- Email: [email protected]
Do all asthmatics with atopy have atopic asthma?
1202 LETTERS TO THE EDITOR
or responders (complete or partial) in our series were sensitized to legumes (4/5 vs 4/10) and tomato (4/5 vs 3/10) compared with nonresponders. This observation coincides with a recent report from Spain7 and might reflect a geographic pattern of sensitization different from that seen in other countries. Unfortunately, the small sample size prevents us from drawing definitive conclusions. APT results were negative in most patients, a result that was similar to that seen in the other 3 adults patients from Spain recently reported as undergoing dietary therapy.7 Interestingly, in our series PPTs showed a higher rate of detecting food sensitization compared with SPTs and APTs, and therefore they deserve further consideration in patients with EoE. A relevant matter that might have influenced the response to a selective elimination diet is sensitization to aeroallergens. Adults with EoE are commonly sensitized to inhalant allergens that are believed to play a major causative role in adults with EoE. It is well known that EoE activity has seasonal variations, with documented exacerbations during pollen seasons and resolution during winter months.9 In our series responders 2, 3, and 4 were sensitized to plant foods holding homologies with pollen sensitivities, but the diet was carried out in autumn and winter in 2 of 5 responders and 5 of 10 nonresponders. Undoubtedly, this is an unresolved issue that should be considered for future research in dietary therapy for adult EoE. In conclusion, a selective elimination diet based on skin testing has suboptimal efficacy for adults with EoE. The usefulness of skin testing for detecting offending foods in adults with EoE is questionable, especially APTs. Our results do not support the development of dietary therapy based exclusively on skin test results for adults with EoE, in whom the SFED seems the better dietary approach until more accurate food testing or genetic profiling predicting the response to diet are available. Javier Molina-Infante, MDa Elisa Martin-Noguerol, MDa Manuela Alvarado-Arenas, MDb Sergio L. Porcel-Carre~ no, MD, PhDb Soledad Jimenez-Timon, MDb Francisco Javier Hernandez-Arbeiza, MDb From the Departments of aGastroenterology and bAllergy, Hospital San Pedro de Alcantara, Caceres, Spain. E-mail: [email protected]. Disclosure of potential conflict of interest: The authors declare that they have no relevant conflicts of interest. REFERENCES 1. Liacouras CA, Furuta GT, Hirano I, Atkins D, Attwood SE, Bonis PE, et al. Eosinophilic esophagitis: updated consensus recommendations for children and adults. J Allergy Clin Immunol 2011;128:3-10. 2. Kagalwalla AF, Shah A, Li BU, Sentongo TA, Ritz S, Manuel-Rubio M, et al. Identification of specific foods responsible for inflammation in children with eosinophilic esophagitis successfully treated with empiric elimination diet. J Pediatr Gastroenterol Nutr 2011;53:145-9. 3. Gonsalves N, Yang GY, Doerfler B, Ritz S, Ditto AM, Hirano I. Elimination diet effectively treats eosinophilic esophagitis in adults: food reintroduction identifies causative factors. Gastroenterology 2012;142:1451-9, e1. 4. Spergel JM, Andrews T, Brown-Whitehorn TF, Beausoleil JL, Liacouras CA. Treatment of eosinophilic esophagitis with specific elimination diet directed by a combination of skin prick and patch test. Ann Allergy Asthma Immunol 2005;95:336-43. 5. Spergel JM, Brown-Whitehorn T, Beausoleil JL, Shuker M, Liacouras CA. Predictive values for skin prick test and atopy patch test for eosinophilic esophagitis. J Allergy Clin Immunol 2007;119:509-11. 6. Simon D, Straumann A, Wenk A, Spichtin H, Simon H-U, Braathen AR. Eosinophilic esophagitis in adults: no clinical relevance of wheat and rye sensitizations. Allergy 2006;61:1480-3.
