 .
Figure 1.recruitment flowchart of study participants.Source: prepared by the authors *Patients who, during data collection, took oral food challenge with a negative result

• "Processed foods" including
amino acid-based formula, and soy infant formula) and toddler milk, and plant-based substitutes (PBS)

Table 1 .
General characteristics of children and adolescents with food allergy (n = 110).
*** unassociated with use of hypoallergenic formula or plant-based substitute; PBS -plant-based substitutes.

Table 2 .
clinical and nutritional variables of children and adolescents with food allergy (n = 110).

Table 3 .
ultra-processed food intake (percentage of total energy intake) by children and adolescents with food allergy according to the variables studied (n = 100).
1 P -Significance level of the t-Student test.