Asthma in farm children is more determined by genetic polymorphisms and in non-farm children by environmental factors.

BACKGROUND: The asthma syndrome is influenced by hereditary and environmental factors. With the example of farm exposure, we study whether genetic and environmental factors interact for asthma. METHODS: Statistical learning approaches based on penalized regression and decision trees were used to predict asthma in the GABRIELA study with 850 cases (9% farm children) and 857 controls (14% farm children). Single-nucleotide polymorphisms (SNPs) were selected from a genome-wide dataset based on a literature search or by statistical selection techniques. Prediction was assessed by receiver operating characteristics (ROC) curves and validated in the PASTURE cohort. RESULTS: Prediction by family history of asthma and atopy yielded an area under the ROC curve (AUC) of 0.62 [0.57-0.66] in the random forest machine learning approach. By adding information on demographics (sex and age) and 26 environmental exposure variables, the quality of prediction significantly improved (AUC = 0.65 [0.61-0.70]). In farm children, however, environmental variables did not improve prediction quality. Rather SNPs related to IL33 and RAD50 contributed significantly to the prediction of asthma (AUC = 0.70 [0.62-0.78]). CONCLUSIONS: Asthma in farm children is more likely predicted by other factors as compared to non-farm children though in both forms, family history may integrate environmental exposure, genotype and degree of penetrance.

Development of atopic sensitization in Finnish and Estonian children: A latent class analysis in a multicenter cohort.

BACKGROUND: The prevalence of atopy is associated with a Western lifestyle, as shown by studies comparing neighboring regions with different socioeconomic backgrounds. Atopy might reflect various conditions differing in their susceptibility to environmental factors. OBJECTIVE: We sought to define phenotypes of atopic sensitization in early childhood and examine their association with allergic diseases and hereditary background in Finland and Estonia. METHODS: The analysis included 1603 Finnish and 1657 Estonian children from the DIABIMMUNE multicenter young children cohort. Specific IgE levels were measured at age 3, 4, and 5 years, respectively, and categorized into 3 CAP classes. Latent class analysis was performed with the statistical software package poLCA in R software. RESULTS: Both populations differed in terms of socioeconomic status and environmental determinants, such as pet ownership, farm-related exposure, time spent playing outdoors, and prevalence of allergic diseases (all P < .001). Nevertheless, we found similar latent classes in both populations: an unsensitized class, a food class, 2 inhalant classes differentiating between seasonal and perennial aeroallergens, and a severe atopy class. The latter was characterized by high total and specific IgE levels and strongly associated with wheeze (odds ratio [OR], 5.64 [95% CI, 3.07-10.52] and 4.56 [95% CI, 2.35-8.52]), allergic rhinitis (OR, 22.4 [95% CI, 11.67-44.54] and 13.97 [95% CI, 7.33-26.4]), and atopic eczema (OR, 9.39 [95% CI, 4.9-19.3] and 9.5 [95% CI, 5.2-17.5] for Finland and Estonia, respectively). Environmental differences were reflected in the larger seasonal inhalant atopy class in Finland, although composition of classes was comparable between countries. CONCLUSION: Despite profound differences in environmental exposures, there might exist genuine patterns of atopic sensitization. The distribution of these patterns might determine the contribution of atopic sensitization to disease onset.

Genome-wide interaction study of early-life smoking exposure on time-to-asthma onset in childhood.

BACKGROUND: Asthma, a heterogeneous disease with variable age of onset, results from the interplay between genetic and environmental factors. Early-life tobacco smoke (ELTS) exposure is a major asthma risk factor. Only a few genetic loci have been reported to interact with ELTS exposure in asthma. OBJECTIVE: Our aim was to identify new loci interacting with ELTS exposure on time-to-asthma onset (TAO) in childhood. METHODS: We conducted genome-wide interaction analyses of ELTS exposure on time-to-asthma onset in childhood in five European-ancestry studies (totalling 8273 subjects) using Cox proportional-hazard model. The results of all five genome-wide analyses were meta-analysed. RESULTS: The 13q21 locus showed genome-wide significant interaction with ELTS exposure (P = 4.3 × 10-8 for rs7334050 within KLHL1 with consistent results across the five studies). Suggestive interactions (P < 5 × 10-6 ) were found at three other loci: 20p12 (rs13037508 within MACROD2; P = 4.9 × 10-7 ), 14q22 (rs7493885 near NIN; P = 2.9 × 10-6 ) and 2p22 (rs232542 near CYP1B1; P = 4.1 × 10-6 ). Functional annotations and the literature showed that the lead SNPs at these four loci influence DNA methylation in the blood and are located nearby CpG sites reported to be associated with exposure to tobacco smoke components, which strongly support our findings. CONCLUSIONS AND CLINICAL RELEVANCE: We identified novel candidate genes interacting with ELTS exposure on time-to-asthma onset in childhood. These genes have plausible biological relevance related to tobacco smoke exposure. Further epigenetic and functional studies are needed to confirm these findings and to shed light on the underlying mechanisms.