J ALLERGY CLIN IMMUNOL NOVEMBER 2012
7. Gonzalez-Cervera J, Angueira T, Rodriguez-Domınguez B, Arias A, Yag€ ue-Compadre JL, Lucendo AJ. Successful food elimination therapy in adult eosinophilic esophagitis: not all patients are the same. J Clin Gastroenterol 2012 [Epub ahead of print]. 8. Henderson CJ, Abonia JP, King EC, Putnam PE, Collins MH, Franciosi JP, et al. Comparative dietary therapy effectiveness in remission of pediatric eosinophilic esophagitis. J Allergy Clin Immunol 2012;129:1570-8. 9. Fogg MI, Ruchelli E, Spergel JM. Pollen and eosinophilic esophagitis. J Allergy Clin Immunol 2003;112:796-7. Available online August 3, 2012. http://dx.doi.org/10.1016/j.jaci.2012.06.027
Do all asthmatics with atopy have atopic asthma? To the Editor: Distinguishing between atopic (extrinsic or allergic) asthma and nonatopic (intrinsic or nonallergic) asthma is important for clinicians in making appropriate treatment decisions, such as whether to recommend anti-IgE therapy. It is also important in epidemiological and clinical research for defining an accurate asthma phenotype. In their asthmatic patients, clinicians often confirm atopy by allergy skin testing or allergen-specific IgE testing. An important question with both clinical and research relevance is ‘‘Do all asthmatics with atopy have atopic asthma?’’ I believe that most clinicians and researchers would agree that the presence of atopy among asthmatic patients is not always indicative of atopic asthma. However, I suspect that most would be surprised at the high probability of atopy and asthma being present in an individual without being related. For almost any disease, people who have a risk factor for a disease can develop the disease through a mechanism that does not include that risk factor. Greenland and Rothman,1 in the book Modern Epidemiology, illustrated this point with the following example: [A] smoker may develop lung cancer through some mechanism that does not involve smoking (e.g., one involving asbestos or radiation exposure). For such lung cancer cases, their smoking was incidental; it did not contribute to the cancer causation. .Therefore, exposed cases include some cases of disease caused by the exposure, if the exposure is indeed a cause, and some cases of disease that occur through mechanisms that do not involve the exposure. The attributable risk proportion, a fundamental measure in epidemiology that estimates the proportion of disease among the exposed that is due to that exposure, can reveal how much asthma among atopic patients is due to atopy. In the following analysis of data from the 2005-2006 National Health and Nutrition Examination Survey (NHANES), the results provide evidence that a large proportion of asthma cases among atopics are not atopic asthma cases. A detailed description of the NHANES 2005-2006 survey design and methods can be found online at http://www.cdc.gov/ nchs/nhanes.htm. For this analysis, current asthma was defined as a positive response to both of the following questions: ‘‘Has a doctor or other health professional ever told you that you have asthma?’’ and ‘‘Do you still have asthma?’’ The NHANES tested participants 6 years of age or older for allergen-specific IgE antibodies to 15 aeroallergens and 4 foods. Atopy was defined as at _0.35 kU/L). A sensitivity least 1 positive test result (specific IgE > analysis was conducted with 3 other definitions of atopy (Table I).
LETTERS TO THE EDITOR 1203
J ALLERGY CLIN IMMUNOL VOLUME 130, NUMBER 5
TABLE I. Proportion of asthma cases among atopics attributable to atopy in the US population aged 6 years and older Prevalence of current asthma (%, SE)y Varying definitions of atopy*
> _1 > _1 > _2 > _3
Positive Positive Positive Positive
test of 19 allergens test of 15 aeroallergens tests of 19 allergens tests of 19 allergens
Atopics A
13.0 13.4 15.5 18.7
(0.98) (0.97) (1.28) (1.59)
Nonatopics B
5.6 5.6 5.6 5.7
(0.41) (0.41) (0.39) (0.39)
Attributable to atopy A–B
7.4 7.8 9.9 12.9
(0.99) (0.96) (1.33) (1.59)
Proportion of asthma cases among atopics attributable to atopy (A – B)/A
0.57 0.58 0.64 0.69
*Allergen-specific IgE tests of serum were conducted for 15 aeroallergens and 4 foods. A positive test result was defined as 0.35 kU/L or more. _55), and race-ethnicity (Mexican American, non-Hispanic white, non-Hispanic black, and all others). Adjusted for sex, age in years (6-17, 18-40, 41-54, and >
FIG 1. Prevalence of current asthma in the US population aged 6 years and older, adjusted for sex, age in _55), and race-ethnicity (Mexican American, non-Hispanic white, non-Hispanic years (6-17, 18-40, 41-54, and > black, and all others).
