Christina B. Johns

Food Allergy: Epidemiology and Natural History

The prevalence of food allergy is rising for unclear reasons, with prevalence estimates in the developed world approaching 10%. Knowledge regarding the natural course of food allergies is important because it can aid the clinician in diagnosing food allergies and in determining when to consider evaluation for food allergy resolution. Many food allergies with onset in early childhood are outgrown later in childhood, although a minority of food allergy persists into adolescence and even adulthood. More research is needed to improve food allergy diagnosis, treatment, and prevention.

Elevated total serum IgE in nonatopic patients with aspirin-exacerbated respiratory disease.

Background Aspirin-exacerbated respiratory disease (AERD), also known as Samters triad, is characterized by asthma, recurrent nasal polyps, and by allergic-like reactions to aspirin and other nonsteroidal anti-inflammatory drugs, although it is not a true immunoglobulin E (IgE)–mediated allergy. Atopy, although common in patients with AERD, is not a characteristic of the disease. Recently, we have observed a subgroup of patients with AERD who have no history of atopy but have abnormally elevated total serum IgE, a phenomenon that has been observed in patients with asthma has not been further explored. We sought to explore this phenomenon of elevated total serum IgE in the absence of atopy in a subset of patients with AERD. Methods Patients were diagnosed with AERD with an oral aspirin challenge at the Brigham and Womens Hospital Allergy and Clinics. Atopy was defined as a positive test result to at least one of the common aeroallergens. Elevated total serum IgE was defined as IgE of >100 I

Tolerance of daily low-dose aspirin does not preclude aspirin-exacerbated respiratory disease.

U/mL. Results We present six patients with AERD and elevated total serum IgE in the absence of any clear atopy. Total serum E in the patients ranged from 110 to 1760 I

Automated identification of an aspirin-exacerbated respiratory disease cohort

U/mL. Mean blood eosinophil levels for these patients were not significantly different from those of the entire cohort of patients with AERD included in the study. Conclusion In a subset of patients with AERD, we observed elevated total serum IgE even when atopy was not present. To better understand the disease, the cause and clinical relevance of this phenomenon deserves further exploration.

Bacterial metabolites of diet-derived lignans and isoflavones inversely associate with asthma and wheezing