Parents know it best: Prediction of asthma and lung function by parental perception of early wheezing episodes.

BACKGROUND: Childhood asthma is often preceded by early wheeze. Usually, wheezing episodes are recorded retrospectively, which may induce recall bias. AIMS AND OBJECTIVES: The aim of this study was to investigate true-positive recall of parent-reported wheeze at 1 year of age, its determinants, and its implications for asthma and lung function at 6 years of age. METHODS: The PASTURE (Protection Against Allergy-Study in Rural Environments) study followed 880 children from rural areas in 5 European countries from birth to age 6 years. Wheeze symptoms in the first year were asked weekly. At age 6, parent-reported asthma diagnosis was ascertained and lung function measurements were conducted. Correct parental recall of wheeze episodes at the end of the first year was assessed for associations with lung function, asthma, and the asthma risk locus on chromosome 17q21. RESULTS: Parents correctly recalled wheeze after the first year in 54% of wheezers. This true-positive recall was determined by number of episodes, timing of the last wheeze episode, and parental asthma. Independently from these determinants, true-positive recall predicted asthma at age 6 years (odds ratio 4.54, 95% confidence interval (CI) [1.75-14.16]) and impaired lung function (ß = -0.62, 95% CI [-1.12; -0.13], P-value = .02). Associations were stronger in children with asthma risk SNPs on chromosome 17q21. CONCLUSION: Correct parental recall of wheezing episodes may reflect clinical relevance of early wheeze and its impact on subsequent asthma and lung function impairment. Questions tailored to parental perception of wheezing episodes may further enhance asthma prediction.

Bacterial microbiota of the upper respiratory tract and childhood asthma.

BACKGROUND: Patients with asthma and healthy controls differ in bacterial colonization of the respiratory tract. The upper airways have been shown to reflect colonization of the lower airways, the actual site of inflammation in asthma, which is hardly accessible in population studies. OBJECTIVE: We sought to characterize the bacterial communities at 2 sites of the upper respiratory tract obtained from children from a rural area and to relate these to asthma. METHODS: The microbiota of 327 throat and 68 nasal samples from school-age farm and nonfarm children were analyzed by 454-pyrosequencing of the bacterial 16S ribosomal RNA gene. RESULTS: Alterations in nasal microbiota but not of throat microbiota were associated with asthma. Children with asthma had lower α- and ß-diversity of the nasal microbiota as compared with healthy control children. Furthermore, asthma presence was positively associated with a specific operational taxonomic unit from the genus Moraxella in children not exposed to farming, whereas in farm children Moraxella colonization was unrelated to asthma. In nonfarm children, Moraxella colonization explained the association between bacterial diversity and asthma to a large extent. CONCLUSIONS: Asthma was mainly associated with an altered nasal microbiota characterized by lower diversity and Moraxella abundance. Children living on farms might not be susceptible to the disadvantageous effect of Moraxella. Prospective studies may clarify whether Moraxella outgrowth is a cause or a consequence of loss in diversity.

Latent class analysis reveals clinically relevant atopy phenotypes in 2 birth cohorts.