This analysis was limited to the 7245 persons aged 6 years and older who had complete information on the 19 allergens and asthma. The prevalence of current asthma, adjusted for sex, age, and race-ethnicity, was estimated with the PREDMARG statement in the SUDAAN logistic regression procedure (release 10.0.1, SAS-Callable; RTI International, Research Triangle Park, NC). The attributable risk and attributable risk proportion were calculated.2 The sample weight variable WTMEC2YR was used in all analyses to obtain nationally representative statistics. The survey design variables SDMVSTRA and SDMVPSU were used to adjust the variances. The adjusted prevalence of current asthma among persons aged 6 years and older in the US population was 5.6% among nonatopics and 13.0% among atopics (Fig 1). The elevated prevalence of asthma among the nonatopics is the prevalence of asthma due to background risk factors, such as viral infections, tobacco smoke exposure, obesity, and psychological stress, to name a few. The risk of asthma that these background factors confer among atopics cannot be known with certainty because not all the background factors are known, but for the calculation of attributable risk proportion, it is assumed that everyone in the population shares the background risk,2 even the atopics. Thus, the total prevalence of asthma among the atopics is the sum of the background prevalence, shared by everyone, and the prevalence attributable to atopy alone (Fig 1). The prevalence of asthma attributable to atopy alone is the total prevalence of asthma among atopics (13.0%) minus the background prevalence (5.6%)—a difference of 7.4% (Fig 1). The proportion of the total prevalence of asthma among atopics that was attributable to atopy was 0.57 (7.4% divided by 13.0%). In other words, only 57% of the asthma
cases among atopics could be considered atopic asthma cases. For the remaining 43% of the asthma cases among atopics, their atopy was incidental. The attributable risk proportions by age group were 0.59 for 6 to 17 years, 0.63 for 18 to 40 years, 0.57 for 41 to 54 years, and 0.47 for 55 years or older. Table I shows the attributable risk proportions for atopy defined as a positive test to at least 1, 2, or 3 of the 19 allergens and as a positive test to at least 1 of the 15 aeroallergens. Even with the more stringent definitions, the results suggest that at least 30% of asthmatic patients with atopy do not have atopic asthma. The clinical implication of these findings is that a proportion of asthma patients with atopy may not have an allergic etiology to their asthma and may not, therefore, respond to allergy-related treatment strategies, such as anti-IgE therapy, subcutaneous immunotherapy, or allergen avoidance. To improve the probability of a correct diagnosis of atopic asthma, and thus effective treatment, clinicians should always consider allergy skin or IgE test results in the context of a thorough patient history that provides evidence of an association between allergen exposures and asthma symptoms. Limitations of this study include the cross-sectional nature of the survey and the potential misclassification of participants by atopy and asthma. Although the information on current asthma and atopy used in this analysis was collected cross-sectionally, the results should be relevant to clinical practice because, in most circumstances, a clinician assesses atopy and asthma at or near the same point in time, that is, cross-sectionally. Because not all known allergens were included in the panel of 19 allergens, and because the test results could have been negative at the time of testing due to the absence of allergen exposure, some nonatopics
1204 LETTERS TO THE EDITOR
J ALLERGY CLIN IMMUNOL NOVEMBER 2012
may be misclassified. Such misclassification could affect the background rate of asthma and ultimately the attributable risk proportion. However, because most atopics were sensitive to multiple allergens (mean 5 4.3), the probability of an atopic participant being positive to any untested allergen and not to at least 1 of the tested allergens is likely low, as is the probability of a transiently negative IgE panel. Finally, it would seem unlikely that the use of subject-reported, doctor-diagnosed asthma as opposed to a clinical assessment or medical chart review would change the main finding of this study, which is that a large proportion of asthma patients with atopy do not have atopic asthma. Samuel J. Arbes, Jr, DDS, MPH, PhD From Rho, Inc, Chapel Hill, NC. E-mail: [email protected]. Financial support was provided by Rho, Inc. Disclosure of potential conflict of interest: The author declares that he has no relevant conflicts of interest. REFERENCES 1. Greenland S, Rothman KJ. Measures of effect and measures of association. In: Rothman KJ, Greenland S, editors. Modern epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven; 1998: xiii, p. 737. 2. Gordis L. Epidemiology. Philadelphia, PA: W.B. Saunders; 1996. Available online August 22, 2012. http://dx.doi.org/10.1016/j.jaci.2012.06.040