Aspirin-exacerbated respiratory disease (AERD) is characterized by asthma, eosinophilic nasal polyposis, rhinosinusitis, and respiratory reactions to COX-1 inhibitors including aspirin and other nonsteroidal anti-inflammatory drugs. Patients with AERD with symptoms refractory to guideline-based management of asthma, nasal polyposis, or rhinosinusitis benefit from aspirin desensitization and treatment with high-dose aspirin. Daily doses of at least 300 mg of aspirin, with some patients requiring up to 1300 mg, are effective in reducing the burden of disease. Studies investigating daily doses of lower than 300 mg have yielded inconsistent results, with the weight of evidence suggesting that low doses of aspirin are not effective. Clinical observation suggests that a subset of patients with AERD report taking daily low-dose (81 mg) aspirin for months to years without clinical history of an adverse reaction to aspirin. These patients come to attention because of recurrent eosinophilic nasal polyposis. When they undergo oral aspirin challenge after holding their daily aspirin for 10 days, they develop stereotypical respiratory reactions. To our knowledge, such patients with AERD have not been previously described in the literature. The goal of the present study was to identify a group of these patients and investigate their characteristics, with attention to their response to treatment with high-dose aspirin. We performed a retrospective chart review of 163 subjects with AERD who were referred to Brigham and Women’s Hospital or Massachusetts General Hospital for evaluation of possible AERD and who agreed to participate in our AERD registry between September 2013 and August 2014. We identified subjects who were taking 81 mg aspirin daily at the time of AERD evaluation and who reported no clinical history of aspirin hypersensitivity. All subjects who were taking daily low-dose aspirin at the time of AERD evaluation stopped taking aspirin for at least 10 days and then underwent aspirin challenge. The electronic medical record was reviewed for all subjects. Comprehensive clinical and questionnaire data were available for 91 subjects. There were 7 subjects (4.3% of 163) with AERD who had tolerated chronic daily aspirin (81-mg aspirin group). Six of the 7 subjects underwent positive aspirin challenge tests at our institution and developed a characteristic upper and/or lower airway reaction; the seventh reported a positive nonsteroidal anti-inflammatory drug challenge done elsewhere. Their demographic characteristics are given in Table I, and aspirin challenge results are given in Table II. The dose of aspirin that provoked a clinical reaction was 81 mg or less in 5 of the 6 patients. Subjects in the 81 mg aspirin group were significantly older at the time of aspirin challenge than other subjects with AERD (59.6 vs 48.3 years; P < .05), though there was no significant difference between groups in time between subjectreported onset of nasal polyposis and date of aspirin challenge. There was no significant difference between groups in the number of lifetime polypectomies, but fewer subjects in the 81mg aspirin group had a clinical history of asthma (5 of 7 vs 83 of 84; P < .05). Of the subjects in the 81-mg aspirin group, one started taking daily low-dose aspirin before the onset of AERD symptoms and another started taking daily low-dose aspirin before the onset of nasal polyps but had had mild intermittent asthma since childhood. Five had symptoms of AERD at the time they started taking daily low-dose aspirin. Of these, 3 were not taking montelukast and 2 could not recall their use of montelukast when they initiated daily low-dose aspirin. Four subjects in the 81-mg aspirin group were desensitized to aspirin and began high-dose aspirin therapy of at least 325 mg twice daily, and have been followed for a mean of 26 months (range, 5-37 months). Since beginning high-dose aspirin, none has required repeat polypectomy and all report improvement in nasal symptoms. Those with asthma (n 1⁄4 3) report improvement in asthma symptoms, with increases in FEV1 of 12.2%, 15.3%, and 41.4% at the first visit after the initiation of high-dose aspirin, which occurred 6, 12, and 6 weeks after aspirin desensitization, respectively. The patient whose FEV1 increased by 41.4% restarted zileuton, which he had temporarily stopped taking, around the same time that he started taking high-dose aspirin. Of the remaining 3 subjects in the 81-mg aspirin group, 2 will pursue polypectomy before aspirin desensitization and 1 has elected not to pursue highdose aspirin therapy. There are several explanations for why some patients with AERD apparently tolerate daily low-dose aspirin. First, they may initiate daily aspirin early in the clinical course of AERD before they develop aspirin hypersensitivity. However, only 1 subject in our group began taking baby aspirin before the onset of AERD symptoms. Second, the use of montelukast at the time of aspirin initiation could blunt their reaction, producing a “silent desensitization,” though most of the subjects we studied were not taking montelukast when they began taking daily low-dose aspirin. Given our findings, the most likely explanation is that they develop a reaction to their first ingestion of low-dose aspirin, but fail to connect the reaction with their aspirin use and subsequently become desensitized to aspirin through daily use. Although asthma is a prominent clinical feature of AERD, it is possible to have AERD without asthma. Subjects in our 81-mg aspirin group had significantly lower prevalence of asthma than did subjects who had not been taking aspirin daily. Patients with AERD who have more severe asthma may be less likely to tolerate daily low-dose aspirin without clinically obvious

