Anxiety among youth with food allergy

Abstract

Objective

Anxiety is among the major psychological concerns for children living with food allergy (FA). Yet research exploring the variables driving anxiety symptoms in FA remains sparse, and most studies still utilize homogeneous samples to assess anxiety symptoms. The current study seeks to evaluate the rates of clinically significant anxiety symptoms among a diverse sample of youth with FA and examine whether a heightened risk perception of FA outcomes and FA burden (vs. FA medical history) is associated with anxiety in youth.

Methods

94 youth ages 10–14 and their parents were recruited from FA clinics at a mid-Atlantic children’s hospital. Both youth and parents completed demographic and FA medical history questionnaires, the Screen for Child Anxiety Related Emotional Disorders, and the Food Allergy Independent Measure as part of a longitudinal study about FA adjustment and adherence.

Results

Over a third (37%) of youth scored above clinical cut-offs for overall anxiety symptoms. At least 25% of youth reported clinically significant scores on panic disorder, generalized anxiety, social anxiety, separation anxiety, and school avoidance subscales. Perception of risk of adverse FA outcomes and burden—but not FA medical history–were associated with total anxiety, generalized anxiety, panic disorder, and school avoidance symptoms, but not social anxiety and separation anxiety. Having more FAs was associated with higher social anxiety scores but not with other anxiety subscales.

Conclusions

Youth with FA might benefit from psychosocial interventions that address FA risk perception management and promote appropriate FA vigilance to cope with anxiety symptoms.

Food allergy (FA) affects about 8% of children in the United States (Gupta et al., 2018). Primary methods for disease management include daily vigilance for and avoidance of possible allergens in all aspects of life, alertness to symptoms of allergic reactions, and reliable access to rescue medications such as epinephrine auto-injectors (Sicherer & Sampson, 2014). The need for surveillance and avoidance may place children and adolescents living with FA at risk for poor quality of life, including impairment in socioemotional and psychosocial functioning, limited social activities, bullying, anxiety, and stress (Feng & Kim, 2019; Ravid et al., 2012; Teufel et al., 2007).

Among these concerns, anxiety is consistently identified as one of the primary adverse psychosocial experiences of children with FA (Cummings et al., 2010; Ferro et al., 2016; Polloni & Muraro, 2020). Namely, anxiety about and fear of accidental allergen exposure and subsequent allergic reactions can interfere with many aspects of daily life, such as eating outside the home, engaging in social activities, attending school, and traveling (Bollinger et al., 2006). LeBovidge et al. (2009) found that in a sample of children with FA who were 8–17 years old, about 40% scored above average for anxious coping behaviors, about 30% scored above average for separation/panic symptoms, and 17% were above average on tense/restless symptoms, as measured by the Multidimensional Anxiety Scale for Children. In a different study, Shanahan et al. (2014) assessed anxiety symptoms and disorders in children with FA using the Child and Adolescent Psychiatric Assessment structured interview. Similar to LeBovidge et al. (2009), Shanahan and colleagues reported that having FA was significantly associated with separation and generalized anxiety. More specifically, separation anxiety symptoms, such as worrying about potentially harmful events, were reported by more than 6% of children with FA versus 2% in children without FA; symptoms of generalized anxiety, such as restlessness, were reported by more than 15% in children with FA versus 5% in children without FA. Although limited, there is research to suggest that racial and ethnic disparities in anxiety symptoms may exist among children with FA. For instance, a study by Goodwin and colleagues in 2017 demonstrated that among children who identify with a racial/ethnic minority group and low socioeconomic status (SES), having FA was associated with rejection anxiety and social anxiety.