Skin barrier abnormality caused by filaggrin (FLG) mutations is associated with increased serum 25-hydroxyvitamin D concentrations To the Editor: Epidermal exposure to solar UVB radiation results in synthesis of vitamin D3, which is converted to 25-hydroxy (25-OH) vitamin D in the liver.1 In addition to skin pigmentation, epidermal protection against UVB radiation is provided by using trans-urocanic acid (UCA),2 which is generated when histidase metabolizes histidine in filaggrin.3 Loss-of-function mutations in the filaggrin gene (FLG) affect 10% of Europeans and their descendants4 and reduce filaggrin protein levels in the skin, as well as its metabolites, including UCA.5 Because experimental knockdown of filaggrin markedly increases the UV sensitivity of keratinocytes,6 we hypothesized that the concentration of serum 25-OH vitamin D would be increased in carriers of common FLG mutations. We included 5 general population cohorts from Denmark and Germany: 2 children’s cohorts, Copenhagen Prospective Studies on Asthma in Childhood (COPSAC) and German Infant Study on the Influence of Nutrition Intervention (GINI)/Influence of Lifestyle Factors on the Development of the Immune System and Allergies in East and Wes Germany (LISA), and 3 cross-sectional adult cohorts, Health2006, Monitoring of Trends and Determinants in Cardiovascular Disease Healthy Survey (Monica), and Cooperative Health Research in the Region of Augsburg (KORA) F4. Atopic dermatitis (AD), dietary vitamin D intake, and vitamin D status were assessed as described in the Methods section and Table E1 in this article’s Online Repository at www.jacionline.org. Subjects were dichotomized into those not carrying an FLG mutation or carriers of at least 1 of the R501X or 2282del4 FLG mutations. Serum 25-OH vitamin D concentrations were log transformed to approximate a normal distribution. One-way ANOVA was applied to detect differences in log-transformed values of serum measurements of 25-OH vitamin D between
study groups (Table I). Multiple linear regression models were used to adjust for potential confounding and known risk factors, including sex, age, body mass index, dietary vitamin supplements, AD, and season of vitamin D testing.1 Results were computed as percentage differences compared with the reference group corresponding to the back-transformed b-coefficients from the linear regression analyses on log-transformed outcomes multiplied by 100. This was done to compensate for the heterogeneous 25-OH vitamin D measurements in the 5 cohorts. Betweenstudy heterogeneity was tested by using the x2-based Q-statistic and quantified by using the I2 value as a measure of the proportion of variance between the study-specific estimates that are attributable to between-study difference rather than random variation (I2 describes the percentage of total variation across studies that is due to heterogeneity rather than chance). Because there was no evidence for between-study heterogeneity, the study-specific estimates were combined according to the inverse-variance weighted fixed-effects model. The forest plot was prepared by using Review Manager Computer program Version 5.1 (2011; the Nordic Cochrane Centre, the Cochrane Collaboration, Copenhagen, Denmark). All P values were 2-tailed, and statistical significance was defined as a P value of less than .05, except for the heterogeneity test, which was considered to be significant at the 10% level. SAS version 9.2 software (SAS Institute, Cary, NC) was used for the COPSAC, Health2006, Monica and KORA F4 study data analyses, and R version 2.13.1 software (http://www.R-project.org) was used for the GINI/LISA study. The concentrations of 25-OH vitamin D in cord blood were independent of FLG mutation status (P 5 .65), indicating no difference at birth. A joint analysis including 9950 subjects aged 4 to 81 years using the raw model showed a strong effect of FLG mutations, with 10.7% (95% CI, 7.0% to 14.6%; P < .0001) higher 25-OH vitamin D concentrations in FLG mutation carriers. An adjusted joint analysis showed a 10.1% (95% CI, 6.7% to 13.6%; P <.0001) difference, suggesting that the association was not confounded (Fig 1). An analysis excluding subjects with AD (n 5 8849) produced a similar result of 9.1% (95% CI, 5.2% to 13.1%; P < .0001). The I2 measure indicated no evidence for significant between-study heterogeneity (I2 5 0%; 95% CI, 0% to 79.2%; Q-test P 5 .51). No interaction was detected between the FLG genotype and age (P 5 .22) or sex (P 5 .56). The FLG effect was stronger in the summer half year (interaction P 5 .12; April-September: 13.1% [95% CI, 8.3% to 18.1%], Q-test P 5 .6, n 