Asthma, rhinitis, other respiratory diseases. Hay fever and asthma in relation to markers of infection in the United States

Transcription

1 Hay fever and asthma in relation to markers of infection in the United States Paolo Maria Matricardi, MD, a Francesco Rosmini, DSc, b Valentina Panetta, DSc, a Luigina Ferrigno, BSc, b and Sergio Bonini, MD a Rome, Italy Background: The hygiene hypothesis proposes that declining exposure to infections is implicated in the rising trend of allergy and asthma. Objective: We sought to test this hypothesis by examining the relationship of hay fever, asthma, and atopic sensitization with markers of infection in a large general population sample of the United States. Methods: We analyzed the data of 33,994 US residents recorded in a public database of a nationally representative cross-sectional survey (Third National Health and Nutrition Examination Survey, ). The variables examined were sociodemographic information, lifetime diagnosis and age at first diagnosis of hay fever or asthma, current skin sensitization to 9 airborne allergens and peanut, and current serology for Toxoplasma gondii, herpes simplex viruses type 1 and 2, and hepatitis A, B, and C viruses. Results: Hay fever (adjusted odds ratio, 0.27; 95% CI, ; P <.001) and asthma (adjusted odds ratio, 0.45; 95% CI, ; P <.001) were less frequent in subjects seropositive for hepatitis A virus (HAV), T gondii, and herpes simplex virus 1 versus seronegative subjects after adjusting for age, sex, race, urban residence, census region, family size, income, and education. Skin sensitization to peanut and to all the airborne allergens examined, except for cockroach, was less frequent among HAV-seropositive versus HAV-seronegative subjects younger than 40 years of age. The prevalence of hay fever and asthma diagnosed at or before 18 years of age in HAV-seronegative subjects increased progressively from 2.7% (95% CI, 0.7%-4.7%) and 0.4% (95% CI, 0.1%-1.6%), respectively, in cohorts born before 1920 to 8.5% (95% CI, 7.3%-9.7%) and 5.8% (95% CI, 4.8%-6.8%), respectively, in cohorts born in the 1960s, whereas they remained constant at around 2% in all cohorts of HAV-seropositive subjects. Conclusion: In the United States serologic evidence of acquisition of certain infections, mainly food-borne and orofecal infection, is associated with a lower probability of having hay fever and asthma. Third National Health and Nutrition Examination Survey data support the hypothesis that hygiene is a major factor contributing to the increase in hay fever, asthma, and atopic sensitization in westernized countries. (J Allergy Clin Immunol 2002;110:381-7.) From a the Institute of Neurobiology and Molecular Medicine, Consiglio Nazionale delle Ricerche, and b the Laboratory of Epidemiology and Biostatistics, Istituto Superiore di Sanità, Rome. Received for publication March 20, 2002; revised May 8, 2002; accepted for publication May 21, Reprint requests (present address): Paolo M. Matricardi, MD, WHO, World Health Organization, Chronic Respiratory Diseases and Arthritis Unit, 20 Avenue Appia, CH-1211 Geneve 27, Switzerland Mosby, Inc. All rights reserved /2002 $ /81/ doi: /mai Key words: Asthma, hay fever, hepatitis A virus, hygiene, infection, epidemiology, National Health and Nutrition Examination Survey III Allergic are spreading among populations living a Western lifestyle. 1 This epidemic has been variously attributed to pollution, 2 changes of diet, 3 and allergen exposure. 4 Another theory assumes that hygiene, by reducing exposure to infections, facilitates atopic responses and their inflammatory consequences at mucosa and skin surfaces, namely allergic asthma, rhinitis, and atopic eczema (the hygiene hypothesis). 5-7 In Europe atopy was found to be less frequent in children raised in large 5,8,9 and poor families, 5,9 on farms, 10 or in communities living a traditional-type lifestyle 11 and in children attending daycare centers. 12 Moreover, allergic rhinitis and asthma were inversely associated to positive serology for hepatitis A virus (HAV) 13,14 and for Toxoplasma gondii, 15 suggesting that the level of exposure to orofecal and food-borne infections might influence the inception of respiratory allergic. The emergence of allergic asthma in unsanitary inner-city areas in the United States seems irreconcilable with the hygiene hypothesis. 