Although anxiety in FA can be adaptive because it may reduce risk-taking behaviors and sustain vigilance to allergens and allergic reactions, high levels of anxiety can be disruptive and maladaptive (Polloni & Muraro, 2020; Ravid et al., 2012; Shanahan et al., 2014). A growing body of research has explored which factors may predispose children with FA to experience elevated and maladaptive anxiety (Mandell et al., 2005). Some researchers have noted that certain FA medical events, such as recent anaphylaxis, treatment with an epinephrine auto-injector, and FA-related hospitalizations may trigger significant anxiety symptoms (Mandell et al., 2005; Polloni & Muraro, 2020; Ravid et al., 2012). However, other research has also identified subjective FA perception as a crucial contributor to FA anxiety. For instance, LeBovidge et al. (2009) reported that patients’ and families’ perception of risk for severe allergic reactions was more predictive of anxiety than FA medical history (e.g., numbers of FAs, recent FA diagnosis, history of anaphylaxis) and that a more negative perception of FA and its impact on daily activities was associated with higher anxiety symptoms (LeBovidge et al., 2009). Other studies have also demonstrated that an overestimated risk perception of fatal anaphylaxis could contribute to clinically significant anxiety levels (Boyle et al., 2017; Hanna et al., 2016). More recently, Newman et al. (2022) suggested that a focus on managing risk perception could improve anxiety about FA, as adolescents can range from being dismissive of risky FA behaviors to expressing a significant fear of allergic reactions, both of which could differentially influence their perceived risk and in turn, their levels of anxiety. Another potential factor influencing anxiety risk is the presence of an asthma diagnosis. FA and asthma have high comorbidity, and asthma attacks may be associated with anxiety similar to allergic reactions. Indeed, there is a high co-occurrence of asthma and anxiety and mood disorders associated with asthma (Friedlander et al., 2013). Further, since asthma is highly prevalent among children who identify as ethnic minorities from low SES (Centers for Disease Control and Prevention, 2022), experiences related to asthma and SES may be particularly relevant to understanding the relationship between FA and anxiety.

Overall, studies examining the risk factors associated with elevated anxiety symptoms in children with FA remain sparse, especially studies that sample an ethnically diverse population. Previous research that investigated the topic of anxiety outcomes either utilized homogeneous convenience samples of mostly non-Hispanic White children (>70%) or lacked detailed demographic breakdowns of participant race and ethnicity, and infrequently included anxiety measures with clinical cut-offs (Fox & Masia Warner, 2017; King et al., 2009; Knibb et al., 2012; Shanahan et al., 2014). Petrovic-Dovat et al. (2016) demonstrated that children with FA did not report more anxiety than their normative counterparts in a primarily White sample. Yet findings from Goodwin et al. (2017) indicated that overall anxiety was significantly higher among ethnically diverse children with FA than children without FA. Clearly, these mixed findings necessitate more inclusive research with diverse samples to thoroughly examine risk factors and the rates of anxiety in the context of FA. As such, the current study seeks to address these gaps in the literature by exploring the factors associated with five domains of anxiety (panic, generalized, separation, social, and school avoidance anxiety) in a racially and ethnically diverse sample of youth with FA.

The current study’s aims are twofold: (1) assess the rates of clinically significant anxiety symptoms among a diverse sample of youth with FA and (2) examine whether anxiety symptoms in youth with FA are most related to a heightened risk perception of FA outcomes or their demographic profile and medical history (i.e., identifying with female gender, having more FAs, and having a recent FA reaction), while controlling for SES and asthma diagnosis as confounds. To generate our aim 1 hypothesis, we conducted a literature review. To the best of our knowledge, only Petrovic-Dovat et al. (2016) evaluated anxiety rates among children with FA using the 41-item Screen for Child Anxiety Related Emotional Disorders (SCARED). The rates of anxiety found in the primarily White sample by Petrovic-Dovat et al. (2016) and the findings from Goodwin et al. (2017) in an ethnically diverse sample of children with FA helped inform our hypothesis for aim 1. Therefore, we hypothesized that percentage rates of total, panic, generalized, separation, social, and school avoidance anxiety symptoms measured via the SCARED will be higher in the current sample as compared to Petrovic-Dovat et al. (2016). In aim 2, we hypothesized that risk perception of allergen exposure and adverse FA outcomes would be significantly associated with anxiety symptoms but not demographic factors and FA medical history.

Methods

Participants and study procedures

Study participants were 94 youth (ages 10–14 years) diagnosed with IgE-mediated FA to at least one of the top eight food allergens in the United States (peanut, tree nut, egg, milk, wheat, soy, shellfish, fish) and a primary caregiver. IgE-mediated FA is characterized as a chronic condition in which patients experience an adverse immune response to specific food proteins when they ingest the food. Symptoms of an allergic reaction arise quickly after ingestion as histamine is released from mast cells throughout the body (Anvari et al., 2019). Youth with a comorbid atopic condition (e.g., asthma, atopic dermatitis) were eligible to participate due to the high rates of co-occurrence with FA. Exclusionary criteria included diagnoses of developmental disorders and/or non-atopic medical conditions as these may impact children’s FA-related psychosocial functioning and adherence.

Prior to data collection, all study procedures were approved by the hospital's Institutional Review Board (IRB Protocol 00007112). Participants were recruited from allergy clinics at a pediatric hospital in the mid-Atlantic region of the United States during a larger study about FA adjustment and adherence from May 2017 to December 2019. Research staff pre-screened allergy clinic lists and identified 474 potentially eligible participants. Of these potential participants, 229 were approached at their in-person appointment or contacted through calls/emails to further screen for eligibility. Of these, 165 youth–caregiver dyads were eligible and completed informed consent/assent with a trained research assistant during their allergy clinic appointments. Post-consenting, youth and their caregivers completed separate questionnaires on paper or online via Research Electronic Data Capture (REDCap; Harris et al., 2009). One hundred three participants provided data on variables specific to the current analyses, and the final sample after excluding missing data was N = 94 participants. This study’s TREND checklist is available as Supplemental File 1. The datasets associated with the current study are available from the corresponding author upon request.

Measures

Demographic and food allergy medical questionnaire

Demographic questions included the child’s age, gender, race, ethnicity, caregiver’s age, and caregiver’s relationship to the child. Caregivers also reported subjective SES (options were 1 = We are poor, 2 = We are almost poor, 3 = We live from check to check, 4 = We live comfortably, 5 = We live very well). An FA medical questionnaire queried information about the child’s age at first FA diagnosis, number of FAs, current FA diagnoses, and when their last allergic reaction occurred (1 = I’ve never had an allergic reaction, 2 = Within the past week, 3 = Within the past month, 4 = Within the past 6 months, 5 = Within the past year, 6 = Within the past 5 years, 7 = Within the past 10 years, 8 = I can’t remember or I’m not sure). Participants also reported whether they had atopic conditions, including atopic dermatitis, asthma, or environmental allergies.

Screen for Child Anxiety Related Emotional Disorders

Youth completed the Screen for Child Anxiety Emotional Disorders (SCARED), a 41-item measure that has consistently demonstrated adequate validity and reliability for screening anxiety disorder symptoms in normative youth and youth with medical conditions (Birmaher et al., 1997, 1999; Hale et al., 2011; Jastrowski Mano et al., 2012). The measure consists of five anxiety subscales: panic or significant somatic symptoms, generalized anxiety, separation anxiety, social anxiety, and school avoidance. Each item describes a feeling or an emotional state, to which participants respond on a 3-point Likert scale (0 = Not true or hardly ever true, 1 = Somewhat true or sometimes true, 2 = Very true or often true). Both domain-specific subscale scores and an overall anxiety score are summed; higher scores indicate more anxiety symptoms. Cronbach’s alpha values for each anxiety subscale were: panic disorder (α = .82), generalized anxiety (α = .86), separation anxiety (α = .83), social anxiety (α = .89), and school avoidance (α = .76). The overall Cronbach’s alpha for the full 41-item scale was α = .94, indicating that in this sample, the scale assessed a composite set of highly related anxiety items.

Food Allergy Independent Measure

To measure the perceived risk of accidental allergen exposure and adverse outcomes, as well as the perceived burden of FA, youth completed the Food Allergy Independent Measure (FAIM) (Van Der Velde et al., 2010). The FAIM has demonstrated strong validity and reliability and has been used in the FA literature to capture the psychosocial burden of FA experienced by patients (e.g., Dunn Galvin & Hourihane, 2018; Warren et al., 2021). The first four items assess the expectation of outcomes, which capture the perception of the risk for allergen exposure, the risk of having a severe allergic reaction, the risk of dying from an allergic reaction, and the ability to engage in a proper reaction treatment. Responses are recorded on a 7-point Likert scale (anchors: 0 = Never—0% chance; 7 = Always—100% chance). Two additional Independent Measure items query how many foods the child must avoid because of FA (anchors: 1 = Almost none; 7 = Almost all) and how great of an impact FA has on the child’s social life (anchors: 1 = Negligibly small; 7 = Extremely great). A mean score is calculated across the six items, with higher scores indicating higher perceived risk and FA burden. The full six items demonstrated adequate reliability among this sample (α = .67); the reliability for each subscale was α = .72 for expectation of outcomes items and α = .60 for independent-measure items.

Statistical analysis

Prior to conducting statistical analyses, assumption of independence was tested and satisfied (Durbin-Watson value = 2.333), indicating that residuals were uncorrelated (Farebrother, 1980). Multicollinearity diagnosis between predictors revealed no significant concerns, with all statistics ≤9.347. Cook’s distance test did not demonstrate potential problems with outliers (residual statistics = 0.05). Assumptions of normality, linearity, and homoscedasticity were assessed and satisfied using a standardized residual histogram, normal P-P Plot, and scatterplot, respectively.

Categorical variables were dummy coded with the following reference categories: no allergic reaction within the past 6 months for recency of FA reaction, male for gender, and no diagnosis for asthma. Subsequently, a hierarchical regression analysis was conducted to examine total anxiety symptoms and five anxiety subscales (panic disorder, generalized anxiety, separation anxiety, social anxiety, school avoidance anxiety) as a function of four predictors—number of FAs, recency of FA reaction, gender, and perceived risk of adverse allergy outcomes/psychosocial burden of FA (i.e., FAIM) while controlling for SES and asthma. We excluded 9 participants out of 103 due to missing data on two key independent variables of gender and recency of FA reaction, resulting in N = 94. Additionally, due to missing data on SES from participants who chose not to provide answers, regression analyses were conducted with available data points from n = 82. All other study statistics were examined using the full N = 94. All statistical analyses were conducted in IBM SPSS Statistics 24.0. Results were deemed statistically significant at p ≤ .05.

Results

Sample characteristics

The final sample included 94 youth (M_age = 11.81 ± 1.30 years) who provided answers to all study variables examined in the statistical analyses. Almost half of participants were female (46.8%) and 40.4% identified as Black or African American. The average number of FAs was 3 ± 1.95, with peanut and tree nuts being the most common FAs. Most participants (73.4%) reported more than one FA diagnosis, 36.2% of participants had an allergic reaction within the past 6 months, and 53.2% had an asthma diagnosis. See Table 1 for full descriptive statistics.

Table 1.

Participant demographic and medical characteristics (N = 94).

	M (SD)	Range
Child age (years)	11.81 (1.30)	10–14
	N (%)
Child gender
Male	50 (53.2)
Female	44 (46.8)
Child race
Black or African American	38 (40.4)
White or Caucasian	26 (27.7)
More than one race	14 (14.9)
Asian or Asian American	7 (7.4)
Prefer not to answer	4 (4.3)
Other	3 (3.2)
Don’t know	2 (2.1)
Child ethnicity
Not Hispanic or Latino	78 (83.0)
Hispanic or Latino	13 (13.8)
Prefer not to answer	2 (2.1)
Don’t know	1 (1.1)
Child asthma diagnosis
Diagnosed	50 (53.2)
Not diagnosed	44 (46.8)
Subjective socioeconomic status
We are poor	1 (1.1)
We are almost poor	1 (1.1)
We live from check to check	25 (27.2)
We live comfortably	43 (46.7)
We live very well	12 (13.0)
Prefer not to answer	10 (10.9)
	M (SD)	Range
Food allergy medical characteristics
Age at food allergy diagnosis (years)	3.30 (2.90)	.25–13
Number of food allergies	3.0 (1.95)	1–11
	N (% Yes)
Allergic reaction in past 6 months	34 (36.2)
Use of epinephrine auto-injector in past 6 months	8 (8.5)
Specific allergens
Tree nuts	71 (76.3)
Peanuts	65 (69.1)
Shellfish	33 (35.5)
Fish	23 (25.0)
Direct egg	22 (23.9)
Sesame	13 (14.0)
Direct cow’s milk	10 (11.0)
Baked egg	10 (10.9)
Soy	8 (8.7)
Wheat	7 (7.5)
Baked cow’s milk	6 (6.5)

Total SCARED score results

See Table 2 for the mean, standard deviation, and clinical cut-off for the total SCARED score, as well as the means and standard deviations for the total FAIM and subscale scores. Among youth, 37.2% met the clinical cut-off of 25 and above for total anxiety symptoms. The final regression analysis examined the relationship between total anxiety symptoms and four predictor variables: number of FAs, recency of FA reaction, gender, and FAIM scores while controlling for asthma and SES. The result was not significant (F [6,75] = 1.422, p = .218), and the model accounted for approximately 10% of the variance in total SCARED scores (R² = .102). FAIM scores emerged as a significant predictor in the final model (B = 0.084, p = .049) and explained an additional 4.8% of the variance in anxiety symptoms (ΔR² = .048). This result suggested that a higher perceived risk of adverse FA outcomes and FA burden was significantly associated with anxiety symptoms, regardless of SES and asthma diagnosis. See Table 3 for all inferential statistics.

Table 2.

Anxiety and food allergy independent measure outcomes: descriptive statistics (N = 94).

	M (SD)	Range	Cut-off	Above clinical cut-off N (%)
SCARED scores
Total score	21.10 (14.27)	0–64	≥25	35 (37.2)
Subscales
Separation anxiety disorder	4.07 (3.53)	0–14	≥5	42 (44.7)
Generalized anxiety disorder	5.90 (4.40)	0–17	≥9	32 (34)
Social anxiety	5.31 (3.95)	0–14	≥8	30 (31.9)
Panic disorder	4.41 (4.05)	0–16	≥7	25 (26.6)
School avoidance	1.97 (2.02)	0–8	≥3	24 (25.5)
FAIM scores
Total score	3.17 (0.98)	1–5.33
Expectation of outcomes	3.41 (1.22)	1–6.75
Independent measure	2.70 (1.22)	0–5.50

Note. SCARED = Screen for Child Anxiety Related Emotional Disorders; FAIM = Food Allergy Independent Measure.

Table 3.

Hierarchical results of total anxiety symptoms and five anxiety subscales (N=82).

Independent variables	B	SE	t	ΔR2
Total SCARED
Step 1				.054
Gender	0.133	0.081	1.639
Number of FAs	0.026	0.021	1.198
Reaction within last 6 months	−0.005	0.085	−0.056
Asthma diagnosis	−0.051	0.085	−0.596
Subjective SES	−0.003	0.056	−0.056
Step 2
Gender	0.115	0.080	1.446
Number of FAs	0.028	0.021	1.328
Reaction within last 6 months	0.029	0.085	0.338
Asthma diagnosis	−0.041	0.084	−0.487
Subjective SES	−0.003	0.055	−0.050
FAIM Score	0.084*	0.042	2.000	0.048*
Total R2 = 0.102
Panic disorder
Step 1				0.086
Gender	2.249*	0.913	2.464
Number of FAs	0.089	0.241	0.371
Reaction within last 6 months	−0.255	0.963	−0.265
Asthma diagnosis	−0.512	0.965	−0.531
Subjective SES	−0.062	0.629	−0.099
Step 2
Gender	2.025*	0.892	2.269
Number of FAs	0.119	0.234	0.508
Reaction within last 6 months	0.185	0.954	0.194
Asthma diagnosis	−0.381	0.939	−0.406
Subjective SES	−0.057	0.611	−0.093
FAIM Score	1.101*	0.471	2.339	0.062*
Total R2=0.148
General anxiety
Step 1				0.104
Gender	2.090*	0.966	2.164
Number of FAs	0.269	0.255	1.057
Reaction within last 6 months	0.784	1.018	0.770
Asthma diagnosis	−1.446	1.021	−1.417
Subjective SES	0.347	0.665	0.521
Step 2
Gender	1.875	0.950	1.973
Number of FAs	0.298	0.249	1.194
Reaction within last 6 months	1.207	1.016	1.188
Asthma diagnosis	−1.320	1.000	−1.320
Subjective SES	0.352	0.650	0.540
FAIM score	1.058*	0.501	2.110	0.05*
Total R2=0.154
Separation anxiety
Step 1				0.007
Gender	0.080	0.838	0.095
Number of FAs	0.007	0.221	0.033
Reaction within last 6 months	−0.311	0.884	−0.352
Asthma diagnosis	0.453	0.886	0.512
Subjective SES	−0.126	0.577	−0.219
Step 2
Gender	−0.008	0.843	−0.010
Number of FAs	0.019	0.221	0.086
Reaction within last 6 months	−0.138	0.902	−0.153
Asthma diagnosis	0.505	0.888	0.569
Subjective SES	−0.124	0.577	−0.215
FAIM Score	0.432	0.445	0.971	0.012
Total R2=0.02
Social anxiety
Step 1				0.063
Gender	0.752	0.926	0.812
Number of FAs	0.504*	0.244	2.065
Reaction within last 6 months	−0.139	0.976	−0.142
Asthma diagnosis	−0.751	0.978	−0.768
Subjective SES	0.086	0.638	0.135
Step 2
Gender	0.664	0.933	0.712
Number of FAs	0.515*	0.245	2.106
Reaction within last 6 months	0.032	0.997	0.033
Asthma diagnosis	−0.700	0.982	−0.713
Subjective SES	0.088	0.639	0.137
FAIM Score	0.429	0.492	0.872	0.009
Total R2 = 0.073
School avoidance
Step 1				0.071
Gender	0.048	0.468	0.104
Number of FAs	0.153	0.123	1.243
Reaction within last 6 months	−0.326	0.493	−0.661
Asthma diagnosis	0.250	0.494	0.505
Subjective SES	−0.453	0.322	−1.408
Step 2
Gender	−0.049	0.462	−0.106
Number of FAs	0.166	0.121	1.371
Reaction within last 6 months	−0.135	0.494	−0.273
Asthma diagnosis	0.306	0.486	0.631
Subjective SES	−0.451	0.316	−1.427
FAIM Score	0.478	0.244	1.963	0.045
Total R2=0.116

Note. SCARED = Screen for Child Anxiety Related Emotional Disorders; FAIM = Food Allergy Independent Measure; SES = Socioeconomic status; FA = Food allergy.

^*Significant at the p ≤.05 level.

SCARED subscale results

See Table 2 for the means, standard deviations, and clinical cut-offs for each SCARED subscale. Approximately 26.6% of youth scored above the clinical cut-off for panic disorder symptoms. The regression analysis examining panic disorder as a function of the number of FAs, recency of FA reaction, gender, and FAIM while controlling for SES and asthma was marginally significant (F [6,75] = 2.169, p = .055) and accounted for 14.8% of the variance (R² = .148). Only female youth (B = 2.025, p = .026) and higher FAIM scores (B = 1.101, p = .022) were associated with more panic disorder symptoms in the final model.

Approximately 34% of youth scored above the clinical cut-off for generalized anxiety symptoms. The regression model in which generalized anxiety symptoms were specified as the outcome was significant (F [6,75] = 2.281, p = .045) and accounted for 15.4% of the variance (R² = .154). Similar to the findings for panic disorder symptoms, female gender (B = 1.875, p = .052) was marginally associated and FAIM scores (B = 1.058, p = .038) were significantly associated with general anxiety symptoms.

About 31.9% of youth scored above the clinical cut-off for social anxiety symptoms. The regression model in which social anxiety symptoms were specified as the outcome was not statistically significant (F [6,75] = .978, p = .446), R² = .073. Among the predictors in the model, only the total number of FA was significantly associated with the outcome, such that having more FAs was positively associated with higher social anxiety symptom scores (B = 0.515, p = .039).

The percentages of youth scoring above clinical cut-offs for separation anxiety and school avoidance were 44.7% and 25.5%, respectively. Regression results regarding separation anxiety and school avoidance symptoms were not statistically significant ps > .1. Only FAIM scores were positively and marginally associated with school avoidance symptoms (B = .478, p = .053). The number of FAs, recent FA reaction, gender, and FAIM scores were not related to separation anxiety symptoms, all ps > .1. See Table 3 for all inferential statistics.

Discussion

Youth with FA often report poor quality of life and anxiety while managing FA, which affects their diet, meal preparation, and participation in school and social activities (Feng & Kim, 2019; Johnson & Woodgate, 2017; Ravid et al., 2012; Teufel et al., 2007). This study evaluated anxiety among a racially and ethnically diverse sample of youth with FA from allergy clinics at a pediatric hospital. Our findings illustrated that youth with FA reported significantly elevated anxiety when assessed by the SCARED, a validated measure of anxiety symptoms. Notable results revealed that over 37% of youth reported clinically elevated total anxiety scores. The following percentages of youth scored above the clinical cut-offs on subscales of panic disorder (26.6%), generalized anxiety (34%), social anxiety (31.9%), separation anxiety (44.7%), and school avoidance (25.5%). Comparatively, one study assessing anxiety in youth with FA from general pediatric care clinics documented the following percentages of clinical elevations on the SCARED: total anxiety (27.3%), panic disorder (32%), generalized anxiety (20.6%), social anxiety (26.8%), separation anxiety (20%), and school avoidance (28.6%) (Petrovic-Dovat et al., 2016).

Except for panic disorder and school avoidance, the pattern of our findings is consistent with our prediction in aim 1, such that a racially and ethnically diverse sample of youth with FA reported higher rates of overall anxiety, generalized, social, and separation anxiety symptoms. The high rates of total anxiety and social anxiety in our sample are consistent with previous research on racial and ethnic minority children with FA by Goodwin et al. (2017). With 44.7% of youth scoring above the clinical threshold for separation anxiety in our sample, this finding bolsters previous literature showcasing separation anxiety symptoms in children with FA (King et al., 2009; LeBovidge et al., 2009; Shanahan et al., 2014). Evidence from our study and Goodwin et al. (2017) suggest that anxiety symptoms are more commonly observed among diverse samples of youth with FA, as opposed to findings from previous studies with homogeneous samples that noted comparable anxiety levels between patients with FA and the general population (LeBovidge et al., 2009; Petrovic-Dovat et al., 2016). These mixed result patterns strongly call for more research with diverse samples to document the psychosocial experiences of minoritized youth with FA. As our study only collected data from children with FA and cannot statistically compare anxiety rates with normative populations, we suggest that future researchers investigate these comparisons using validated measures with clinical cut-offs like the SCARED while ensuring racial and ethnic representation for generalizability.

Our findings also indicate that youth who perceived that they had a greater risk of adverse FA-related outcomes and FA burden reported more anxiety symptoms, regardless of their SES and asthma diagnosis. This relationship has been suggested by other researchers as well (i.e., Chong & Turner, 2019; Dahlsgaard et al., 2020). Notably, the number of FAs and having an allergic reaction in the past 6 months were not related to anxiety scores. Although there is some support in the literature that having more FAs is associated with a worse quality of life (Warren et al., 2016), the number of FAs may have less impact on the patient’s anxiety if the patient had perceived themselves as high-risk for adverse FA outcomes regardless of how many FAs they were diagnosed with. This finding aligns with prior research suggesting that perceptions of FA severity are more related to FA psychosocial outcomes than objective FA markers (Herbert & Dahlquist, 2008). However, it is surprising that having a recent allergic reaction was not associated with anxiety scores. It may be that youth who are anxious about experiencing an allergic reaction engage in more avoidance behaviors, which reduces their chances of having an allergic reaction, or that recently experiencing an allergic reaction diminishes the uncertainty about what an allergic reaction and treatment will be like that could contribute to elevated anxiety.

This study also identified unique variables associated with specific types of anxiety symptoms reported by youth with FA. Greater perceived risk of adverse reaction was associated with higher generalized anxiety and panic disorder scores but not social and separation anxiety scores. It makes sense that risk perception of allergen exposure and allergic reactions would be related to generalized anxiety and panic disorder, as these types of anxiety symptoms focus on overall vigilance of the environment and vigilance to physiological symptoms in the body. It is also worth noting that there is an overlap between symptoms of anxiety and symptoms of an allergic reaction reflected in some SCARED items (e.g., “When I get frightened, I feel like throwing up,” “When I get frightened, I feel like I am choking”). Thus, we are not necessarily interpreting these results to mean that there is a higher prevalence of panic disorder in our sample of children with FA, but rather that symptoms of FA-specific anxiety can mimic panic disorder symptoms and may be limited to FA-related situations. There is a growing body of literature regarding FA-specific anxiety among the FA population, suggesting that some patients only experience anxiety related to FA management (e.g., potential allergic reactions) rather than generalized anxiety symptoms. A deeper understanding of FA-specific concerns has led to recently developed FA-specific anxiety measures (Dahlsgaard et al., 2022; De Holanda Coelho et al., 2021; To et al., 2022). Such assessment tools will help clinicians understand the extent, characteristics, and predictors of FA-specific anxiety, which will aid in differentiating patients with non-condition specific anxiety concerns from those with FA-specific anxiety. Additional research utilizing these newly developed tools is needed to understand their relevance and establish appropriate clinical pathways for patients based on their presenting anxiety symptoms.

Having more FAs was associated with higher social anxiety scores but not generalized anxiety, panic disorder, separation anxiety, or school avoidance. These results have face validity in that regardless of SES and asthma diagnosis, youth who must manage more allergens may be primarily concerned about situations where they face challenges partaking in social meals or eating what their peers are eating, which could lead to social anxiety (Bollinger et al.,2006). We believe that social anxiety pertaining to FA is a newly identified component of FA-specific anxiety. Although skills to navigate social situations are frequently cited as an area of need for youth with FA (Engel et al., 2022; Herbert et al., 2023), these skills are not assessed using new measures of FA-specific anxiety and hence may be missed if youth with FA are only screened for anxiety with a generalized anxiety/panic disorder framework or anaphylaxis phobia framework in mind. Thus, this FA-related social anxiety phenomenon warrants additional research, measure development, and intervention development. Further, our findings suggested a trend where the perceived risk of adverse FA outcomes was related to school avoidance symptoms. However, since the SCARED school avoidance items assess the general school experience rather than food-related events, we might not fully capture the above association in youth with FA. Neither perceptions of risk of adverse outcomes nor the number of FAs were related to separation anxiety. This null finding might be because the SCARED separation anxiety items pertain to something negative happening to a loved one or being apart from caregivers at night. Yet in our experience, youth with FA often anecdotally report concerns about being without a trusted caregiver in food-related situations. Altogether, these observations further support the development and use of FA-specific anxiety measures alongside the SCARED to fully capture the spectrum of anxiety symptoms and associated risk factors.