5 4398) than in the winter half year (October-March: 7.2% [95% CI, 2.5% to 12.2%], Q-test P 5 .7, n 5 5552). Only 20 subjects were compound heterozygous/homozygous FLG mutation carriers in the 5 cohorts. No overall dose-response effect of FLG mutation status on 25-OH vitamin D concentrations was detected. The 9 compound heterozygous/homozygous carriers from the Health2006 study had 29.6% (95% CI, 29.0% to 84.6%) higher 25-OH vitamin D concentrations compared with 9.6% (95% CI, 2.3% to 17.4%) in heterozygous carriers. Although rather different measures were used, we showed significantly higher concentrations of serum 25-OH vitamin D in carriers of FLG mutations with homogeneity across the studies. Although not significant, the effect of FLG mutations increased over the summer half year, supporting the hypothesis that the difference in vitamin D status was caused by UVexposure. Filaggrin is hydrolyzed in a humidity-sensitive fashion to its constituent amino acids and their deiminated carboxylic acid derivates, including UCA.3 This process typically accelerates during the winter in temperate
or responders (complete or partial) in our series were sensitized to legumes (4/5 vs 4/10) and tomato (4/5 vs 3/10) compared with nonresponders. This observation coincides with a recent report from Spain7 and might reflect a geographic pattern of sensitization different from that seen in other countries. Unfortunately, the small sample size prevents us from drawing definitive conclusions. APT results were negative in most patients, a result that was similar to that seen in the other 3 adults patients from Spain recently reported as undergoing dietary therapy.7 Interestingly, in our series PPTs showed a higher rate of detecting food sensitization compared with SPTs and APTs, and therefore they deserve further consideration in patients with EoE. A relevant matter that might have influenced the response to a selective elimination diet is sensitization to aeroallergens. Adults with EoE are commonly sensitized to inhalant allergens that are believed to play a major causative role in adults with EoE. It is well known that EoE activity has seasonal variations, with documented exacerbations during pollen seasons and resolution during winter months.9 In our series responders 2, 3, and 4 were sensitized to plant foods holding homologies with pollen sensitivities, but the diet was carried out in autumn and winter in 2 of 5 responders and 5 of 10 nonresponders. Undoubtedly, this is an unresolved issue that should be considered for future research in dietary therapy for adult EoE. In conclusion, a selective elimination diet based on skin testing has suboptimal efficacy for adults with EoE. The usefulness of skin testing for detecting offending foods in adults with EoE is questionable, especially APTs. Our results do not support the development of dietary therapy based exclusively on skin test results for adults with EoE, in whom the SFED seems the better dietary approach until more accurate food testing or genetic profiling predicting the response to diet are available. Javier Molina-Infante, MDa Elisa Martin-Noguerol, MDa Manuela Alvarado-Arenas, MDb Sergio L. Porcel-Carre~ no, MD, PhDb Soledad Jimenez-Timon, MDb Francisco Javier Hernandez-Arbeiza, MDb From the Departments of aGastroenterology and bAllergy, Hospital San Pedro de Alcantara, Caceres, Spain. E-mail: [email protected]. Disclosure of potential conflict of interest: The authors declare that they have no relevant conflicts of interest. REFERENCES 1. Liacouras CA, Furuta GT, Hirano I, Atkins D, Attwood SE, Bonis PE, et al. Eosinophilic esophagitis: updated consensus recommendations for children and adults. J Allergy Clin Immunol 2011;128:3-10. 2. Kagalwalla AF, Shah A, Li BU, Sentongo TA, Ritz S, Manuel-Rubio M, et al. Identification of specific foods responsible for inflammation in children with eosinophilic esophagitis successfully treated with empiric elimination diet. J Pediatr Gastroenterol Nutr 2011;53:145-9. 3. Gonsalves N, Yang GY, Doerfler B, Ritz S, Ditto AM, Hirano I. Elimination diet effectively treats eosinophilic esophagitis in adults: food reintroduction identifies causative factors. Gastroenterology 2012;142:1451-9, e1. 4. Spergel JM, Andrews T, Brown-Whitehorn TF, Beausoleil JL, Liacouras CA. Treatment of eosinophilic esophagitis with specific elimination diet directed by a combination of skin prick and patch test. Ann Allergy Asthma Immunol 2005;95:336-43. 5. Spergel JM, Brown-Whitehorn T, Beausoleil JL, Shuker M, Liacouras CA. Predictive values for skin prick test and atopy patch test for eosinophilic esophagitis. J Allergy Clin Immunol 2007;119:509-11. 6. Simon D, Straumann A, Wenk A, Spichtin H, Simon H-U, Braathen AR. Eosinophilic esophagitis in adults: no clinical relevance of wheat and rye sensitizations. Allergy 2006;61:1480-3.