BACKGROUND: Phenotypes of childhood-onset asthma are characterized by distinct trajectories and functional features. For atopy, definition of phenotypes during childhood is less clear. OBJECTIVE: We sought to define phenotypes of atopic sensitization over the first 6 years of life using a latent class analysis (LCA) integrating 3 dimensions of atopy: allergen specificity, time course, and levels of specific IgE (sIgE). METHODS: Phenotypes were defined by means of LCA in 680 children of the Multizentrische Allergiestudie (MAS) and 766 children of the Protection against allergy: Study in Rural Environments (PASTURE) birth cohorts and compared with classical nondisjunctive definitions of seasonal, perennial, and food sensitization with respect to atopic diseases and lung function. Cytokine levels were measured in the PASTURE cohort. RESULTS: The LCA classified predominantly by type and multiplicity of sensitization (food vs inhalant), allergen combinations, and sIgE levels. Latent classes were related to atopic disease manifestations with higher sensitivity and specificity than the classical definitions. LCA detected consistently in both cohorts a distinct group of children with severe atopy characterized by high seasonal sIgE levels and a strong propensity for asthma; hay fever; eczema; and impaired lung function, also in children without an established asthma diagnosis. Severe atopy was associated with an increased IL-5/IFN-γ ratio. A path analysis among sensitized children revealed that among all features of severe atopy, only excessive sIgE production early in life affected asthma risk. CONCLUSIONS: LCA revealed a set of benign, symptomatic, and severe atopy phenotypes. The severe phenotype emerged as a latent condition with signs of a dysbalanced immune response. It determined high asthma risk through excessive sIgE production and directly affected impaired lung function.

Identification of fungal candidates for asthma protection in a large population-based study.

BACKGROUND: Exposure to molds has been related to asthma risk both positively and negatively, depending on the environmental setting. The pertinent results are based on generic markers or culturing methods although the majority of present fungi cannot be cultured under laboratory conditions. The aim of the present analysis was to assess environmental dust samples for asthma-protective fungal candidates with a comprehensive molecular technique covering also non-cultivable and non-viable fungi. METHODS: Mattress dust samples of 844 children from the GABRIELA study were analyzed by polymerase chain reaction-single-strand conformation polymorphism (PCR-SSCP) of the fungus-specific internal transcribed spacer (ITS) region. Known asthma candidate species were tested for their associations with asthma, and further gel positions were sought to explain the above. As a second, data-driven, analysis, we tested the association of each individual gel position with asthma. RESULTS: In the hypothesis-driven approach, Penicillium chrysogenum emerged with an odds ratio of 0.80 (95% confidence interval 0.66-0.96; p = 0.020). The effect size was changed by 39% toward the null when adjusting for the two bands 683 (DNA of Metschnikowia sp., Aureobasidium spp.) and 978 (DNA of Epicoccum spp., Galactomyces spp., uncultured Penicillium). The data-driven approach yielded an additional band (containing DNA of Pseudotaeniolina globosa) with reduced risk of asthma (OR = 0.80 [0.66-0.96], p = 0.012). CONCLUSIONS: A large population-based study revealed several fungal taxa with inverse associations with childhood asthma. Molds produce a variety of bioactive compounds with detrimental but also beneficial immunoregulatory capacities, which renders them promising targets for further asthma research.

The Early Development of Wheeze. Environmental Determinants and Genetic Susceptibility at 17q21.

RATIONALE: Growing up on a farm protects from childhood asthma and early wheeze. Virus-triggered wheeze in infancy predicts asthma in individuals with a genetic asthma risk associated with chromosome 17q21. OBJECTIVES: To test environmental determinants of infections and wheeze in the first year of life, potential modifications of these associations by 17q21, and the implications for different trajectories of wheeze. METHODS: We followed 983 children in rural areas of Europe from birth until age 6 years. Symptoms of wheeze, rhinitis, fever, and environmental exposures were documented with weekly diaries during year 1. Asthma at age 6 was defined as ever having a reported doctor's diagnosis. Single-nucleotide polymorphisms related to ORMDL3 (rs8076131) and GSDMB (rs7216389, rs2290400) at 17q21 were genotyped. MEASUREMENTS AND MAIN RESULTS: Early wheeze was positively associated with presence of older siblings among carriers of known asthma risk alleles at 17q21 (e.g., rs8076131) (adjusted odds ratio [aOR], 1.53; 95% confidence interval [CI], 1.16-2.01). Exposure to farm animal sheds was inversely related to wheeze (aOR, 0.44; 95% CI, 0.33-0.60). Both effects were similarly observed in children with transient wheeze up to age 3 years without subsequent development of asthma (aOR, 1.71 [95% CI, 1.09-2.67]; and aOR, 0.48 [95% CI, 0.30-0.76], respectively). CONCLUSIONS: These findings suggest that the chromosome 17q21 locus relates to episodes of acute airway obstruction common to both transient wheeze and asthma. The previously identified asthma risk alleles are the ones susceptible to environmental influences. Thus, this gene-environment interaction reveals two faces of 17q21: The same genotype constitutes genetic risk and allows for environmental protection, thereby providing options for prospective prevention strategies.