16,17 However, longitudinal studies in Tucson, Ariz, provided the formal demonstration that early exposure to other children in the family or at a daycare center protects children from asthma and atopy, and this was attributed to an earlier and more frequent acquisition of infections. 18 Should this interpretation be correct, markers of exposure to a higher microbial burden, such as positive serology to HAV and T gondii, should be independently associated with less allergy and asthma in the United States, as they are in Europe. The Third National Health and Nutrition Examination Survey, (NHANES III), examined a large national sample of Americans who responded to questionnaires on allergic and respiratory and who underwent allergy skin testing and blood testing for markers of infections. 19 This public database provided a unique opportunity to investigate whether hay fever and asthma are indeed correlated with serology for HAV, T gondii, and other markers of infection in the US general population. METHODS Study design We examined a public registry of 33,994 US residents aged 1 to 90 years or older that includes clinical and laboratory data from NHANES III and analyzed the following variables: age, sex, race, 381

2 382 Matricardi et al J ALLERGY CLIN IMMUNOL SEPTEMBER 2002 Abbreviations used EIA: Enzyme immunoassay HAV: Hepatitis A virus HSV: Herpes simplex virus NHANES: National Health and Nutrition Examination Survey OR: Odds ratio urban-rural residence, census, family size, family income, education, diagnosis ever, age at first diagnosis, current status of hay fever or asthma, skin tests for 9 airborne allergens and peanut, and serology for T gondii, herpes simplex viruses type 1 (HSV-1) and 2 (HSV-2), and hepatitis A, B, and C viruses. Skin prick tests and laboratory tests were performed only in certain age ranges (see below). Therefore analyses were made on samples of different sizes, according to the data available. NHANES III NHANES III was conducted from 1988 through 1994 by the National Center for Health Statistics of the Centers for Disease Control and Prevention, Atlanta, Ga, 20 and approved by the National Center for Health Statistics Institutional Review Board. The survey was based on a stratified multistage clustered probability design to select a representative sample of the civilian, noninstitutionalized US population. Eighty-one geographic sites were included in the final sample. All survey participants completed extensive questionnaires. Questionnaires for participants younger than 17 years were completed by a knowledgeable proxy. A large array of clinical and laboratory tests were performed in those who provided informed consent for in vivo testing and blood sampling. The variables we examined were defined as follows in NHANES III. Sociodemographic information. Age was defined as age in years at the last birthday. Year of birth, not recorded in the database, was calculated from age and date at interview. The latter was not recorded in the database and was arbitrarily established at March 1990 for participants examined between October 1988 and October 1991 and at March 1993 for participants examined between September 1991 and October Self-reported race was classified as white, black, or other. Socioeconomic status was determined by family income, defined as the total combined family income (expressed in US dollars per year), and by education, defined as the number of school years of the reference adult in the family (available only in the age range from years). Family size represents the total number of related persons living in a household (single-dwelling unit). NHANES III classified residence as rural or urban according to the United States Department of Agriculture Rural/Urban continuum codes and as Northeast, Midwest, South, and West according to the census region. Clinical data. Reported lifetime diagnoses of asthma or hay fever, age at first diagnosis, and current disease were ascertained by answers to the following questions: Has a doctor ever told you that you had asthma? or Has a doctor ever told you that you had hay fever? ; Do you still have asthma? or Do you still have hay fever? ; and How