Access to health care and food in children with food allergy

Background: Aspirin‐exacerbated respiratory disease (AERD) is characterized by 3 clinical features: asthma, nasal polyposis, and respiratory reactions to cyclooxygenase‐1 inhibitors (nonsteroidal anti‐inflammatory drugs). Electronic health records (EHRs) contain information on each feature of this triad. Objective: We sought to determine whether an informatics algorithm applied to the EHR could electronically identify patients with AERD. Methods: We developed an informatics algorithm to search the EHRs of patients aged 18 years and older from the Partners Healthcare system over a 10‐year period (2004‐2014). Charts with search terms for asthma, nasal polyps, and record of respiratory (cohort A) or unspecified (cohort B) reactions to nonsteroidal anti‐inflammatory drugs were identified as “possible AERD.” Two clinical experts reviewed all charts to confirm a diagnosis of “clinical AERD” and classify cases as “diagnosed AERD” or “undiagnosed AERD” on the basis of physician‐documented AERD‐specific terms in patient notes. Results: Our algorithm identified 731 “possible AERD” cases, of which 638 were not in our AERD patient registry. Chart review of cohorts A (n = 511) and B (n = 127) demonstrated a positive predictive value of 78.4% for “clinical AERD,” which rose to 88.7% when unspecified reactions were excluded. Of those with clinical AERD, 12.4% had no mention of AERD by any treating caregiver and were classified as “undiagnosed AERD.” “Undiagnosed AERD” cases were less likely than “diagnosed AERD” cases to have been seen by an allergist/immunologist (38.7% vs 93.2%; P < .0001). Conclusions: An informatics algorithm can successfully identify both known and previously undiagnosed cases of AERD with a high positive predictive value. Involvement of an allergist/immunologist significantly increases the likelihood of an AERD diagnosis.

Urinary triclosan levels and recent asthma exacerbations

To the editor: Lignans and isoflavones are plant-derived chemicals with potent anti-inflammatory and anti-oxidant effects1. Lignans are ubiquitous in Western diets with flaxseeds containing the greatest amounts, while isoflavones are most abundant in soybeans (Table E1). The end product(s) of gut microbial metabolism of lignans is enterolactone2; metabolism of the isoflavone daidzein gives rise to equol and o-desmethylangolensin (O-DMA). Human studies demonstrate inverse associations between lignans and isoflavones and hormone-dependent cancers, cardiovascular disease, and osteoporosis1, and improvements in asthma control are reported in those with high soy consumption3. Experimental studies suggest that bacterial metabolites of isoflavones attenuate allergic airway inflammation,4 however similar studies are unavailable for enterolactone. Prospective human studies demonstrate that microbial exposures are important for asthma development and control5, but the mechanism of this association is unclear. Given their anti-inflammatory and anti-oxidative properties, we hypothesized that bacterial metabolites of lignans and isoflavones may mediate the beneficial effects of bacteria on asthma and lower airway disease. To provide evidence supporting our hypothesis, we conducted a cross-sectional population-based study using data from the National Health and Nutrition Examination Survey (NHANES), which collected urinary levels of lignan and isoflavone metabolites and medical history of asthma and wheeze. Data were obtained from the 2003–2004, 2005–2006, 2007–2008, and 2009–2010 NHANES. 9,633 subjects ages 6–85 had complete data for the primary analyses of the association between urinary levels of lignan and isoflavone metabolites (e.g. enterolactone, equol, and O-DMA) and physician-diagnosed current asthma and self-reported non-asthmatic wheeze, adjusting for age, gender, race/ethnicity, log-transformed urinary creatinine level, poverty index ratio (PIR), and body mass index (BMI). We performed three different sensitivity analyses: A. adjusting for smoke exposure (we adjusted for presence of household smoke exposure among subjects <20 years old and never, former, or current smoking status among subjects ≥20 years old), B. adjusting for plant intake (using total dietary fiber intake), and C. stratifying each analysis by atopic status (presence or absence of allergen-specific IgE, available in 2005–2006 only). Statistical analyses were performed with STATA12.0 (College Station, TX) using the sampling and weighting variables provided by NHANES. See this articles Online Repository for additional methods (www.jacionline.org). Participants reporting current asthma were more likely to be female, younger, African American, aeroallergen sensitized, and have a PIR less than 1. Those reporting non-asthmatic wheeze were more likely to be older and Caucasian. Current asthmatics were more likely to report being former smokers, and those with non-asthmatic wheeze were more likely to be current smokers or have household smoke exposure. Asthma and wheeze status was positively associated with BMI (Table E2). Urinary levels of bacterial metabolites by tertile are presented in Table E3. These levels varied by race/ethnicity, PIR, age, BMI category, and tobacco exposure status (Tables E4–E8). NHANES participants were selected at random for urinary lignan and isoflavone analysis. The sample with data available on urinary bacterial metabolite measurements was similar to the general NHANES population, although more adults (82% vs. 72%, P = <0.001), and more subjects with BMI greater than 30 (29% vs. 26%, P = <0.001) composed the subsample (Table E9). Urinary enterolactone levels were inversely associated with current asthma (Table I, Figure 1). In logistic regression models adjusted for race/ethnicity, PIR, gender, age, creatinine, and BMI, the odds ratio (OR) for current asthma for the highest tertile of urinary enterolactone was 0.69 (95% CI 0.56–0.85). Levels of urinary enterolactone and O-DMA were inversely associated with non-asthmatic wheeze. The OR for non-asthmatic wheeze for the highest tertile of urinary enterolactone was 0.53 (95% CI 0.42–0.67). The OR for non-asthmatic wheeze for the highest tertile of urinary O-DMA was 0.64 (95% CI 0.52–0.80). In a model including all metabolites, urinary enterolactone remained independently inversely associated with current asthma, while both enterolactone and O-DMA remained independently inversely associated with non-asthmatic wheeze (Table E10). Figure 1 Predicted probability of current asthma (A), and non-asthmatic wheeze (B) with 95% CIs by urinary enterolactone levels. Table I Odds ratios (and 95% CI) for current asthma and non-asthmatic wheeze by level of urinary metabolites of lignans and isoflavones In sensitivity analyses, the inverse association between isoflavone and lignan metabolites and current asthma and non-asthmatic wheeze remained significant after the addition of smoke exposure to the models, stratifying by age (Tables E11–12). Adjustment for total dietary fiber intake, a global measure of plant intake, led to minimal change (0–3%, data not shown) in the point estimates for the association between enterolactone and current asthma, and enterolactone and O-DMA and non-asthmatic wheeze. There were similar associations between enterolactone and O-DMA and current asthma and non-asthmatic wheeze when the outcomes were stratified by atopic status (data not shown). We examined the relationship between bacterial metabolites of dietary lignans and isoflavones and asthma and wheeze using a large, nationally representative sample. We hypothesized that these metabolites would be inversely associated with asthma and wheezing because of their effects on molecular pathways involved in allergic inflammation and oxidation, and limited human studies supporting an association6. We found that higher urinary levels of the lignan metabolite enterolactone was inversely associated with a physician diagnosis of current asthma, and both enterolactone and O-DMA were inversely associated with non-asthmatic wheezing within the last year. Our findings were similar among adults and children and robust to smoke exposure, fiber intake, and atopic status. These results suggest that interventions to increase levels of enterolactone and O-DMA may be beneficial in the prevention and treatment of asthma and non-asthmatic wheeze. Several epidemiologic and laboratory-based studies have demonstrated associations between bacterial flora and lower respiratory disease7. The presence of specific organisms and the diversity of microflora are thought to be relevant in this relationship5. Notably, bacteria of the genus Bacteroides which are positively correlated with lignan metabolism, are inversely associated with the development of allergy, while the presence of Clostridium, which is negatively correlated with lignan metabolism, is positively associated with the development of allergy8,9. These observations together with our findings suggest that the bacterial metabolites of lignans and isoflavones, in particular enterolactone, may have a causal role in disease prevention and control. Our study has several limitations including incomplete characterization of respiratory outcomes and cross-sectional studies such as this are only a starting point for identifying causal relationships between exposures and disease. Further research will be necessary to determine if our observations result from different dietary choices among those with and without lower respiratory disease, that we could not detect by adjusting for fiber. Although lignan and isoflavone metabolites have antioxidant, anti-inflammatory, and other biological activities which we hypothesize underlie their association with lower respiratory disease, these may be biomarkers of another causal factor rather than themselves being causal in disease amelioration. Nonetheless, treatment with other (non-metabolite) lignans such as macelignan and honokiol decreases Th2-mediated pulmonary inflammation in animal models of asthma, supporting our overall hypothesis4. In summary, we have identified significant concentration-dependent associations between bacterial metabolites of diet-derived lignan and isoflavone metabolites and asthma and wheezing in a large, nationally-representative population-based sample. To our knowledge, this is the first report of an association between biologically active bacterial metabolites and lower respiratory outcomes in a nationally-representative population. Future clinical and laboratory-based studies are needed to support a causal relationship between lignan and isoflavone metabolites and lower respiratory disease. The potential diagnostic or therapeutic role for these bacterial metabolites, in particular enterolactone, warrants further study.

Dietary Fatty Acid Modification for the Treatment of Aspirin-Exacerbated Respiratory Disease: A Prospective Pilot Trial

To the editor: Reduced access to healthcare and food in the United States is associated with poor health outcomes1-3 and facilitating access to providers, medications, and food has led to measured improvements in public health.4, 5 Food allergy is a common, chronic condition, affecting 4-8% of US children, and is increasing in prevalence for unclear reasons.6, 7 Whether patients with food allergy experience impaired access to healthcare and food is currently unknown. Minority populations share a significant burden of food allergy,7, 8 and the rate of increase in food allergy in Black children may be twice that in Caucasian children.7 We were interested in whether subjects with food allergy report reduced access to healthcare and food and how this is associated with race/ethnicity, as this may influence disease outcome. We examined data from the 2011 and 2012 National Health Interview Survey (NHIS), a household interview survey of the US population covering a range of health topics. In each household, an adult answered questions about a randomly chosen child in the household. We considered a child to have food allergy if the responding adult answered yes to the question “during the past 12 months, has the sample child had any kind of food or digestive allergy?” We used the adults responses to questions regarding the childs access to healthcare and the familys access to food as measures of access to healthcare and food. Access to food was defined using the USDAs definition of “food security,” a measure of consistency of access to enough food for an active, healthy life. Please see the Online Repository at www.jacionline.org for additional information regarding the access measures, NHIS methodology and variable definitions. Statistical analyses were performed with STATA 12.0 software (StataCorp, College Station, Tex.). We used the chi square test to determine whether subjects with and without food allergy differed by demographic and access factors. We used logistic regression to determine the association between race/ethnicity and access, and adjusted for gender, age, family income and education, in a nested fashion. We incorporated survey weighting, sampling units, and strata in the primary chi square analysis, but because subjects were not equally distributed among the strata, only survey weights were incorporated in the chi square analysis stratified by race and in the logistic regression models. Complete data were available for 26,021 children from the combined 2011-2012 dataset, of whom 1,351 (5.59%) reported food allergy. Of the food allergic children, 54.8% were White, 17.1% Black/African American, 17.7% Hispanic/Latino/Spanish, and 10.4% Other (Table E1). Food allergy in the sample child was more common in families with a higher education level and with a higher household income, in line with previously reported demographic trends.6, 9 The survey population was equally distributed between genders and among age groups. Among children with food allergy, 20.95% were determined to have low food security, 33.53% reported having problems paying family medical bills, 4.47% reported not being able to afford needed prescriptions, 4.14% reported not being able to afford needed specialist care, 2.76% reported not being able to afford needed follow-up care, 2.45% percent reported having trouble finding a doctor to see the child, and 4.11% reported having no family member with health insurance (Figure 1A and Table E1). With the exception of having family members without insurance, these values are all significantly higher than those for children without food allergy (p≤0.05), and similar trends were observed when stratifying by race/ethnicity (Table E2). Figure 1 A: Percentage of subjects with food allergy reporting impaired food security or reduced access to healthcare metric. B: Distribution of subjects with food allergy reporting impaired food security or reduced access to healthcare metric by race/ethnicity. ... Compared to White children with food allergy, after adjusting for age and gender, Black children with food allergy were significantly more likely to have low food security (OR 3.31, 95% CI 2.17-5.06), to have problems paying family medical bills (OR 2.28, 95% CI 1.55-3.35), and to be unable to afford needed prescriptions (OR 3.44, 95% CI 1.68-7.02; Table 1 and Figure 1B). Hispanic children with food allergy were more likely to have low food security (OR 2.44, 95% CI 1.61-3.70), to have problems paying family medical bills (OR 1.56, 95% CI 1.08-2.23), and were more likely to be unable to afford needed prescriptions (OR 2.38, 95% CI 1.13-5.03) and follow-up care (OR 3.74, 95% CI 1.70-8.24). Many of these associations were attenuated after further adjusting for income and parental education. However, even after incorporating these variables, Black respondents with food allergy were significantly more likely to have low food security (OR 2.15, 95% CI 1.30-3.53), to have problems paying family medical bills (1.68, 95% CI 1.09-2.59), and to have trouble affording prescriptions for the child (OR 2.40, 95% CI 1.14-5.05) and Hispanic respondents with food allergy were significantly more likely to have trouble affording follow-up care (OR 3.02, 95% CI 1.34-6.81; Table 1) compared to White respondents with food allergy. There were no significant race/ethnicity differences in ability to afford specialist care or difficulty finding a doctor to see the child. Black respondents with food allergy were more likely in all models to have any family member with health insurance. We next compared children with food allergy to those with other chronic medical conditions and found that children with food allergy have similar or greater difficulty with access to care and food as children with other chronic medical conditions, with similar racial/ethnic disparities as in the previous analysis (Table E3). Table 1 Racial/ethnic disparities in likelihood of poor food security and reduced healthcare access among children with food allergy In this large national survey, we examined access to healthcare and food among subjects with food allergy, a chronic disease rising in prevalence. We found that compared to subjects without food allergy, subjects with food allergy are significantly more likely to report difficulty with access to care and food. Furthermore, parents of non-White children with food allergy were significantly more likely to report difficulty affording medical care and medications and low food security compared to parents of White children with food allergy. Not surprisingly, many of these associations were attenuated when we included parental income and education in the analysis. However, we were surprised that even after adjusting for income and education, Black respondents with food allergy were significantly more likely to report low food security and trouble affording prescriptions, and Hispanic respondents with food allergy were significantly more likely to report trouble affording follow-up care compared to White respondents. While it may be unsurprising that families of children with food allergies report more trouble accessing healthcare than families of children without food allergy, we did find that families of children with food allergy report at least as much, if not more, trouble accessing healthcare as families of children with other chronic diseases (Table E3 and supplementary information). Our results suggest there may be a barrier to accessing healthcare and food in children with food allergy, particularly among non-White children. Poor access to healthcare and food may increase morbidity, especially among minority children, by imposing poor nutrition and delayed treatment for allergic reactions. Associations drawn from cross-sectional studies are only a first step in understanding the association between food allergy and access to care. Our study is limited by the use of parental report of food or digestive allergy within the last year. This may over- or under-estimate food allergy prevalence, and further validation studies are needed to perform population based studies of food allergy. However, parent-reported food allergy prevalence in our sample falls within the range of previously reported estimates, and has been used in many epidemiologic studies of food allergy.6, 7, 9 Our cross-sectional study is also limited by the possibility of reverse causation in that decreased access to healthcare and food may increase the likelihood of self-report of food allergy. However, we incorporated potentially important socio-economic confounders such as income and education into our analyses, making this effect less likely. We were also limited by our inability to incorporate the full sampling design into our analysis due to the distribution of subjects within strata. Our estimates are therefore not necessarily nationally representative. However, this study is notable as it is the first to examine access to care among patients with food allergy and includes over 1,000 subjects with parent-reported food allergy, nearly 50% of whom are non-White. In summary, we have demonstrated that subjects with food allergy report difficulty with access to medical care and food, and that there are significant disparities in access associated with race/ethnicity. We were surprised that many of these disparities persisted after adjusting for income and education, which may be explained by socio-cultural factors and needs further investigation. Given the rising burden of food allergy, particularly among children of Black/African-American ethnicity, our results may have significant public health implications. Further study is necessary to determine if impaired access to care in patients with food allergy is associated with increased morbidity, and if improvements in access can improve disease outcome, as has been shown for other allergic disease such as asthma.4