J ALLERGY CLIN IMMUNOL NOVEMBER 2012
7. Gonzalez-Cervera J, Angueira T, Rodriguez-Domınguez B, Arias A, Yag€ ue-Compadre JL, Lucendo AJ. Successful food elimination therapy in adult eosinophilic esophagitis: not all patients are the same. J Clin Gastroenterol 2012 [Epub ahead of print]. 8. Henderson CJ, Abonia JP, King EC, Putnam PE, Collins MH, Franciosi JP, et al. Comparative dietary therapy effectiveness in remission of pediatric eosinophilic esophagitis. J Allergy Clin Immunol 2012;129:1570-8. 9. Fogg MI, Ruchelli E, Spergel JM. Pollen and eosinophilic esophagitis. J Allergy Clin Immunol 2003;112:796-7. Available online August 3, 2012. http://dx.doi.org/10.1016/j.jaci.2012.06.027
Do all asthmatics with atopy have atopic asthma? To the Editor: Distinguishing between atopic (extrinsic or allergic) asthma and nonatopic (intrinsic or nonallergic) asthma is important for clinicians in making appropriate treatment decisions, such as whether to recommend anti-IgE therapy. It is also important in epidemiological and clinical research for defining an accurate asthma phenotype. In their asthmatic patients, clinicians often confirm atopy by allergy skin testing or allergen-specific IgE testing. An important question with both clinical and research relevance is ‘‘Do all asthmatics with atopy have atopic asthma?’’ I believe that most clinicians and researchers would agree that the presence of atopy among asthmatic patients is not always indicative of atopic asthma. However, I suspect that most would be surprised at the high probability of atopy and asthma being present in an individual without being related. For almost any disease, people who have a risk factor for a disease can develop the disease through a mechanism that does not include that risk factor. Greenland and Rothman,1 in the book Modern Epidemiology, illustrated this point with the following example: [A] smoker may develop lung cancer through some mechanism that does not involve smoking (e.g., one involving asbestos or radiation exposure). For such lung cancer cases, their smoking was incidental; it did not contribute to the cancer causation. .Therefore, exposed cases include some cases of disease caused by the exposure, if the exposure is indeed a cause, and some cases of disease that occur through mechanisms that do not involve the exposure. The attributable risk proportion, a fundamental measure in epidemiology that estimates the proportion of disease among the exposed that is due to that exposure, can reveal how much asthma among atopic patients is due to atopy. In the following analysis of data from the 2005-2006 National Health and Nutrition Examination Survey (NHANES), the results provide evidence that a large proportion of asthma cases among atopics are not atopic asthma cases. A detailed description of the NHANES 2005-2006 survey design and methods can be found online at http://www.cdc.gov/ nchs/nhanes.htm. For this analysis, current asthma was defined as a positive response to both of the following questions: ‘‘Has a doctor or other health professional ever told you that you have asthma?’’ and ‘‘Do you still have asthma?’’ The NHANES tested participants 6 years of age or older for allergen-specific IgE antibodies to 15 aeroallergens and 4 foods. Atopy was defined as at _0.35 kU/L). A sensitivity least 1 positive test result (specific IgE > analysis was conducted with 3 other definitions of atopy (Table I).
LETTERS TO THE EDITOR 1203
J ALLERGY CLIN IMMUNOL VOLUME 130, NUMBER 5
TABLE I. Proportion of asthma cases among atopics attributable to atopy in the US population aged 6 years and older Prevalence of current asthma (%, SE)y Varying definitions of atopy*
> _1 > _1 > _2 > _3
Positive Positive Positive Positive
test of 19 allergens test of 15 aeroallergens tests of 19 allergens tests of 19 allergens
Atopics A
13.0 13.4 15.5 18.7
(0.98) (0.97) (1.28) (1.59)
Nonatopics B
5.6 5.6 5.6 5.7
(0.41) (0.41) (0.39) (0.39)
Attributable to atopy A–B
7.4 7.8 9.9 12.9
(0.99) (0.96) (1.33) (1.59)
Proportion of asthma cases among atopics attributable to atopy (A – B)/A
0.57 0.58 0.64 0.69
*Allergen-specific IgE tests of serum were conducted for 15 aeroallergens and 4 foods. A positive test result was defined as 0.35 kU/L or more. _55), and race-ethnicity (Mexican American, non-Hispanic white, non-Hispanic black, and all others). Adjusted for sex, age in years (6-17, 18-40, 41-54, and >
FIG 1. Prevalence of current asthma in the US population aged 6 years and older, adjusted for sex, age in _55), and race-ethnicity (Mexican American, non-Hispanic white, non-Hispanic years (6-17, 18-40, 41-54, and > black, and all others).
This analysis was limited to the 7245 persons aged 6 years and older who had complete information on the 19 allergens and asthma. The prevalence of current asthma, adjusted for sex, age, and race-ethnicity, was estimated with the PREDMARG statement in the SUDAAN logistic regression procedure (release 10.0.1, SAS-Callable; RTI International, Research Triangle Park, NC). The attributable risk and attributable risk proportion were calculated.2 The sample weight variable WTMEC2YR was used in all analyses to obtain nationally representative statistics. The survey design variables SDMVSTRA and SDMVPSU were used to adjust the variances. The adjusted prevalence of current asthma among persons aged 6 years and older in the US population was 5.6% among nonatopics and 13.0% among atopics (Fig 1). The elevated prevalence of asthma among the nonatopics is the prevalence of asthma due to background risk factors, such as viral infections, tobacco smoke exposure, obesity, and psychological stress, to name a few. The risk of asthma that these background factors confer among atopics cannot be known with certainty because not all the background factors are known, but for the calculation of attributable risk proportion, it is assumed that everyone in the population shares the background risk,2 even the atopics. Thus, the total prevalence of asthma among the atopics is the sum of the background prevalence, shared by everyone, and the prevalence attributable to atopy alone (Fig 1). The prevalence of asthma attributable to atopy alone is the total prevalence of asthma among atopics (13.0%) minus the background prevalence (5.6%)—a difference of 7.4% (Fig 1). The proportion of the total prevalence of asthma among atopics that was attributable to atopy was 0.57 (7.4% divided by 13.0%). In other words, only 57% of the asthma
cases among atopics could be considered atopic asthma cases. For the remaining 43% of the asthma cases among atopics, their atopy was incidental. The attributable risk proportions by age group were 0.59 for 6 to 17 years, 0.63 for 18 to 40 years, 0.57 for 41 to 54 years, and 0.47 for 55 years or older. Table I shows the attributable risk proportions for atopy defined as a positive test to at least 1, 2, or 3 of the 19 allergens and as a positive test to at least 1 of the 15 aeroallergens. Even with the more stringent definitions, the results suggest that at least 30% of asthmatic patients with atopy do not have atopic asthma. The clinical implication of these findings is that a proportion of asthma patients with atopy may not have an allergic etiology to their asthma and may not, therefore, respond to allergy-related treatment strategies, such as anti-IgE therapy, subcutaneous immunotherapy, or allergen avoidance. To improve the probability of a correct diagnosis of atopic asthma, and thus effective treatment, clinicians should always consider allergy skin or IgE test results in the context of a thorough patient history that provides evidence of an association between allergen exposures and asthma symptoms. Limitations of this study include the cross-sectional nature of the survey and the potential misclassification of participants by atopy and asthma. Although the information on current asthma and atopy used in this analysis was collected cross-sectionally, the results should be relevant to clinical practice because, in most circumstances, a clinician assesses atopy and asthma at or near the same point in time, that is, cross-sectionally. Because not all known allergens were included in the panel of 19 allergens, and because the test results could have been negative at the time of testing due to the absence of allergen exposure, some nonatopics
1204 LETTERS TO THE EDITOR
J ALLERGY CLIN IMMUNOL NOVEMBER 2012
may be misclassified. Such misclassification could affect the background rate of asthma and ultimately the attributable risk proportion. However, because most atopics were sensitive to multiple allergens (mean 5 4.3), the probability of an atopic participant being positive to any untested allergen and not to at least 1 of the tested allergens is likely low, as is the probability of a transiently negative IgE panel. Finally, it would seem unlikely that the use of subject-reported, doctor-diagnosed asthma as opposed to a clinical assessment or medical chart review would change the main finding of this study, which is that a large proportion of asthma patients with atopy do not have atopic asthma. Samuel J. Arbes, Jr, DDS, MPH, PhD From Rho, Inc, Chapel Hill, NC. E-mail: [email protected]. Financial support was provided by Rho, Inc. Disclosure of potential conflict of interest: The author declares that he has no relevant conflicts of interest. REFERENCES 1. Greenland S, Rothman KJ. Measures of effect and measures of association. In: Rothman KJ, Greenland S, editors. Modern epidemiology. 2nd ed. Philadelphia, PA: Lippincott-Raven; 1998: xiii, p. 737. 2. Gordis L. Epidemiology. Philadelphia, PA: W.B. Saunders; 1996. Available online August 22, 2012. http://dx.doi.org/10.1016/j.jaci.2012.06.040
Skin barrier abnormality caused by filaggrin (FLG) mutations is associated with increased serum 25-hydroxyvitamin D concentrations To the Editor: Epidermal exposure to solar UVB radiation results in synthesis of vitamin D3, which is converted to 25-hydroxy (25-OH) vitamin D in the liver.1 In addition to skin pigmentation, epidermal protection against UVB radiation is provided by using trans-urocanic acid (UCA),2 which is generated when histidase metabolizes histidine in filaggrin.3 Loss-of-function mutations in the filaggrin gene (FLG) affect 10% of Europeans and their descendants4 and reduce filaggrin protein levels in the skin, as well as its metabolites, including UCA.5 Because experimental knockdown of filaggrin markedly increases the UV sensitivity of keratinocytes,6 we hypothesized that the concentration of serum 25-OH vitamin D would be increased in carriers of common FLG mutations. We included 5 general population cohorts from Denmark and Germany: 2 children’s cohorts, Copenhagen Prospective Studies on Asthma in Childhood (COPSAC) and German Infant Study on the Influence of Nutrition Intervention (GINI)/Influence of Lifestyle Factors on the Development of the Immune System and Allergies in East and Wes Germany (LISA), and 3 cross-sectional adult cohorts, Health2006, Monitoring of Trends and Determinants in Cardiovascular Disease Healthy Survey (Monica), and Cooperative Health Research in the Region of Augsburg (KORA) F4. Atopic dermatitis (AD), dietary vitamin D intake, and vitamin D status were assessed as described in the Methods section and Table E1 in this article’s Online Repository at www.jacionline.org. Subjects were dichotomized into those not carrying an FLG mutation or carriers of at least 1 of the R501X or 2282del4 FLG mutations. Serum 25-OH vitamin D concentrations were log transformed to approximate a normal distribution. One-way ANOVA was applied to detect differences in log-transformed values of serum measurements of 25-OH vitamin D between
study groups (Table I). Multiple linear regression models were used to adjust for potential confounding and known risk factors, including sex, age, body mass index, dietary vitamin supplements, AD, and season of vitamin D testing.1 Results were computed as percentage differences compared with the reference group corresponding to the back-transformed b-coefficients from the linear regression analyses on log-transformed outcomes multiplied by 100. This was done to compensate for the heterogeneous 25-OH vitamin D measurements in the 5 cohorts. Betweenstudy heterogeneity was tested by using the x2-based Q-statistic and quantified by using the I2 value as a measure of the proportion of variance between the study-specific estimates that are attributable to between-study difference rather than random variation (I2 describes the percentage of total variation across studies that is due to heterogeneity rather than chance). Because there was no evidence for between-study heterogeneity, the study-specific estimates were combined according to the inverse-variance weighted fixed-effects model. The forest plot was prepared by using Review Manager Computer program Version 5.1 (2011; the Nordic Cochrane Centre, the Cochrane Collaboration, Copenhagen, Denmark). All P values were 2-tailed, and statistical significance was defined as a P value of less than .05, except for the heterogeneity test, which was considered to be significant at the 10% level. SAS version 9.2 software (SAS Institute, Cary, NC) was used for the COPSAC, Health2006, Monica and KORA F4 study data analyses, and R version 2.13.1 software (http://www.R-project.org) was used for the GINI/LISA study. The concentrations of 25-OH vitamin D in cord blood were independent of FLG mutation status (P 5 .65), indicating no difference at birth. A joint analysis including 9950 subjects aged 4 to 81 years using the raw model showed a strong effect of FLG mutations, with 10.7% (95% CI, 7.0% to 14.6%; P < .0001) higher 25-OH vitamin D concentrations in FLG mutation carriers. An adjusted joint analysis showed a 10.1% (95% CI, 6.7% to 13.6%; P <.0001) difference, suggesting that the association was not confounded (Fig 1). An analysis excluding subjects with AD (n 5 8849) produced a similar result of 9.1% (95% CI, 5.2% to 13.1%; P < .0001). The I2 measure indicated no evidence for significant between-study heterogeneity (I2 5 0%; 95% CI, 0% to 79.2%; Q-test P 5 .51). No interaction was detected between the FLG genotype and age (P 5 .22) or sex (P 5 .56). The FLG effect was stronger in the summer half year (interaction P 5 .12; April-September: 13.1% [95% CI, 8.3% to 18.1%], Q-test P 5 .6, n 5 4398) than in the winter half year (October-March: 7.2% [95% CI, 2.5% to 12.2%], Q-test P 5 .7, n 5 5552). Only 20 subjects were compound heterozygous/homozygous FLG mutation carriers in the 5 cohorts. No overall dose-response effect of FLG mutation status on 25-OH vitamin D concentrations was detected. The 9 compound heterozygous/homozygous carriers from the Health2006 study had 29.6% (95% CI, 29.0% to 84.6%) higher 25-OH vitamin D concentrations compared with 9.6% (95% CI, 2.3% to 17.4%) in heterozygous carriers. Although rather different measures were used, we showed significantly higher concentrations of serum 25-OH vitamin D in carriers of FLG mutations with homogeneity across the studies. Although not significant, the effect of FLG mutations increased over the summer half year, supporting the hypothesis that the difference in vitamin D status was caused by UVexposure. Filaggrin is hydrolyzed in a humidity-sensitive fashion to its constituent amino acids and their deiminated carboxylic acid derivates, including UCA.3 This process typically accelerates during the winter in temperate