old were you when you were first told you had asthma? or How old were you when you were first told you had hay fever? Allergy skin testing. Allergy skin prick tests were proposed to all examinees aged 6 to 19 years (n = 5833) and to a random half sample (n = 5014) of examinees aged 20 to 59 years. Ten licensed, commercially available, Food and Drug Association approved standardized allergens were used: Dermatophagoides farinae, cat, German cockroach, short ragweed, perennial rye, Bermuda grass, Russian thistle, white oak, Alternaria alternata, and peanut. Histamine 1% and the glycerinated diluent were used as positive and negative controls, respectively. The length and width of the wheal reactions were measured after 15 minutes. A positive reaction was defined if the mean wheal diameter was 3 mm or larger after subtraction of the mean wheal diameter of the negative control. Serologic assays. 20 Total serum antibodies to HAV were tested in 21,260 participants aged 6 years or older by means of a qualitative, solid-phase competitive enzyme immunoassay (EIA; HAVAB EIA, Abbott Laboratories, North Chicago, Ill). IgG antibodies to T gondii were tested in 17,658 participants aged 12 years or older by using a quantitative, solid-phase indirect EIA (HAVAB EIA, Abbott Laboratories). IgG antibodies to viral glycoproteins specific for HSV-1 and HSV-2 were tested in 13,098 and 13,094 participants, respectively, aged 12 years or older by using a paper immunodot enzyme linked assay. Total antibodies to hepatitis B core antigen were tested in 21,265 participants aged 6 years or older by using a qualitative, solid-phase competitive radioimmunoassay (CORAB, Abbott Laboratories). IgG antibodies to hepatitis C virus were tested in 21,241 participants aged 6 years or older by using a qualitative, direct solid-phase EIA (hepatitis C virus EIA 2.0, Abbott Laboratories). In each instance a positive result was assigned according to kit instructions. Other markers of infection examined in NHANES III were not included in this analysis because they were mainly the expression of response to immunization (eg, anti-tetanus antibodies) or because data were limited to a small fraction of participants (eg, Helicobacter pylori). Statistical methods Contingency tables and multivariate logistic regression models were used to evaluate the crude and adjusted odds ratios (ORs) for skin sensitization, asthma, and hay fever by means of serology for single or cumulated infections. A χ 2 test was used to compare proportions, and linear trends were evaluated by using weighted regression analyses. 21 We consider a P value of.05 or greater to be not statistically significant. SPSS 10.0 software packages were used for analyses. RESULTS The frequency of hay fever ever diagnosed increased progressively with age up to 60 years and declined thereafter. Asthma ever diagnosed peaked in the second and sixth decades of life. The prevalence of a positive skin prick test response to at least one allergen increased with age up to values close to 60% in the 20- to 29-year age range and declined slightly across older groups (Fig 1, A). The prevalences of subjects with antibodies against HSV-1, HAV, or T gondii increased progressively with age; the prevalences of subjects with antibodies against hepatitis B or C viruses or HSV-2 were very low in the first 2 decades and increased in the third and fourth decades of life (Fig 1, B). A lifetime diagnosis of hay fever was significantly less frequent in HAV-seropositive than in HAV-seronegative participants aged 6 through 59 years. A lifetime diagnosis of asthma was also less frequent in HAV-seropositive participants aged 6 through 59 years, but differences between HAV-seropositive and HAV-seronegative participants were less marked than for hay fever and were statistically significant in subjects aged 20 through 59 years (Table I). The observed associations persisted after adjusting for age, sex, race, urban-rural residence, census region, family size, family income, and education (Table I). To evaluate whether a positive HAV serology was just

