**2. Materials and Methods**

We performed a retrospective observational analysis of the clinical charts of children with a history of nut reactions who were referred to the allergy unit of the hospital from January 2015 to December 2019. Written informed consent for all performed procedures was obtained from the children's parents. The code of the event report issued by the hospital is IR904-18-26854.

A skin prick test (SPT) for aeroallergens (including grass, artemisia, cypress, olive tree, hazel, birch, and poplar) and/or specific foods if the clinical history was suggestive of respiratory allergy (asthma and/or oculorhinitis) and food allergy was performed by using commercial extracts (Lofarma, Milan, Italy). When an SPT for specific foods was not available, we performed a PbP with fresh foods. Fresh nuts were used for PbPs, according to Ortolani et al. [16]. Both SPTs and PbPs were performed on the volar surface of the forearm with a lancet, as per the European standard [17]. The results were read after 15 min: a wheal diameter ≥ 3 mm was considered positive. Positive and negative controls were used—histamine (10 mg/mL; Lofarma, Milan, Italy) and normal saline, respectively.

In patients with a positive PbP to nuts, a s-IgE to nuts and the available molecular components (peanuts: (Ara h 1, Ara h 2, Ara h 3, Ara h 8, Ara h 9); hazelnuts: (Cor a 1, Cor a 8, Cor a 9, Cor a 14); walnuts: (Jug r 1, Jug r 3)) were detected by ImmunoCAP (Thermo Fisher Scientific, Uppsala, Sweden), following the manufacturer's instructions. A positive cut-off point was set at 0.1 kUA/L.

All the patients with positive tests in reference to the culprit nut underwent an OFC with the nut suspected as the cause for the reaction or beginning with the nut suspected as the cause for the first reaction in chronological order in the case of multiple reactions (according to clinical history and sensitization profile).

The OFC was performed under an allergist's supervision, and it was carried out according to international standards [18,19] adapted to the context of a one-day hospital setting [20]. The OFC was usually proposed to healthy children and postponed in case of acute diseases like fever, infectious gastroenteritis, or bronchitis. The protocol used for the nuts OFC is summarized in Table 1. The increasing doses were given every 20 min until completing the protocol or reaching the threshold dose for reaction. The OFC was considered positive if there were objective signs like urticaria, angioedema, vomiting, diarrhea, bronchospasm, hoarse voice, rhinitis, conjunctivitis, hypotension, or loss of

consciousness within two hours after administration of the last food dose, which is the frame time of observation in the hospital setting for IgE-mediated food allergies. If there were reactions, patients were treated as needed and observed for a minimum of two hours until the clinical manifestations of the reaction resolved.


**Table 1.** Nut oral food challenge protocol.

According to Niggeman's classification, the OFC clinical manifestations were classified as mild, moderate, and severe [21]. Moreover, for any reaction, we described the threshold dose of the culprit nut (the maximum tolerated dose during the OFC) and its corresponding dose of nut protein. The same classification was used to define the severity of the reaction in the clinical history.

Finally, the studied population was represented by children with confirmed nut allergies, comprising a clinical history of nut reactions, plus positive PbP and/or s-IgE to nut plus positive nut OFC. The diagnostic selection of children with nut allergies is shown in Figure 1. Subsequently, the patients were divided into groups according to the nut responsible for the first reaction in chronological order (according to clinical history). Then, demographic characteristics (gender, age, age at first nut reaction), coexisting allergic diseases (atopic dermatitis, asthma, rhinitis, another food allergy), familiar history of allergy, values of PbP to nut and s-IgE to nut, characteristics of OFC, and clinical manifestations of the first reaction to the nut were extrapolated through chart review. Finally, these characteristics were compared between the various nuts.

Statistical analyses were performed using OpenEpi (version 3.01; Atlanta, GA, USA), Microsoft Excel (2013 version, Redmond, WA, USA), and SPSS (IBM SPSS Statistics 22, Chicago, IL, USA). Qualitative data were expressed as counts and percentages; quantitative data were expressed as mean value ± standard deviation or median value and minimum– maximum value. Data distribution was verified with the Shapiro–Wilk test and equality of variance with the Hartley F-test. Differences between continuous variables were calculated using the Student's t-test and the one-way analysis of variance (ANOVA). Associations between categorical variables were obtained with Fisher's exact test and chi-square test. A *p*-value of less than 0.05 was considered statistically significant.

**Figure 1.** Flow chart used for the diagnostic selection of children with a confirmed nut allergy. Legend: N, number; OFC, oral food challenge; PbP, prick by prick; s-IgE, serum-specific IgE. **Figure 1.** Flow chart used for the diagnostic selection of children with a confirmed nut allergy. Legend: N, number; OFC, oral food challenge; PbP, prick by prick; s-IgE, serum-specific IgE.

### Statistical analyses were performed using OpenEpi (version 3.01; Atlanta, GA, USA), **3. Results**

in Table 2.

Microsoft Excel (2013 version, Redmond, WA, USA), and SPSS (IBM SPSS Statistics 22, Chicago, IL, USA). Qualitative data were expressed as counts and percentages; quantitative data were expressed as mean value ± standard deviation or median value and minimum–maximum value. Data distribution was verified with the Shapiro–Wilk test and equality of variance with the Hartley F-test. Differences between continuous variables were calculated using the Student's t-test and the one-way analysis of variance (ANOVA). Associations between categorical variables were obtained with Fisher's exact test and chisquare test. A *p*-value of less than 0.05 was considered statistically significant. We reviewed the clinical charts of 318 children (201 males (63%) and 117 females (37%)) with a clinical history of nut reaction to the following nuts according to the first reaction in chronological order: hazelnut (112; 35%), walnut (90; 28%), peanut (58; 18%), pine nut (30; 9%), cashew (14; 4%), pistachio (8; 3%), and almond (6; 2%). No patients reported reactions to macadamia, pecan, or Brazil nuts. Overall, 184 subjects had a positive PbP and/or s-IgE to nuts and underwent OFC with the culprit nut. The OFC was negative in 71 children (39%). Conversely, 113 subjects (61%) had a positive OFC and, therefore, a nut allergy was confirmed. The result of the OFC was independent of the type of nut tested (*p* = 0.10).

**3. Results**  We reviewed the clinical charts of 318 children (201 males (63%) and 117 females (37%)) with a clinical history of nut reaction to the following nuts according to the first reaction in chronological order: hazelnut (112; 35%), walnut (90; 28%), peanut (58; 18%), pine nut (30; 9%), cashew (14; 4%), pistachio (8; 3%), and almond (6; 2%). No patients The demographic characteristics of 113 children with a confirmed nut allergy are reported in Table 2. According to the nut responsible for the first reaction considered in chronological order, we identified six different groups of patients: cashew (4; 4%), hazelnut (43; 38%), peanut (22; 19%), pine nut (11; 10%), pistachio (1; 1%), and walnut (32; 28%). The demographic characteristics of patients divided for the different nuts are summarized in Table 2.

reported reactions to macadamia, pecan, or Brazil nuts. Overall, 184 subjects had a positive PbP and/or s-IgE to nuts and underwent OFC with the culprit nut. The OFC was negative in 71 children (39%). Conversely, 113 subjects (61%) had a positive OFC and, therefore, a nut allergy was confirmed. The result of the OFC was independent of the type of nut tested (*p* = 0.10). The demographic characteristics of 113 children with a confirmed nut allergy are reported in Table 2. According to the nut responsible for the first reaction considered in chronological order, we identified six different groups of patients: cashew (4; 4%), hazelnut (43; 38%), peanut (22; 19%), pine nut (11; 10%), pistachio (1; 1%), and walnut (32; 28%). There were no statistically significant differences in gender, concurrent allergic diseases, familiar history positive for allergic diseases, or mean values of PbP and s-IgE to nuts between the various nuts (Table 2). Conversely, the difference in mean age at the first nut reaction was statistically significant (*p* = 0.00017) for the pine nut group, which was higher (8.6 ± 3.7 years) than the mean age at the first reaction to the other nuts (hazelnut 3.7 ± 3.2 years, *p* = 0.00008; peanut 5 ± 3.4 years, *p* = 0.009; walnut 4.2 ± 2.7 years, *p* = 0.0001) except cashew *(p* = 0.97). Other differences in mean age were confirmed for the cashew (8.5 ± 5.9 years) and the hazelnut groups (3.7 ± 3.2 years; *p* = 0.012) but without evidence of statistical relevance with the remaining nuts.

The demographic characteristics of patients divided for the different nuts are summarized


**Table 2.** Demographic characteristics of patients with a nut allergy.

Legend: AD, atopic dermatitis; FA, food allergy; Max, maximum; Min, minimum; N, number; PbP, prick by prick; s-IgE, serum-specific IgE; SD, standard deviation; %, percentage.

> Seventy-nine subjects (70%) denied other allergic food reactions, concurrent or previous, based on their clinical history. Among the 34 children with food co-allergies (30%), the most frequent was egg (17; 50%), followed by milk (14; 41%), fresh fruit (13; 38%), both egg and milk (11; 32%), fish/clams (9; 26%), cereals (3; 9%), legumes (2; 6%), and seeds (1; 3%). The frequency of food co-allergies did not differ between the six nuts groups (Table 2). As regards nut co-allergies, over 75% of subjects in every group, except for the pistachio one, reported a single nut reaction (Table 3), without statistical difference between groups (*p* = 0.45).

**Table 3.** Nut co-allergy.


Legend: N, number; %, percentage.

Fifty-nine patients (52%) referenced a history of respiratory allergy (asthma and/or oculorhinitis). Grass pollen allergy was the most frequent among the pollen species tested (40; 68%), followed by cypress and birch (23; 59%).

Furthermore, 54 patients out of 113 (48%) had a history of anaphylaxis to nuts as the first reaction in chronological order: 25 patients (46% of anaphylaxis) reported moderate reactions while 19 (35% of anaphylaxis) reported severe ones. In 10 patients (19% of anaphylaxis), the severity of anaphylaxis at the first nut reaction was unknown. The occurrence of anaphylaxis (*p* = 0.16) and its severity at the first reaction in chronological order (moderate *p* = 0.77; severe *p* = 0.10) were not statistically different between the various nuts. The PbP and s-IgE values did not differ between the various nuts according to the severity of the first nut reaction. Moreover, in case of positive OFC, they did not differ among nuts according to the severity of the reaction (Table 4).

**Table 4.** Prick by prick and serum-specific IgE levels to the respective nuts, severity of the first nut reaction, and severity of the positive oral food challenge.


Legend: N, number; PbP, prick by prick; s-IgE, serum-specific IgE; SD, standard deviation.

A complete molecular analysis was performed in 62 out of 97 eligible patients (64%): The highest adherence was obtained in the peanut group (86%), followed by the hazelnut (63%) and then walnut (50%) groups. Thus, the available molecular components were detected at least in 50% of the eligible population (Table 5). The mean values of molecular allergens did not correlate with the severity of the first nut reaction. Moreover, in case of positive OFC, they did not differ according to the severity of the reaction (Table 5).

**Table 5.** Molecular allergens, severity of the first nut reaction, and severity of the positive oral food challenge.


Legend: Max, maximum; Min, minimum; N, number; SD, standard deviation; %, percentage.

During the first nut reaction, cutaneous involvement was the most frequent reaction (70; 62%), followed by gastrointestinal (44; 39%) and then respiratory clinical manifestations (25; 22%). Seven children (6%) reported signs and symptoms by contact. No one presented neurological involvement. Instead, five children (4%) referred to cardiovascular manifestations, described only in patients with anaphylaxis. During the OFCs, the signs and symptoms involved mainly the gastrointestinal system (82; 73%), followed by the cutaneous (42; 37%) and respiratory systems (28; 25%). No one presented cardiovascular or neurological involvement. The system involvement during the first nut reaction and the OFC is reported in Figure 2. The clinical features of the reactions did not differ between the various nuts (Figure 2), except for cutaneous involvement at the first nut reaction (*p* = 0.011), referred to mainly in the hazelnut and walnut groups.

**Figure 2.** System involvement during the first nut reaction and the positive oral food challenge. **Figure 2.** System involvement during the first nut reaction and the positive oral food challenge.

> In total, 91 out of 113 subjects (81%) with clinical manifestations during OFC presented a mild reaction with single-system involvement. The remaining 22 subjects had anaphylaxis: moderate in 13 children (11%) and severe in 9 children (8%). Table 6 shows In total, 91 out of 113 subjects (81%) with clinical manifestations during OFC presented a mild reaction with single-system involvement. The remaining 22 subjects had anaphylaxis: moderate in 13 children (11%) and severe in 9 children (8%). Table 6 shows the severity of

the reaction during the OFC. The severity of the reaction did not depend on the type of nut tested (Table 6).


**Table 6.** Severity of reaction of the positive oral food challenge and dose of nut ingested.

Legend: N, number; %, percentage.

The mean dose of nut proteins ingested was 283 mg ± 630 mg (range 2–4500 mg) in mild reactions during OFC, 376 mg ± 686 mg (range 20–2500 mg) in moderate reactions, and 69 mg ± 70 mg (range 5–200 mg) in severe ones (Table 6). The dose of nut proteins ingested was significantly lower in severe reactions if compared with the mild ones (*p* = 0.003) but not with the moderate ones (*p* = 0.13). Furthermore, there were no statistically significant differences between the severity of reaction during the OFC and the mean dose of proteins ingested between the different nuts (mild, *p* = 0.65; moderate, *p* = 0.36; mild, *p* = 0.64) (Table 6).

Furthermore, 88 out of 113 children (78%) with a positive OFC showed a spontaneous resolution of signs and symptoms. The remaining 25 subjects (22%) with reactions during the OFC required the administration of therapy: 19 (76%) were treated with oral antihistamines, 20 (80%) with oral corticosteroids, and 3 (12%) with inhaled bronchodilators. None needed the administration of injectable epinephrine. None required hospitalization or intensive care assistance. The need for therapy did not depend on the type of nut tested (*p* = 0.53).