SCID patients with ARTEMIS vs RAG deficiencies following HCT: increased risk of late toxicity in ARTEMIS-deficient SCID.

A subgroup of severe combined immunodeficiencies (SCID) is characterized by lack of T and B cells and is caused by defects in genes required for T- and B-cell receptor gene rearrangement. Several of these genes are also involved in nonhomologous end joining of DNA double-strand break repair, the largest subgroup consisting of patients with T(-)B(-)NK(+)SCID due to DCLRE1C/ARTEMIS defects. We postulated that in patients with ARTEMIS deficiency, early and late complications following hematopoietic cell transplantation might be more prominent compared with patients with T(-)B(-)NK(+)SCID caused by recombination activating gene 1/2 (RAG1/2) deficiencies. We analyzed 69 patients with ARTEMIS and 76 patients with RAG1/2 deficiencies who received transplants from either HLA-identical donors without conditioning or from HLA-nonidentical donors without or with conditioning. There was no difference in survival or in the incidence or severity of acute graft-versus-host disease regardless of exposure to alkylating agents. Secondary malignancies were not observed. Immune reconstitution was comparable in both groups, however, ARTEMIS-deficient patients had a significantly higher occurrence of infections in long-term follow-up. There is a highly significant association between poor growth in ARTEMIS deficiency and use of alkylating agents. Furthermore, abnormalities in dental development and endocrine late effects were associated with alkylation therapy in ARTEMIS deficiency.

Asthma and the hygiene hypothesis. Does cleanliness matter?

RATIONALE: The early hygiene hypothesis explained the development of allergies by a lack of infections; nowadays, the aspect of excessive cleanliness in affluent populations seems to have replaced this concept. Yet, no investigation has shown that home or personal cleanliness relate to allergic diseases. OBJECTIVES: To relate personal and home cleanliness to risk of asthma and allergies. METHODS: Comprehensive questionnaire information on home or personal cleanliness and allergic health conditions at school age was collected in 399 participants of the urban Perinatale Asthma Umwelt Langzeit Allergie Studie (PAULA) birth cohort. Bacterial markers were assessed in floor and mattress dust and were related to cleanliness and allergic diseases. MEASUREMENTS AND MAIN RESULTS: Personal cleanliness was inversely related to bacterial compounds on floors and mattresses, whereas home cleanliness effectively reduced dust amount but not microbial markers. Exposure to muramic acid related to a lower prevalence of school-age asthma (adjusted odds ratio, 0.59 [95% confidence interval, 0.39; 0.90]). Mattress endotoxin in the first year of life was inversely associated with atopic sensitization (0.73 [0.56-0.96]) and asthma at school age (0.72 [0.55-0.95]). Despite the associations of dust parameters both with cleanliness and allergic health conditions, the development of allergies was not related to home and personal cleanliness. CONCLUSIONS: Bacterial exposure in house dust determined childhood asthma and allergies. Personal cleanliness, such as washing hands, and home cleanliness were objectively reflected by dust parameters in homes. However, neither personal nor home cleanliness was associated with a risk for asthma and allergies. Other microbial components in house dust not affected by personal hygiene are likely to play a role.

Consumption of unprocessed cow's milk protects infants from common respiratory infections.

BACKGROUND: Breast-feeding is protective against respiratory infections in early life. Given the co-evolutionary adaptations of humans and cattle, bovine milk might exert similar anti-infective effects in human infants. OBJECTIVE: To study effects of consumption of raw and processed cow's milk on common infections in infants. METHODS: The PASTURE birth cohort followed 983 infants from rural areas in Austria, Finland, France, Germany, and Switzerland, for the first year of life, covering 37,306 person-weeks. Consumption of different types of cow's milk and occurrence of rhinitis, respiratory tract infections, otitis, and fever were assessed by weekly health diaries. C-reactive protein levels were assessed using blood samples taken at 12 months. RESULTS: When contrasted with ultra-heat treated milk, raw milk consumption was inversely associated with occurrence of rhinitis (adjusted odds ratio from longitudinal models [95% CI]: 0.71 [0.54-0.94]), respiratory tract infections (0.77 [0.59-0.99]), otitis (0.14 [0.05-0.42]), and fever (0.69 [0.47-1.01]). Boiled farm milk showed similar but weaker associations. Industrially processed pasteurized milk was inversely associated with fever. Raw farm milk consumption was inversely associated with C-reactive protein levels at 12 months (geometric means ratio [95% CI]: 0.66 [0.45-0.98]). CONCLUSIONS: Early life consumption of raw cow's milk reduced the risk of manifest respiratory infections and fever by about 30%. If the health hazards of raw milk could be overcome, the public health impact of minimally processed but pathogen-free milk might be enormous, given the high prevalence of respiratory infections in the first year of life and the associated direct and indirect costs.