3 J ALLERGY CLIN IMMUNOL VOLUME 110, NUMBER 3 Matricardi et al 383 A B FIG 1. Prevalence of atopic sensitization (left axis), lifetime diagnosis of hay fever and asthma (right axis; A), and lifetime diagnosis of seropositivity for several infections (B) by age classes in the United States (NHANES III, ). HBV, Hepatitis B virus; HCV, hepatitis C virus; TG, T gondii; SPT, skin prick test. a surrogate of poverty, we tested the same associations in the subgroup of participants (n = 2711) aged 6 through 59 years and living with an annual income of less than $10,000 US. In this subgroup HAV seropositivity was inversely associated with hay fever (OR, 0.47; 95% CI, ; P <.0001) or asthma (OR, 0.66; 95% CI, ; P <.02) after adjusting for age, race, and sex. The prevalence of hay fever and asthma diagnosed before the age of 18 years among HAV-seronegative participants increased progressively from 2.7% (95% CI, 0.7%-4.7%) and 0.4% (95% CI, 0.1%-1.6%), respectively, in cohorts born before 1920 up to 8.5% (95% CI, 7.3%-9.7%) and 5.8% (95% CI, 4.8%-6.8%), respectively, in cohorts born in the 1960s. By contrast, there was no relevant increase of prevalence in any of the cohorts of individuals with HAV antibodies at examination (Fig 2). In subjects aged 6 through 59 years, hay fever was inversely associated with positive serology for HAV (OR, 0.5; 95% CI, ), T gondii (OR, 0.7; 95% CI, ), and HSV-1 (OR, 0.6; 95% CI, ; Fig 3). Asthma was inversely associated with seropositivity for HAV (OR, 0.6; 95% CI, ) and HSV-1 (OR, 0.6; 95% CI, ). Neither hay fever nor asthma were inversely associated with seropositivity for HSV-2 or hepatitis B or C viruses (Fig 3). Individuals simultaneously seropositive for HAV, T gondii, and HSV-1 had an odds of having hay fever or asthma, respectively, 4 times (OR, 0.27; 95% CI, ; P <.001) and 2 times (OR, 0.45; 95% CI, ; P <.001) lower than that of individuals seronegative for all 3 markers of infection. Skin sensitization to 8 of 9 inhalant allergens and to peanut was less frequent among HAV-seropositive versus HAV-seronegative subjects, with the differences being significant in individuals younger than 40 years. By contrast, sensitization to cockroach was not lower among HAV-seropositive subjects at any age (Table II). The adjusted odds of being seropositive for HAV decreased steadily with the overall degree of sensitization, which was obtained by means of summation of the mean wheal diameters induced by the airborne allergens tested (P for trend <.0001, Fig 4). DISCUSSION This study shows that hay fever, asthma, and skin sensitization to several inhalant allergens and peanut are inversely related to serologic markers of infections (ie, HAV, T gondii, and HSV-1 antibodies) in a representative population sample of the United States. In fact, the adjusted odds of a lifetime diagnosis of hay fever or asthma in subjects seronegative for all 3 infections (HAV, T gondii, and HSV-1) were, respectively, about 4 and 2 times higher than in seropositive individuals. In addition, this study provides the first evidence that during the last century, hay fever and asthma cases markedly increased in the United States but only among subsets with a lifestyle hygienic enough to prevent lifetime acquisition of HAV infection. Taken together, these data support the hypothesis that a higher exposure to certain infections protects from allergic rhinitis and asthma and that a reduced overall microbial burden contributes to the rising trend of allergic in westernized countries. 5,6 Our interpretation of the observed associations might be questioned because of the cross-sectional and retro-

4 384 Matricardi et al J ALLERGY CLIN IMMUNOL SEPTEMBER 2002 A B FIG 2. Prevalence of participants with a history of hay fever (A) or asthma (B) diagnosed before 18 years of age in relation to serology for HAV by decade of birth (NHANES III, ). Bars represent 95% CIs. P values reflect comparisons of prevalences of hay fever (A) or asthma (B) between participants seropositive versus seronegative for HAV at examination. spective nature of the study design. However, the NHANES III data confirm, in a larger, multiethnic population of a country characterized by diverse lifestyles, dietary habits, and climates, the inverse association of a positive serology for HAV with respiratory allergies identified among army recruits 13,15 and among adults 14 in Italy. The associations persisted after adjustment for confounders, including 2 reliable indicators of socioeconomic status, and therefore it is unlikely that positive serology for HAV is simply a nonspecific surrogate of low social economic status. 