When considering the clinical implications of these findings, it may first be appropriate for clinicians to note that more youth with FA presenting for care in an allergy or pediatric clinic may be experiencing more anxiety than youth without FA. As such, clinicians should be prepared to assess for and provide resources related to anxiety management, particularly anxiety about FA risk and social situations. These resources could include proactive education for all youth and their caregivers about the intersection of anxiety, FA, and asthma, as well as specific screening for clinically elevated anxiety during routine allergy appointments. It is also crucial for clinicians to note that some degrees of anxiety are adaptive for youth with FA because they promote appropriate vigilance to allergens in the environment and awareness of allergic reaction symptoms (Dahlsgaard et al., 2020). Thus, not every youth reporting FA anxiety needs to participate in formal mental health treatments. Some youth may benefit from education about what is or is not adaptive anxiety and ways to manage this anxiety when needed. Yet high anxiety levels may predispose youth to engage in overly restrictive behavior, which could reduce quality of life, or avoidance of disclosing FA to others, which could increase the risk for allergic reactions (Polloni & Muraro, 2020). Clinicians are encouraged to use the SCARED alongside FA-specific anxiety measures to help clarify whether patients are balancing vigilant safety behaviors with engaging in age-appropriate activities or experiencing anxiety symptoms that warrant further intervention. Some research suggests that evidence-based anxiety treatments can be adapted to support FA-specific needs. Studies using cognitive behavioral therapy with exposure and response prevention (e.g., being near an allergen, eating safe foods, reducing hand washing) have shown promising anxiety reduction in youth with FA (Dahlsgaard et al., 2023). Additionally, FA education with direct exposure to allergens (e.g., smelling and touching specific allergens) has also led to reduced fear (Weinberger et al., 2019). The promising outcomes from these interventions suggest that balancing risk and burden perceptions with appropriate FA management could mitigate excessive and impairing anxiety symptoms.

Strengths of this study include that this was a racially and ethnically diverse sample of youth with documented physician-diagnosed FA. Youth responded to questionnaires by themselves as opposed to parent-proxy, which may be particularly important when assessing internalizing disorder symptoms. The use of a validated anxiety assessment tool like the SCARED allowed for a detailed analysis of different types of anxiety symptoms, creating more nuanced examinations of the relationship between anxiety and FA characteristics. However, our current study has limitations. Some participants preferred not answering the SES question, thus making our regression results on 82 cases potentially underpowered and difficult to definitively interpret null findings. As this is a cross-sectional study, the directionality of the observed associations is unknown. Anxiety symptoms might predict FA burden rather than FA burden predicting anxiety. Therefore, future research studies would benefit from examining anxiety and FA burden longitudinally among youth with FA to better understand the development of FA-related anxiety and its intersection with FA burden and to ascertain appropriate intervention points and content. Finally, although our analyses accounted for asthma diagnosis, future studies may consider also accounting for environmental allergies and eczema when assessing the relationship between FA risk factors and anxiety outcomes, given the common comorbidity of these diagnoses in youth with FA.

In sum, anxiety among youth with FA is common, and findings from the current study provide evidence that youth with FA experience an array of anxiety symptoms. Although standardized measures of anxiety help assess some of these symptoms, this study highlights that they cannot adequately capture the nuances of anxiety in youth with FA because they lack FA-specific questions. There may be two different components of FA-specific anxiety—fear of allergen exposure and anaphylaxis, which has been the predominant type of FA anxiety discussed in the literature thus far—and social anxiety, which warrants further research. Interventions that harness evidence-based anxiety management techniques, such as cognitive-behavioral therapy, to promote balanced perceptions of FA risk and participation in school/social activities are likely to result in decreased FA burden and FA-specific anxiety, overall improving quality of life for youth with FA.

Supplementary material

Supplementary material is available online at Journal of Pediatric Psychology (https://academic.oup.com/jpepsy/).

Author contributions

Sally Ho (Conceptualization [lead], Formal analysis [lead], Visualization [lead], Writing—original draft [lead], Writing—review & editing [equal]), Frances Cooke (Conceptualization [equal], Data curation [equal], Project administration [lead], Writing—original draft [supporting], Writing—review & editing [supporting]), Ashley Ramos (Conceptualization [supporting], Writing—review & editing [supporting]), Elizabeth L McQuaid (Writing—review & editing [supporting]), Hemant Sharma (Writing—review & editing [supporting]), and Linda Jones Herbert (Conceptualization [equal], Funding acquisition [lead], Investigation [lead], Methodology [lead], Writing—original draft [equal], Writing—review & editing [equal])

Funding

This work was supported by the National Institute of Allergy and Infectious Diseases (5K23AI30184-02) as awarded to Linda J. Herbert.

Conflicts of interest

None declared.