Clinical and epidemiologic phenotypes of childhood asthma.

RATIONALE: Clinical and epidemiologic approaches have identified two distinct sets of classifications for asthma and wheeze phenotypes. OBJECTIVES: To compare epidemiologic phenotype definitions identified by latent class analysis (LCA) with clinical phenotypes based on patient histories, diagnostic work-up, and treatment responses. To relate phenotypes to genetic and environmental determinants as well as diagnostic and treatment-related parameters. METHODS: LCA was performed in an international multicenter birth cohort based on yearly questions about current wheeze until age 6 years. Associations of wheeze classes and clinical phenotypes with asthma-related characteristics such as atopy, lung function, fraction of exhaled nitric oxide, and medication use were calculated using regression models. MEASUREMENTS AND MAIN RESULTS: LCA identified five classes, which verified the clinically defined wheeze phenotypes with high sensitivity and specificity; the respective receiver operating characteristics curves displayed an area under the curve ranging from 84% (frequent wheeze) to 85% (asthma diagnosis) and 87% (unremitting wheeze) to 97% (recurrent unremitting wheeze). Recurrent unremitting wheeze was the most specific and unremitting wheeze at least once the most sensitive definition. The latter identified a subgroup of children with decreased lung function, increased genetic risk, and in utero smoke exposure (ODDS RATIO, 2.03; 95% CONFIDENCE INTERVAL, 1.12-3.68; P = 0.0191), but without established asthma diagnosis and treatment. CONCLUSIONS: Clinical phenotypes were well supported by LCA analysis. The hypothesis-free LCA phenotypes were a useful reference for comparing clinical phenotypes. Thereby, we identified children with clinically conspicuous but undiagnosed disease. Because of their high area under the curve values, clinical phenotypes such as (recurrent) unremitting wheeze emerged as promising alternative asthma definitions for epidemiologic studies.

Exposure to environmental microorganisms and childhood asthma.

BACKGROUND: Children who grow up in environments that afford them a wide range of microbial exposures, such as traditional farms, are protected from childhood asthma and atopy. In previous studies, markers of microbial exposure have been inversely related to these conditions. METHODS: In two cross-sectional studies, we compared children living on farms with those in a reference group with respect to the prevalence of asthma and atopy and to the diversity of microbial exposure. In one study--PARSIFAL (Prevention of Allergy-Risk Factors for Sensitization in Children Related to Farming and Anthroposophic Lifestyle)--samples of mattress dust were screened for bacterial DNA with the use of single-strand conformation polymorphism (SSCP) analyses to detect environmental bacteria that cannot be measured by means of culture techniques. In the other study--GABRIELA (Multidisciplinary Study to Identify the Genetic and Environmental Causes of Asthma in the European Community [GABRIEL] Advanced Study)--samples of settled dust from children's rooms were evaluated for bacterial and fungal taxa with the use of culture techniques. RESULTS: In both studies, children who lived on farms had lower prevalences of asthma and atopy and were exposed to a greater variety of environmental microorganisms than the children in the reference group. In turn, diversity of microbial exposure was inversely related to the risk of asthma (odds ratio for PARSIFAL, 0.62; 95% confidence interval [CI], 0.44 to 0.89; odds ratio for GABRIELA, 0.86; 95% CI, 0.75 to 0.99). In addition, the presence of certain more circumscribed exposures was also inversely related to the risk of asthma; this included exposure to species in the fungal taxon eurotium (adjusted odds ratio, 0.37; 95% CI, 0.18 to 0.76) and to a variety of bacterial species, including Listeria monocytogenes, bacillus species, corynebacterium species, and others (adjusted odds ratio, 0.57; 95% CI, 0.38 to 0.86). CONCLUSIONS: Children living on farms were exposed to a wider range of microbes than were children in the reference group, and this exposure explains a substantial fraction of the inverse relation between asthma and growing up on a farm. (Funded by the Deutsche Forschungsgemeinschaft and the European Commission.).

Atopic sensitization in the first year of life.

BACKGROUND: There is conflicting evidence on whether allergen-specific memory is primed prenatally, whether this priming affects persistent immunologic effects, and whether it is modulated by the first environmental exposures in infancy. OBJECTIVE: We sought to explore the course of atopic sensitization between birth and 12 months of age. METHODS: Specific IgE levels for 6 food and 13 common inhalant allergens were assessed in cord blood and 1-year blood samples in the Protection against Allergy-Study in Rural Environments (PASTURE) birth cohort including 793 children from rural regions of 5 European countries. Detailed information on children's health, nutrition, and farm-related exposures was gathered by using a pregnancy questionnaire, 2 questionnaires at 2 and 12 months of age, and a diary covering the time in between. RESULTS: Sensitization was more common at 12 months of age than at birth for almost all specificities. On an individual level, persistent sensitization to the same allergens was rare (1%), whereas transient (only at birth, 11%) and incident (only at 12 months, 34%) sensitization was seen in substantial proportions of children. Associations of transient sensitization with maternal sensitization differed with the allergen specificities, with the strongest associations for food allergens (odds ratio [OR], 10.6; 95% CI, 6.0-18.6) and the weakest associations for seasonal allergens (OR, 1.64; 95% CI, 0.94-2.86). Associations of maternal sensitization with incident sensitization were also seen. Incident sensitization was related to distinct prenatal and postnatal environmental exposures of mother and child, such as consumption of cereals for incident sensitization to seasonal allergens (OR, 0.66; 95% CI, 0.50-0.88). CONCLUSION: IgE sensitization patterns change between birth and 12 months and are related to maternal and environmental influences.

Comparisons of power of statistical methods for gene-environment interaction analyses.

Any genome-wide analysis is hampered by reduced statistical power due to multiple comparisons. This is particularly true for interaction analyses, which have lower statistical power than analyses of associations. To assess gene-environment interactions in population settings we have recently proposed a statistical method based on a modified two-step approach, where first genetic loci are selected by their associations with disease and environment, respectively, and subsequently tested for interactions. We have simulated various data sets resembling real world scenarios and compared single-step and two-step approaches with respect to true positive rate (TPR) in 486 scenarios and (study-wide) false positive rate (FPR) in 252 scenarios. Our simulations confirmed that in all two-step methods the two steps are not correlated. In terms of TPR, two-step approaches combining information on gene-disease association and gene-environment association in the first step were superior to all other methods, while preserving a low FPR in over 250 million simulations under the null hypothesis. Our weighted modification yielded the highest power across various degrees of gene-environment association in the controls. An optimal threshold for step 1 depended on the interacting allele frequency and the disease prevalence. In all scenarios, the least powerful method was to proceed directly to an unbiased full interaction model, applying conventional genome-wide significance thresholds. This simulation study confirms the practical advantage of two-step approaches to interaction testing over more conventional one-step designs, at least in the context of dichotomous disease outcomes and other parameters that might apply in real-world settings.

Continuous Rather Than Solely Early Farm Exposure Protects From Hay Fever Development.

BACKGROUND: An important window of opportunity for early-life exposures has been proposed for the development of atopic eczema and asthma. OBJECTIVE: However, it is unknown whether hay fever with a peak incidence around late school age to adolescence is similarly determined very early in life. METHODS: In the Protection against Allergy-Study in Rural Environments (PASTURE) birth cohort potentially relevant exposures such as farm milk consumption and exposure to animal sheds were assessed at multiple time points from infancy to age 10.5 years and classified by repeated measure latent class analyses (n = 769). Fecal samples at ages 2 and 12 months were sequenced by 16S rRNA. Hay fever was defined by parent-reported symptoms and/or physician's diagnosis of hay fever in the last 12 months using questionnaires at 10.5 years. RESULTS: Farm children had half the risk of hay fever at 10.5 years (adjusted odds ratio [aOR] 0.50; 95% CI 0.31-0.79) than that of nonfarm children. Whereas early life events such as gut microbiome richness at 12 months (aOR 0.66; 95% CI 0.46-0.96) and exposure to animal sheds in the first 3 years of life (aOR 0.26; 95% CI 0.06-1.15) were determinants of hay fever, the continuous consumption of farm milk from infancy up to school age was necessary to exert the protective effect (aOR 0.35; 95% CI 0.17-0.72). CONCLUSIONS: While early life events determine the risk of subsequent hay fever, continuous exposure is necessary to achieve protection. These findings argue against the notion that only early life exposures set long-lasting trajectories.