7 Moreover, the associations were confirmed by the objective tests of allergic sensitization, which also demonstrated a gradient between the overall degree of atopic sensitization and the odds of being seropositive for HAV. Consequently, our results were not biased by the type of access to medical care or by a different use of diagnostic labels. Finally, positive serology to T gondii and to HSV-1 were also associated with less hay fever and asthma, and the associations were more pronounced in cases of multiple markers of infections or a high degree of sensitization, which supports the quantitative nature of the underlying phenomenon. Thus on the basis of the foregoing considerations, the associations observed are consistent, dose dependent, and strong. This study does not provide information about age at acquisition of the infections examined. However, the

5 J ALLERGY CLIN IMMUNOL VOLUME 110, NUMBER 3 Matricardi et al 385 FIG 3. Adjusted odds of hay fever or asthma among NHANES III individuals younger than 60 years who were seropositive verus seronegative for various infections. ORs were adjusted for age, sex, race, urban-rural residence, geographic area, family size, family income, and education. Bars represent 95% CIs. The number of subjects included in the analysis and the percentage of those with positive serology is shown in parentheses. Results are shown on a log scale. HBV, Hepatitis B virus; HCV, hepatitis C virus; TG, T gondii. FIG 4. Gradient of association between positive serology for infections and overall degree of atopic sensitization. Overall, atopic sensitization was estimated as cumulative wheal diameter induced by the examined allergen. Adjusted ORs of seropositivity for HAV were calculated in a logistic regression analysis that included all the potential confounding variables (age, sex, race, urban-rural residence, geographic area, family size, and family income). The analysis is restricted to subjects with at least one wheal reaction of 3 mm or larger; the lowest degree of sensitization (cumulative wheal diameter <10 mm) is the reference category. Bars represent 95% CIs, and the results are shown on a log scale. The degrees of cumulative wheal diameter were also considered a single ordinal variable (P for trend <.0001). associations were stronger among younger subjects. Moreover, the infections inversely associated with respiratory allergies represent those that can be acquired early in life from the community (ie, HSV-1, HAV, and T gondii). By contrast, there was no inverse association with markers of infections typically acquired after puberty (eg, HSV-2 and hepatitis B and C viruses). Thus the NHANES III data are consistent with the concept that the earlier certain infections are acquired, the stronger their putative protecting effects. 5,6 Accordingly, a recent prospective study demonstrated that exposure to older children in the family or at a daycare center, which are indirect markers of more frequent acquisition of infections early in life, protected school-age children from asthma and atopy. 18 Proposals for mechanisms by which infections might protect against allergies evoke either the T H 1/T H 2 paradigm (stimulation of T H 1 immunity by certain infections would suppress T H 2 responses against allergens) 6 or the so-called anti-inflammatory network (certain infections would trigger and ensure strong immunoregulation that would prevent or suppress allergic and other immunemediated inflammations). 22 Our study, however, does not have the potential to investigate the biologic processes underlying the associations observed, although it provides novel clues as to these mechanisms. First, the inverse relationship between serum antibodies to infectious agents and respiratory disease was stronger for hay fever than for asthma, suggesting that infections affect allergic sensitization more than nonspecific bronchial reactivity. Second, the associations involved 2 foodborne and orofecal infections (HAV and T gondii) and one food allergen (ie, peanut), which is an emerging trigger of fatal allergic reactions in the United States. 23 This suggests that mechanisms underlying the associations reported here are not limited to the respiratory tract and to airborne allergens. Thus protection from hay fever and asthma by using a high exposure to orofecal and foodborne infections is likely an outcome at the level of the respiratory mucosae of a more complex and systemic process that encompasses structural components of other mucosa surfaces, including those of the gastrointestinal tract, 6,24 and their interactions with the innate and adaptive immune system. 25,26 This conjecture is consistent with the recent finding that ingestion of unpasteurized milk and exposure to stables account for most of the protection against the development of asthma and atopy linked to the farming environment. 27 However, although mechanisms of a direct protective role of HAV from the development of allergy and asthma have been recently proposed, 28 the nature of the food-borne and orofecal infections underlying the observed associations (viruses, bacteria, or parasites) remains uncertain. 22,24 Finally, sensitization to cockroach slightly prevailed among HAV-seropositive participants, possibly because of a higher exposure both to this allergen and to HAV of people with a lower socioeconomic status. 29,30 Nevertheless, the inverse association of HAV seropositivity with hay fever or asthma persisted among the poorest subjects, even after adjusting for race, suggesting that the hygiene

6 386 Matricardi et al J ALLERGY CLIN IMMUNOL SEPTEMBER 2002 TABLE I. Hay fever and asthma in relation to serology for HAV by age (NHANES III, United States) Antibodies to HAV Age 6-19 y y Positive Negative Adjusted OR Positive Negative Adjusted OR (%) (%) (95% CI)* (%) (%) (95% CI) Target population Hay fever ever diagnosed ( ) ( )# At interview ( ) ( )# Asthma ever diagnosed ( ) ( )# At interview ( )* ( ) *Adjusted for age, sex, race, urban-rural residence, census, family size, and family income. Adjusted for education also. Not all subjects had a complete data set. P <.05. P <.01. P <.001. #P < TABLE II. Skin sensitization to common airborne allergens and to peanut in relation to serology for HAV by age (NHANES III, United States) Antibodies to HAV Age 6-19 y y y Positive Negative Positive Negative Positive Negative (%) (%) Ratio (%) (%) Ratio (%) (%) Ratio No. examined Pollens Ragweed Bermuda grass Rye grass * White oak Russian thistle Molds Alternaria alternata Indoor allergens Cat Mite German cockroach Food allergen Peanut *P <.05, P <.01, and P <.001 for the comparison of HAV-seropositive and HAV-seronegative groups. hypothesis can be defended among inner-city areas, where allergic asthma emerged as a public health problem only after the 1970s. 17 Indeed, it is conceivable that in the United States a decline in the circulation of foodborne and orofecal infections and the consequent increase in susceptibility to development of allergy and asthma started much earlier among wealthier classes and only in recent decades spread among the less advantaged strata of the population, making them susceptible to risk factors of allergic asthma, such as high exposure to cockroaches, which is typical at any time only of a lower socioeconomic status. In conclusion, NHANES III data support the hypothesis that a declining exposure to infections, mainly foodborne and orofecal infections, is a major factor contributing to the increase in hay fever, asthma, and atopic sensitization in westernized countries. We acknowledge the US Department of Health and Human Services, National Center for Health Statistics, as the original source of data, and we thank Jean Ann Gilder (Scientific Communication) for her help in editing the text. REFERENCES 1. Von Mutius E, Martinez FD, Fritzsch C, Nicolai T, Roell G, Thiemann

7 J ALLERGY CLIN IMMUNOL VOLUME 110, NUMBER 3 Matricardi et al 387 Antibodies to HAV y y >80 y Positive Negative Adjusted OR Positive Negative Adjusted OR Positive Negative Adjusted OR (%) (%) (95% CI) (%) (%) (95% CI) (%) (%) (95% CI) ( ) ( ) ( ) ( )# ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) HH. Prevalence of asthma and atopy in two areas of West and East Germany. Am J Respir Crit Care Med 1994;149: Behrendt H, Friedrichs KH, Kramer U, Hitzfeld B, Becker WM, Ring J. The role of indoor and outdoor pollution in allergic. Prog Allergy Clin Immunol 1995;3: Weiss ST. Diet as a risk factor for asthma. Ciba Found Symp 1997; 206: Woolcock AJ, Peat JK, Trevillion LM. Is the increase in asthma prevalence linked to increase in allergen load? Allergy 1995;50: Strachan DP. Hay fever, hygiene and household size. BMJ 1989;299: Martinez FD, Holt PG. Role of microbial burden in aetiology of allergy and asthma. Lancet 1999;354(suppl II): Strachan DP. Family size, infection and atopy: the first decade of the hygiene hypothesis. Thorax 2000;55:S Von Mutius E, Martinez FD, Fritzsch C, Nicolai T, Reitmer P, Thiemann HH. Skin test reactivity and number of siblings. BMJ 1994;308: Matricardi PM, Franzinelli F, Franco A, Caprio G, Murru F, Cioffi D, et al. Sibship size, birth order, and atopy in 11,371 Italian young men. J Allergy Clin Immunol 1998;101: Braun-Fahrlander C, Gassner M, Grize L, Neu U, Sennhauser FH, Varonier HS, et al. Prevalence of hay fever and allergic sensitization in farmer s children and their peers living in the same rural community. SCARPOL team. Clin Exp Allergy 1999;29: Alm JS, Swartz J, Lilja G, Scheynius A, Pershagen G. Atopy in children with an anthroposophic lifestyle. Lancet 1999;353: Kramer U, Heinrich J, Wjst M, Wichmann HE. Age of entry to day nursery and allergy in later childhood. Lancet 1999;353: Matricardi PM, Rosmini F, Ferrigno L, Nisini R, Rapicetta M, Chionne P, et al. Cross-sectional retrospective study of prevalence of atopy among Italian military students with antibodies against hepatitis A virus. BMJ 1997;314: Matricardi PM, Rosmini F, Rapicetta M, Gasbarrini G, Stroffolini T. Atopy, hygiene and anthroposophic lifestyle. Lancet 1999;354: Matricardi PM, Rosmini F, Riondino S, Fortini M, Ferrigno L, Rapicetta M, et al. Exposure to foodborne and orofecal microbes versus airborne viruses in relation to atopy and allergic asthma: epidemiological study. BMJ 2000;320: Busse WW, Lemanske RF Jr. Asthma. N Engl J Med 2001;344: Platts-Mills TAE, Woodfolk JA, Sporik RA. The increase in asthma cannot be ascribed to cleanliness. Am J Respir Crit Care Med 2001;164: Ball TM, Castro-Rodriguez JA, Griffith KA, Holberg CJ, Martinez FD, Wright AL. Siblings, day-care attendance, and the risk of asthma and wheezing during childhood. N Engl J Med 2000;343: Plan and Operation of the Third National Health and Nutrition Examination Survey, , series 1: programs and collection procedures. Vital Health Stat ;32: Gunter EW, Lewis BG, Koncikowski SM. Laboratory procedures used for the Third National Health and Nutrition Examination Survey (NHANES III), In: NHANES III Reference Manuals and Reports. Atlanta (GA): NCHS-CDC; Publication no. PB INC. 21. Snedecor GW, Cochran WG. Statistical methods. 6th ed. Ames (IA): Iowa State University Press; p Yazdanbakhsh M, Kremsner PG, van Ree R. Allergy, parasites and the hygiene hypothesis. Science 2002;296: Bock SA, Munoz-Furlong A, Sampson HA. Fatalities due to anaphylactic reactions to foods. J Allergy Clin Immunol 2001;107: Matricardi PM, Rosmini F, Riondino S, Fortini M, Ferrigno L, Rapicetta M, et al. Exposure to foodborne and orofecal microbes versus airborne viruses in relation to atopy and allergic asthma: epidemiological study. BMJ 2000;320: Holgate ST, Davies DE, Lackie PM, Wilson SJ, Puddicombe SM, Lordan JL. Epithelial-mesenchymal interactions in the pathogenesis of asthma. J Allergy Clin Immunol 2000;105: Holgate ST. Science, medicine, and the future. Allergic disorders. BMJ 2000;320: Riedler J, Braun-Fahrländer C, Eder W, Schreuer M, Waser M, Maisch S, et al. Exposure to farming in early life and development of asthma and allergy: a cross-sectional survey. Lancet 2001;358: McIntire JJ, Umetsu SE, Akbari O, Potter M, Kuchroo VK, Barsh GS, et al. Identification of Tapr (an airway hyperreactivity regulatory locus) and the linked Tim gene family. Nat Immunol 2001;2: Kitch BT, Chew G, Burge HA, Muilenberg ML, Platts-Mills TA, O Connor G, et al. Socioeconomic predictors of high allergen levels in homes in the greater Boston area. Environ Health Perspect 2000;108: Stevenson LA, Gergen PJ, Hoover DR, Rosenstreich D, Mannino DM, Matte TD. Sociodemographic correlates of indoor allergen sensitivity among United States children. J Allergy Clin Immunol 2001; 108: