Advertisement

Indoor microbial communities: Influence on asthma severity in atopic and nonatopic children

Published:February 03, 2016DOI:https://doi.org/10.1016/j.jaci.2015.11.027

      Background

      Allergic and nonallergic asthma severity in children can be affected by microbial exposures.

      Objective

      We sought to examine associations between exposures to household microbes and childhood asthma severity stratified by atopic status.

      Methods

      Participants (n = 196) were selected from a cohort of asthmatic children in Connecticut and Massachusetts. Children were grouped according to asthma severity (mild with no or minimal symptoms and medication or moderate to severe persistent) and atopic status (determined by serum IgE levels). Microbial community structure and concentrations in house dust were determined by using next-generation DNA sequencing and quantitative PCR. Logistic regression was used to explore associations between asthma severity and exposure metrics, including richness, taxa identification and quantification, community composition, and concentration of total fungi and bacteria.

      Results

      Among all children, increased asthma severity was significantly associated with an increased concentration of summed allergenic fungal species, high total fungal concentrations, and high bacterial richness by using logistic regression in addition to microbial community composition by using the distance comparison t test. Asthma severity in atopic children was associated with fungal community composition (P = .001). By using logistic regression, asthma severity in nonatopic children was associated with total fungal concentration (odds ratio, 2.40; 95% CI, 1.06-5.44). The fungal genus Volutella was associated with increased asthma severity in atopic children (P = .0001, q = 0.04). The yeast genera Kondoa might be protective; Cryptococcus species might also affect asthma severity.

      Conclusion

      Asthma severity among this cohort of children was associated with microbial exposure, and associations differed based on atopic status.

      Key words

      Abbreviations used:

      ITS (Internal transcribed spacer), OTU (Operational taxonomic unit), qPCR (Quantitative PCR)
      Asthma is a complex disease characterized by enhanced bronchial reactivity and reversible airway obstruction and can be differentiated into subtypes, including allergic and nonallergic.
      • Douwes J.
      • Gibson P.
      • Pekkanen J.
      • Pearce N.
      Non-eosinophilic asthma: importance and possible mechanisms.
      • Pearce N.
      • Pekkanen J.
      • Beasley R.
      How much asthma is really attributable to atopy?.
      • Kim H.Y.
      • DeKruyff R.H.
      • Umetsu D.T.
      The many paths to asthma: phenotype shaped by innate and adaptive immunity.
      Both of these subtypes are associated with inflammatory changes in the airways, although different cells are involved.
      • Amin K.
      • Ludviksdottir D.
      • Janson C.
      • Nettelbladt O.
      • Bjornsson E.
      • Roomans G.M.
      • et al.
      Inflammation and structural changes in the airways of patients with atopic and nonatopic asthma.
      Allergic asthma is mediated by an allergic response of the immune system to allergens from sources such as dust mite, cockroach, or fungi. IgE-mediated sensitization to allergen exposure results in a TH2 CD4+ lymphocyte response; overproduction of cytokines, including IL-4, IL-5, IL-9, and IL-13; eosinophilic airway inflammation; and asthma symptoms.
      • Kim H.Y.
      • DeKruyff R.H.
      • Umetsu D.T.
      The many paths to asthma: phenotype shaped by innate and adaptive immunity.
      • Umetsu D.T.
      • McIntire J.J.
      • Akbari O.
      • Macaubas C.
      • DeKruyff R.H.
      Asthma: an epidemic of dysregulated immunity.
      • Levine S.J.
      • Wenzel S.E.
      Narrative review: the role of Th2 immune pathway modulation in the treatment of severe asthma and its phenotypes.
      Nonallergic asthma is an acute inflammatory response dominated by neutrophils (as opposed to eosinophils) and might be responsible for half of asthma cases.
      • Douwes J.
      • Gibson P.
      • Pekkanen J.
      • Pearce N.
      Non-eosinophilic asthma: importance and possible mechanisms.
      Asthma severity is influenced by environmental exposures, including exposures to microbes. For instance, bacterial endotoxin exposure can affect asthma severity.
      • Michel O.
      • Kips J.
      • Duchateau J.
      • Vertongen F.
      • Robert L.
      • Collet H.
      • et al.
      Severity of asthma is related to endotoxin in house dust.
      • Zambelli-Weiner A.
      • Stockton M.
      • Ehrlich E.
      • Whitmer M.
      • Lewis D.
      • Olenchock S.
      • et al.
      House dust endotoxin exposure mitigates asthma severity in a tropical environment.
      Fungal chitin elicits both an innate and adaptive immune response,
      • Goldman D.L.
      • Vicencio A.G.
      The chitin connection.
      and fungal glucan exposure has been associated with increased asthma severity in children.
      • Blatter J.
      • Forno E.
      • Brehm J.
      • Acosta-Pérez E.
      • Alvarez M.
      • Colón-Semidey A.
      • et al.
      Fungal exposure, atopy, and asthma exacerbations in Puerto Rican children.
      • Douwes J.
      • Zuidhof A.
      • Doekes G.
      • van der Zee S.
      • Wouters I.
      • Boezen M.H.
      • et al.
      (1 → 3)- β -d-glucan and endotoxin in house dust and peak flow variability in children.
      Also, exposure to mold or dampness in a home is associated with asthma severity in children and additional respiratory outcomes in subjects with and without allergy.
      • Mendell M.J.
      • Mirer A.G.
      • Cheung K.
      • Tong M.
      • Douwes J.
      Respiratory and allergic health effects of dampness, mold, and dampness-related agents: a review of the epidemiologic evidence.
      Previous findings suggest that response to specific microbial agents can vary with allergic status. For instance, asthma severity is increased in children with allergy who are exposed to the sensitizing allergen,
      • Murray C.S.
      • Poletti G.
      • Kebadze T.
      • Morris J.
      • Woodcock A.
      • Johnston S.L.
      • et al.
      Study of modifiable risk factors for asthma exacerbations: virus infection and allergen exposure increase the risk of asthma hospital admissions in children.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      whereas nonallergic asthma might be influenced by exposure to microbial respiratory irritants, such as bacterial endotoxin.
      • Douwes J.
      • Gibson P.
      • Pekkanen J.
      • Pearce N.
      Non-eosinophilic asthma: importance and possible mechanisms.
      To further investigate this hypothesis, microbial exposures can now be more completely characterized by using next-generation DNA sequencing technologies to better elucidate diversity and taxa identification in homes.
      In the current analysis we examined associations between microbial exposures and asthma severity stratified by atopic status. We hypothesized that in atopic children asthma severity would be associated with microbial community composition or allergenic fungal taxa caused by the allergic response of the atopic lung. Additionally, we also hypothesized that in nonatopic children asthma severity would be associated with total concentration of exposure because of high concentrations of microbial irritants. We measured fungal and bacterial community characteristics in homes of asthmatic children in Connecticut and Massachusetts. We considered fungal and bacterial taxa, diversity, and concentration measured by using next-generation DNA sequencing and quantitative PCR (qPCR). Asthma severity was measured prospectively during the month after sampling, and atopic status was determined based on age-adjusted serum IgE concentration, which is involved in allergic asthma.
      • Kim H.Y.
      • DeKruyff R.H.
      • Umetsu D.T.
      The many paths to asthma: phenotype shaped by innate and adaptive immunity.
      Within the 3 groups (all children, atopic children, and nonatopic children) of this cohort, we determined associations between microbial community features and asthma severity based on asthma subtype.

      Methods

       Study population

      The current analysis was designed to select participants of the parent study at the extremes of asthma severity stratified by atopic status while controlling for socioeconomic status, asthma medication use, and sampling season. Subjects for the present analysis (n = 200) were selected from participants in a study of air quality and respiratory health.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      Between 2006 and 2009, children with asthma and their families were recruited from Connecticut and southwestern Massachusetts.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      Eligibility criteria included age between 5 and 10 years, a caregiver who spoke English, active asthma in the child, and consent to a blood draw for allergy testing. Active asthma required at least 2 of the following criteria: (1) a physician's diagnosis; (2) asthma symptoms, including wheeze, persistent cough, chest tightness, or shortness of breath, in the previous 12 months; and (3) use of prescription asthma medication within the previous 12 months. Based on asthma symptoms and medication use recorded over 1 month (see below), asthma severity during the study was determined by using a 5-level score based on Global Initiative for Asthma Guidelines

      US Department of Health and Human Services. Global Initiative for Asthma, Global Strategy for Asthma Management and Prevention. 2002. Available at: http://www.ginasthma.org/Guidelines/guidelines-2002-original%3a-workshop-report%2c-global-strategy-for-asthma-management-and-prevention.html. Accessed January 1, 2011.

      calculated with a combination of symptom frequency and asthma medication use.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      By including asthma medication use as part of the severity metric, a child with high asthma medication use and consequently low symptom frequency would be considered to have severe asthma, as would a child with high symptom frequency but low asthma medication use.
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      A score of 0 or 1 was considered mild asthma, and a score of 3 or 4 was considered severe asthma; children with a score of 2 were not included in this analysis. Atopic status was defined as serum IgE levels of greater than age-adjusted levels

      Mayo Clinic. Immunoglobulin E (IgE) serum. 2006. Available at: http://mayomedicallaboratories.com/test-catalog/pring.pht?unit_code=8159. Accessed April 5, 2011.

      or allergen-specific IgE sensitivity to a panel of 10 allergens
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      determined based on analysis of blood drawn on the same date that sampling occurred and on the first day of the month of the asthma monitoring period. Total IgE levels were measured, as were levels of IgE specific to the following allergens: Penicillium notatum (Penicillium chrysogenum), Cladosporium herbarum, Der p 1, Der f 1, Fel d 1, Can f 1, Bla g 1, meadow grass (Kentucky blue, Poa pratensis), ragweed (Ambrosia elatior), and egg.
      In the original study a total of 1642 children were eligible, and 1401 enrolled. Subjects selected for the current analysis were first restricted to 1233, with complete information on asthma severity during the first 1-month monitoring period, allergy test results, and successful household environmental sampling at the enrollment visit.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      To control for season, samples were next restricted to the 587 collected during the nonheating season (May-October). Outdoor fungal populations are consistent in this region over these months,
      • Yamamoto N.
      • Bibby K.
      • Qian J.
      • Hospodsky D.
      • Rismani-Yazdi H.
      • Nazaroff W.W.
      • et al.
      Particle-size distributions and seasonal diversity of allergenic and pathogenic fungi in outdoor air.
      and outdoor populations influence indoor populations.
      • Adams R.I.
      • Bhangar S.
      • Pasut W.
      • Arens E.A.
      • Taylor J.W.
      • Lindow S.E.
      • et al.
      Chamber bioaerosol study: outdoor air and human occupants as sources of indoor airborne microbes.
      During these months in the study region, average monthly temperatures range from 11.2°C and 23.1°C, and average monthly precipitation ranges from 9.86 to 11.1 cm.

      National Oceanic and Atmospheric Administration, National Centers for Environmental Information. 1981-2010 U.S. Climate Normals. Hartford Bradley International Airport, CT, US. Available at: http://www.ncdc.noaa.gov/cdo-web/datatools/normals. Accessed June 30, 2015.

      These samples were equally weighted by house type (single family, n = 297; multifamily n = 290), a variable used as a marker for socioeconomic status.
      • Belanger K.
      • Gent J.F.
      • Triche E.W.
      • Bracken M.B.
      • Leaderer B.P.
      Association of indoor nitrogen dioxide exposure with respiratory symptoms in children with asthma.
      Within each house type, 25 subjects were randomly selected among 4 categories: (1) atopic/mild asthma, (2) atopic/severe asthma, (3) nonatopic/mild asthma, and (4) nonatopic/severe asthma (see Fig E1 in this article's Online Repository at www.jacionline.org). Of 200 samples selected, it was not possible to analyze 2, and the fungal DNA in 2 additional samples could not be amplified, leaving 196 samples. The Human Investigation Committee of Yale University approved the study. Written informed consent was obtained from the participating child's primary caregiver, and all children age 7 years and older gave assent.

       Home visit and dust collection

      The initial home visit included a respondent interview (primary caregiver of the participating child) and environmental sampling (described in detail by Belanger et al
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      ). Questions from the home interview included demographic information (eg, race and sex), information about the home (eg, pests, pets, smoking, and presence of mold or water leaks), and the child's medical history. The health outcome monitoring period was the month after the home visit. During this time, the respondent was asked to record the child's daily asthma symptoms and medication use. This information was collected by telephone at the end of the month.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      During the home visit, dust was collected from the main living area in which the child spent the most time (see Belanger et al
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      ). A Eureka Mighty Mite (Eureka, Bloomington, Ill) was fitted with a 19 × 90–mm Whatman cellulose extraction thimble (Whatman, Tewkesbury, Mass), and dust was sampled with a standardized protocol from an exposed seat cushion and the seat back and arms of a couch or chair for 3 minutes and a 1-m2 section of floor for 2 minutes.
      • Belanger K.
      • Beckett W.
      • Triche E.
      • Bracken M.B.
      • Holford T.
      • Ren P.
      • et al.
      Symptoms of wheeze and persistent cough in the first year of life: associations with indoor allergens, air contaminants, and maternal history of asthma.
      • Leaderer B.P.
      • Belanger K.
      • Triche E.
      • Holford T.
      • Gold D.R.
      • Kim Y.
      • et al.
      Dust mite, cockroach, cat, and dog allergen concentrations in homes of asthmatic children in the northeastern United States: impact of socioeconomic factors and population density.
      Dust was stored at −80°C before analysis.

       DNA extraction, sequencing, and analysis

      DNA extraction, sequencing, and analysis, in addition to endotoxin and (1-3)-β-D-glucan measurement, have been described previously.
      • Dannemiller K.C.
      • Gent J.F.
      • Leaderer B.P.
      • Peccia J.
      Influence of housing characteristics on bacterial and fungal communities in homes of asthmatic children.
      Briefly, sieved dust was extracted
      • Yamamoto N.
      • Bibby K.
      • Qian J.
      • Hospodsky D.
      • Rismani-Yazdi H.
      • Nazaroff W.W.
      • et al.
      Particle-size distributions and seasonal diversity of allergenic and pathogenic fungi in outdoor air.
      and determined to have no PCR inhibition through testing spiked samples. The internal transcribed spacer (ITS) region was amplified with ITS1F and ITS4 primers
      • Manter D.
      • Vivanco J.
      Use of the ITS primers, ITS1F and ITS4, to characterize fungal abundance and diversity in mixed-template samples by qPCR and length heterogeneity analysis.
      • Larena I.
      • Salazar O.
      • Gonzalez V.
      • Julian M.C.
      • Rubio V.
      Design of a primer for ribosomal DNA internal transcribed spacer with enhanced specificity for ascomycetes.
      for fungal sequencing. PCR and laboratory negative controls did not amplify. Cleaned, normalized, and pooled samples were sequenced on the 454 GS FLX Titanium DNA sequencing platform (454 Life Sciences, Branford, Conn). Sequences have been archived with accession numbers ERP005149 and ERP002369 and were analyzed with QIIME, version 1.7,
      • Caporaso J.G.
      • Kuczynski J.
      • Stombaugh J.
      • Bittinger K.
      • Bushman F.D.
      • Costello E.K.
      • et al.
      QIIME allows analysis of high-throughput community sequencing data.
      and FHiTINGS, version 1.1.
      • Dannemiller K.C.
      • Reeves D.
      • Bibby K.
      • Yamamoto N.
      • Peccia J.
      Fungal high-throughput taxonomic identification tool for use with next-generation sequencing (FHiTINGS).
      Richness was based on the number of observed species among 450 sequences per sample. Principal coordinate plots were not shown due to an arch distortion in the fungal analysis.
      • Gauch H.G.
      • Whittaker R.H.
      • Wentworth T.R.
      A comparative study of reciprocal averaging and other ordination techniques.
      Distance similarity comparison was based on the Morisita Horn
      • Horn H.S.
      Measurement of “overlap” in comparative ecological studies.
      distance from all quality-trimmed reads. Endotoxin levels were measured with the quantitative kinetic chromogenic LAL assay (Associates of Cape Cod, East Falmouth, Mass), and (1-3)-β-D-glucan levels were measured with the quantitative kinetic Glucatell lysate assay (Associates of Cape Cod), with details provided previously.
      • Dannemiller K.C.
      • Gent J.F.
      • Leaderer B.P.
      • Peccia J.
      Influence of housing characteristics on bacterial and fungal communities in homes of asthmatic children.
      The V4 region of bacterial 16S rRNA was sequenced with 515F/806R primers
      • Caporaso J.G.
      • Lauber C.L.
      • Walters W.A.
      • Berg-Lyons D.
      • Lozupone C.A.
      • Turnbaugh P.J.
      • et al.
      Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample.
      on an Ion Torrent Personal Genome Machine with an Ion 318 chip kit v2. PCR and laboratory negative controls did not amplify. Sequences were archived in the European Nucleotide Archive under accession number ERP005148. Analysis of bacterial sequences was conducted in QIIME, version 1.7.
      • Caporaso J.G.
      • Kuczynski J.
      • Stombaugh J.
      • Bittinger K.
      • Bushman F.D.
      • Costello E.K.
      • et al.
      QIIME allows analysis of high-throughput community sequencing data.
      Richness was based on the number of operational taxonomic units (OTUs) among 2500 sequences per sample. Distance similarity comparison was based on unweighted UniFrac distance
      • Lozupone C.
      • Knight R.
      UniFrac: a new phylogenetic method for comparing microbial communities.
      from all quality-trimmed reads.

       Statistical analysis

      Logistic regression analysis (PROC LOGISTIC) in SAS software (version 9.2; SAS Institute, Cary, NC) was used to examine associations between the dichotomous health outcome of asthma severity (mild or severe) and microbial exposures. Associations between microbial exposures and atopic status (greater than or less than age-adjusted serum IgE levels) were investigated as background information. Separate models were used for each microbial exposure variable. Models were unadjusted because important covariates, such as asthma maintenance medication use, socioeconomic status, and season, were incorporated in the outcome variable and subject selection, as described above. One advantage of using this asthma severity score as an outcome is that, unlike symptom frequency or medication use alone, the score is not associated with socioeconomic variables.
      • Belanger K.
      • Holford T.R.
      • Gent J.F.
      • Hill M.E.
      • Kezik J.M.
      • Leaderer B.P.
      Household levels of nitrogen dioxide and pediatric asthma severity.
      Exposure to environmental tobacco smoke and other allergens was examined in models in which there was a significant association between microbial exposure and the health outcome. Addition of these covariates did not significantly change the results. Variables were dichotomized because of skewed distributions of sequencing data and for consistency throughout. Microbial exposures included dichotomous variables of richness (number of OTUs), total concentration, taxonomic associations, summed allergenic fungal species/genera, inflammatory agents (endotoxin/[1-3]-β-D-glucan), and community composition for fungi and bacteria (Table I). Total fungal and bacterial concentration variables were dichotomized at the median to equally separate high and low exposure. For taxonomic associations (including the group of summed allergenic species/genera), absolute concentration values were calculated by multiplying relative abundance values from sequencing by concentration values from qPCR.
      • Dannemiller K.C.
      • Lang-Yona N.
      • Yamamoto N.
      • Rudich Y.
      • Peccia J.
      Combining real-time PCR and next-generation DNA sequencing to provide quantitative comparisons of fungal aerosol populations.
      These and endotoxin/glucan variables were dichotomized at the 75th percentile to identify homes with high exposure. Dichotomized quantitative variables for each taxa were used in comparisons of asthma severity among all, atopic, and nonatopic children, as well as atopic status, by using the PROC MULTTEST command in SAS with the “pfdr” option and the mean option in the test statement to request the t test for the mean. This analysis generates a q value, which is similar to a P value that has been adjusted for multiple comparisons and measures significance by using the false discovery rate instead of the false-positive rate.
      • Storey J.D.
      • Tibshirani R.
      Statistical significance for genomewide studies.
      Microbial community composition (distance matrix) comparisons were conducted by comparing the mean distance (Morisita Horn or unweighted UniFrac) between homes within a group (eg, severe asthma to severe asthma) with the mean distance between homes in different groups (eg, severe asthma to mild asthma) as a reference. Communities with statistically significantly lower mean distance compared with the reference by using the 2-sample, 2-sided t test were considered to share similarities in composition.
      Table IMicrobial variables used in this study
      Split value
      Allergenic species, allergenic genera, (1-3)-β-D-glucan, and endotoxin were split at the 75th percentile to explore response to high exposures. Microbial richness and concentration were split at the median to compare high and low exposure.
      All childrenAtopic childrenNonatopic children
      Total
      Total values of less than 196 are due to missing data. Samples with less than 450 (fungi) or 2500 (bacteria) sequences per sample were excluded in richness calculations. The (1-3)-β-D-glucan and endotoxin measurements were available for 121 homes.
      YesTotalYesTotalYes
      Fungi
       Low fungal richness90 OTUs16478 (48%)8542 (49%)7936 (46%)
       High allergenic species20,172 spore equivalents/mg19649 (25%)9922 (22%)9727 (28%)
       High allergenic genera66,646 spore equivalents/mg19649 (25%)9924 (24%)9725 (26%)
       High fungal concentration153,773 spore equivalents/mg19698 (50%)9948 (48%)9750 (52%)
       High (1-3)-β-D-glucan178 μg/g12130 (25%)5513 (24%)6617 (26%)
      Bacteria
       Low bacterial richness738 OTUs18190 (50%)9244 (48%)8946 (52%)
       High bacterial concentration195,809 genomes19698 (50%)9943 (43%)9755 (57%)
       High endotoxin0.44 EU/mg12130 (25%)5514 (25%)6616 (24%)
      Allergenic species, allergenic genera, (1-3)-β-D-glucan, and endotoxin were split at the 75th percentile to explore response to high exposures. Microbial richness and concentration were split at the median to compare high and low exposure.
      Total values of less than 196 are due to missing data. Samples with less than 450 (fungi) or 2500 (bacteria) sequences per sample were excluded in richness calculations. The (1-3)-β-D-glucan and endotoxin measurements were available for 121 homes.
      Pearson χ2 analysis was used to test that the distributions were representative between the atopy/asthma severity categories for subjects included in the present analysis, as well as that the distributions were representative between the analysis group compared with the original cohort population for selected covariates.

      Results

      Demographic information for the participants is shown in Table II. By using χ2 analysis, distributions were representative for the variables age, race, sex, mother's education level, or smoking in the home between subjects in the 4 atopy/asthma severity categories (P > .05). Compared with the parent study cohort, sex distribution was representative (P = .47), although the current study group is less white (P < .0001) and the mother's education level is lower (P = .004).
      Table IIDemographic information of asthmatic children participating in this study (n = 196)
      CharacteristicsAll children, no. (%)Atopic children, no. (%)Nonatopic children, no. (%)
      All subjects196 (100)99 (50.5)97 (49.5)
      Asthma severity
       Mild97 (49.5)50 (50.5)47 (48.5)
       Severe99 (50.5)49 (49.5)50 (51.5)
      Sex
       Male119 (60.7)64 (64.6)55 (56.7)
       Female77 (39.3)35 (35.4)42 (43.3)
      Child's age (y)
       542 (21.4)18 (18.2)24 (24.7)
       633 (16.8)20 (20.2)13 (13.4)
       731 (15.8)11 (11.1)20 (20.6)
       835 (17.9)13 (13.1)22 (22.7)
       931 (15.8)21 (21.2)10 (10.3)
       1024 (12.2)16 (16.2)8 (8.2)
      Maintenance medication use
       No117 (59.7)63 (63.6)54 (55.7)
       Yes79 (40.3)36 (36.4)43 (44.3)
      Ethnicity
       White49 (25.0)26 (26.3)23 (23.7)
       Black52 (26.5)27 (27.3)25 (25.8)
       Hispanic
      Hispanics were predominately (87.1%) of Puerto Rican descent.
      85 (43.4)41 (41.4)44 (45.4)
       Mixed and other10 (5.1)5 (5.1)5 (5.2)
      Mother's education (y)
       <1240 (20.4)21 (21.2)19 (19.6)
       12-15116 (59.2)59 (59.6)57 (58.8)
       ≥1640 (20.4)19 (19.2)21 (21.6)
      Housing type
       Single family94 (48.0)51 (51.5)43 (44.3)
       Multifamily102 (52.0)48 (48.5)54 (55.7)
      Smoking in the home
       No174 (88.8)89 (89.9)85 (87.6)
       Yes20 (10.2)10 (10.1)10 (10.3)
       Missing2 (0.5)0 (0.0)2 (2.1)
      Month of sample collection
       May53 (27.0)25 (25.3)28 (28.9)
       June47 (24.0)24 (24.2)23 (23.7)
       July38 (19.4)24 (24.2)14 (14.4)
       August14 (7.1)5 (5.1)9 (9.3)
       September11 (5.6)5 (5.1)6 (6.2)
       October33 (16.8)16 (16.2)17 (17.5)
      Home location
      Urban locations were defined as having less than 45% single-family homes, and suburban locations were defined as having at least 45% single-family homes.
       Urban112 (57)56 (57)56 (58)
       Suburban84 (43)43 (43)41 (42)
      Hispanics were predominately (87.1%) of Puerto Rican descent.
      Urban locations were defined as having less than 45% single-family homes, and suburban locations were defined as having at least 45% single-family homes.
      Sequencing results have been described previously.
      • Dannemiller K.C.
      • Gent J.F.
      • Leaderer B.P.
      • Peccia J.
      Influence of housing characteristics on bacterial and fungal communities in homes of asthmatic children.
      We identified 10 allergenic species and 13 genera containing allergenic species in these homes using a list of previously identified allergenic fungi.
      • Yamamoto N.
      • Bibby K.
      • Qian J.
      • Hospodsky D.
      • Rismani-Yazdi H.
      • Nazaroff W.W.
      • et al.
      Particle-size distributions and seasonal diversity of allergenic and pathogenic fungi in outdoor air.
      Allergenic species found included Alternaria alternata, Alternaria brassicae, Candida albicans, Cladosporium cladosporioides, Epicoccum nigrum, Penicillium brevicompactum, Ulocladium atrum, Ulocladium chartarum, Malassezia sympodialis, and Rhodotorula mucilaginosa. Genera containing allergenic species included the aforementioned genera, in addition to Aspergillus, Curvularia, Fusarium, Pleospora, and Stachybotrys species.

       All children

      Among all children (both atopic and nonatopic), increased asthma severity was associated with a high concentration of summed allergenic species (P = .007) and high total fungal concentration by using logistic regression (P = .02), as well as fungal community composition by using the distance comparison t test (P = .003; Table III and Fig 1, A). For community composition comparison, community similarity is compared among homes within a group (Figs 1 and 2, bars in bar graphs) with the similarity between groups (Figs 1 and 2, solid line in bar graphs). If the mean distance within a group is significantly less than that between groups, then there is low compositional variation (ie, microbial communities are similar) in that group. For fungi, mean distance was lower than expected by chance for homes of children with severe asthma compared with homes of children with mild asthma, which indicates that microbial communities in homes of patients with severe asthma share similarities (Fig 1, A).
      Table IIIORs and 95% CIs from logistic regression models of the association between asthma severity and microbial concentration, taxa, and diversity
      All children, OR (95% CI)Atopic children, OR (95% CI)Nonatopic children, OR (95% CI)
      Fungi
       Low fungal richness1.29 (0.70-2.38)1.05 (0.45-2.46)1.61 (0.66-3.93)
       Allergenic species2.53 (1.28-5.00)2.71 (0.99-7.39)2.38 (0.94-6.01)
       Allergenic genera1.79 (0.92-3.45)2.55 (0.97-6.67)1.27 (0.51-3.18)
       Fungal concentration2.02 (1.14-3.56)1.69 (0.77-3.75)2.40 (1.06-5.44)
       (1-3)-β-D-glucan0.55 (0.24-1.26)0.47 (0.13-1.68)0.60 (0.20-1.83)
      Bacteria
       Low bacterial richness0.55 (0.30-0.99)0.38 (0.16-0.90)0.77 (0.34-1.75)
       Bacterial concentration1.04 (0.60-1.82)1.13 (0.51-2.49)0.94 (0.42-2.11)
       Endotoxin2.46 (0.99-6.11)1.89 (0.54-6.62)3.41 (0.86-13.45)
      All variables are dichotomous. ORs show the risk of being in the severe asthma category compared with the mild asthma category (reference group). Associations with P values of less than .05 are in boldface. Separate models were used for each microbial exposure variable. Models were unadjusted because important covariates, such as asthma maintenance medication use, socioeconomic status, and season, were incorporated in the outcome variable and subject selection, as described above. Exposure to environmental tobacco smoke and other allergens was examined in models in which there was a significant association between microbial exposure and the health outcome. Addition of these covariates did not significantly change the results.
      OR, Odds ratio.
      Figure thumbnail gr1
      Fig 1Community composition analysis for fungi (A) and bacteria (B) among all children. Bar graphs compare β-diversity within groups (bars) with diversity between groups (solid lines). Bars statistically lower than the solid lines indicate low variation (similarity) in community composition within a group. Error bars and dotted lines represent SEs.
      Figure thumbnail gr2
      Fig 2Community composition analysis for fungi and bacteria among the 4 groups (A-D) of children in this study. Bar graphs compare β-diversity within groups (bars) with diversity between groups (solid lines). Bars statistically lower than the solid lines indicate low variation (similarity) in community composition within a group (eg, Fig 2, A; fungi in house dust from atopic children with severe asthma have similar community composition). Error bars and dotted lines represent SEs.
      Among all children, increased bacterial richness (number of unique bacterial OTUs) was associated with increased asthma severity (P = .046), and there was a trend toward increased severity with increased endotoxin exposure (P = .052, Table III). There was less variation in community composition (communities were more similar) in homes of children with severe asthma (Fig 1, B). No bacterial OTUs were significantly associated with asthma severity among atopic, nonatopic, or all children.

       Atopic children

      Asthma severity was also considered separately in atopic and nonatopic children. Among atopic children, there was also a trend toward increased asthma severity with high concentrations of summed allergenic species and genera, as determined by using logistic regression (P = .052 and .057, Table III). The fungal communities in house dust from atopic children with severe asthma had low compositional variation, as determined by using the distance comparison t test (P = .001; Fig 2, A). This indicates that homes of children with severe asthma share fungal community similarities. Table IV lists fungal taxa that are associated with asthma severity in atopic children. Only the genus Volutella remained statistically significantly associated with increased asthma severity after control for multiple comparisons (q < 0.05).
      Table IVFungal taxa associated with asthma severity status in atopic children through multiple-comparison analysis
      Taxa associated with severe asthma in atopic childrenTaxa associated with mild asthma in atopic children (protective)
      TaxonP valueq valueTaxonP valueq value
      SpeciesSpecies
      Volutella colletotrichoides.00030.12Cryptococcus paraflavus.0030.44
      Thelebolus microsporus.0020.41Aspergillus ruber.0080.66
      Cryptococcus nyarrowii.0070.66Cryptococcus heimaeyensis.010.72
      Trichosporon porosum.010.72Xenophacidiella pseudocatenata.020.72
      Candida galli.020.72Cladia schizopora.020.72
      Cryptococcus skinneri.020.72Polyscytalum algarvense.020.72
      Scoliciosporum umbrinum.040.72Plectosphaerella cucumerina.020.72
      Mycocalicium victoriae.040.72Calyptrozyma arxii.030.72
      Cryptococcus podzolicus.040.72Bahusakala australiensis.040.72
      Teratosphaeria ohnowa.040.72Rhizoplaca aspidophora.040.72
      Lambertella tubulosa.0460.72Strelitziana albiziae.040.72
      Pseudaegerita viridis.0470.72Genus
      Rhizocarpon petraeum.0470.72Kondoa.0020.27
      Leptosphaerulina americana.0470.72Exobasidium.0060.66
      Epicoccum nigrum.0470.72Strelitziana.020.66
      GenusXenophacidiella.020.66
      Volutella.00010.04Cladia.020.66
      Thelebolus.020.66Polyscytalum.020.66
      Lambertella.030.66Plectosphaerella.020.66
      Cochliobolus.040.66Microcyclospora.020.66
      Cordyceps.0470.66Calyptrozyma.030.66
      Phaeomoniella.0470.66Bahusakala.040.66
      Phialocephala.0470.66Phaeotheca.040.66
      Rhizocarpon.0470.66Collophora.040.66
      Epicoccum.0470.66Curreya.040.66
      Polyporus.040.66
      Rhizoplaca.040.66
      All variables are dichotomous.
      The association between increased bacterial richness and increased asthma severity was true for atopic children only (Table III).

       Nonatopic children

      In nonatopic children total fungal concentrations were higher in homes of children with severe asthma by using logistic regression (P = .04, Table III). Fungal community composition did not differ in homes of nonatopic children with severe compared with mild asthma (Fig 2, B). No fungal taxa were associated with asthma severity among nonatopic children or all children combined, although there was a trend toward increased asthma severity with high concentrations of allergenic species (P = .07, Table III).
      In nonatopic children there was lower variation in bacterial community composition in homes of patients with severe asthma, indicating that bacterial communities in these homes share similarities (Fig 2, D). There was also a trend toward increased asthma severity with exposure to increased bacterial endotoxin (P = .08, Table III).

      Discussion

      Our results suggest that asthma severity in children is associated with microbial exposures. The observed response might be different in atopic compared with nonatopic children, which deserves further study. Among all children, associations were found with both fungal and bacterial variables (Fig 1, A, and Table III). Asthma severity in atopic children was associated with fungal community composition (Fig 2). There was also a trend toward increased asthma severity with exposure to summed allergenic fungal species and genera. Nonatopic asthma severity was associated with total fungal concentration and possibly with exposure to summed allergenic fungal species (Table III). Asthma severity was associated with bacterial richness in atopic children (Table III) and bacterial community composition in nonatopic children (Fig 2). Bacterial community composition was more similar in homes of nonatopic children with severe asthma than in homes of children with mild asthma (Fig 2, D).
      In children with nonatopic asthma, exposures to microbial inflammatory agents, such as bacterial endotoxin, can increase asthma severity.
      • Douwes J.
      • Gibson P.
      • Pekkanen J.
      • Pearce N.
      Non-eosinophilic asthma: importance and possible mechanisms.
      Here, a trend was seen with endotoxin in all children and in nonatopic children (Table III). Because different species of bacteria release different types or no endotoxin,
      • Trent M.S.
      • Stead C.M.
      • Tran A.X.
      • Hankins J.V.
      Invited review: diversity of endotoxin and its impact on pathogenesis.
      different bacterial communities also likely vary in amounts and types of endotoxin present. This might be the reason that bacterial community composition was also associated with asthma severity in nonatopic children, but this hypothesis should be evaluated in future studies.
      In atopic children exposure to a specific fungal community composition was associated with increased asthma severity, which might be linked to exposure to fungal allergens. Summed concentrations of known allergenic fungal species and genera demonstrated a trend toward increased asthma severity in this group. However, for allergenic fungal species, this trend was also present in nonatopic children and significant for all children (Table III), which indicates that these particular allergenic species might have additional nonallergenic properties that are associated with asthma severity.
      There is a nonsignificant trend in atopic children toward the same association between asthma severity and fungal concentration seen in nonatopic children. This might be due to the association between a higher total fungal concentration and a higher concentration of the most abundant species, some of which are allergenic, which was previously shown in these data.
      • Dannemiller K.C.
      • Gent J.F.
      • Leaderer B.P.
      • Peccia J.
      Influence of housing characteristics on bacterial and fungal communities in homes of asthmatic children.
      Similarly, in nonatopic children summed allergenic species/genera have a nonsignificant trend toward the same association between asthma severity and fungal concentration seen in atopic children. This might also be due to the association between allergenic species/genera and total concentration. In the current analysis fungal concentration has a stronger association with asthma severity in nonatopic compared with atopic children, and the fungal community/taxa have stronger associations with asthma severity in atopic compared with nonatopic children. These results deserve further analysis and await replication in future studies.
      Taxonomic associations with asthma severity were not seen, with the exception of association of the fungal genus Volutella with increased asthma severity among atopic children (Table IV). This might be due to 3 factors: (1) children in different homes experience different exposures and likely become allergic to or irritated by specific agents that vary between homes, (2) we did not have enough statistical power to find associations of rare taxa in our number of homes to identify these more subtle distinctions, and (3) the need to control for multiple comparisons among the hundreds of identified species complicates determination of statistical significance. However, several other q values that were low (eg, q < 0.50) indicate that there are likely other taxa that contribute to asthma severity that cannot be conclusively identified in the current analysis. Taxa listed in Table IV with P values of less than .05 can be considered to be potentially associated with asthma severity and deserve attention in future analyses. Taxa listed on the right in Table IV might be protective against severe asthma. Among the nearly significant associations, other trends are beginning to emerge, such as the potential importance of yeasts. For instance, the genus Kondoa was identified here as potentially protective against severe asthma. Here, Kondoa species was identified in 13 homes, all of which had a child with mild asthma. In 9 (69%) of these homes, the child was atopic. Additionally, several Cryptococcus species were associated with both increased and decreased asthma severity. Both Kondoa and Cryptococcus are yeast genera in the phylum Basidiomycota, which has previously been understudied with regard to allergic disease caused by difficulties in culturing.
      • Simon-Nobbe B.
      • Denk U.
      • Pöll V.
      • Rid R.
      • Breitenbach M.
      The spectrum of fungal allergy.
      Many yeasts, including Cryptococcus species, are inhabitants of human skin.
      • Findley K.
      • Oh J.
      • Yang J.
      • Conlan S.
      • Deming C.
      • Meyer J.A.
      • et al.
      Topographic diversity of fungal and bacterial communities in human skin.
      This “mycobiome” affects human health
      • Cui L.
      • Morris A.
      • Ghedin E.
      The human mycobiome in health and disease.
      and might be inoculated from the environment, including house dust. These potential trends can be used to inform future studies that seek to verify these results, as well as potentially establish mechanisms and definitive associations.

       Other considerations, strengths, and limitations

      Asthma severity might not be associated with the same environmental exposures associated with asthma development. For example, bacterial endotoxin exposure can prevent asthma development but might increase the severity of existing asthma.
      • Liu A.H.
      Endotoxin exposure in allergy and asthma: Reconciling a paradox.
      Early-life exposure to certain bacteria might prevent allergic disease.
      • Lynch S.V.
      • Wood R.A.
      • Boushey H.
      • Bacharier L.B.
      • Bloomberg G.R.
      • Kattan M.
      • et al.
      Effects of early-life exposure to allergens and bacteria on recurrent wheeze and atopy in urban children.
      • Fujimura K.E.
      • Demoor T.
      • Rauch M.
      • Faruqi A.A.
      • Jang S.
      • Johnson C.C.
      • et al.
      House dust exposure mediates gut microbiome Lactobacillus enrichment and airway immune defense against allergens and virus infection.
      Fungal diversity was previously associated with asthma development.
      • Dannemiller K.
      • Mendell M.J.
      • Macher J.M.
      • Kumagai K.
      • Bradman A.
      • Holland N.
      • et al.
      Next-generation DNA sequencing reveals that low fungal diversity in house dust is associated with childhood asthma development.
      but we found no association with asthma severity.
      Limitations of this study include the selection bias associated with a convenience sample. By design for the current analysis, we restricted selection of samples to control for important potential confounders, including asthma maintenance medication use (by using the asthma severity score), socioeconomic status (by using housing type), and sampling season (restricted to nonheating season). Therefore these variables were not included in the statistical models directly but rather indirectly through matching. There might be additional unmeasured potential confounders. Other potential confounders might include other allergens or environmental tobacco smoke. For other allergen exposures, dust mite, cat, dog, and cockroach allergen levels are associated with house type,
      • Leaderer B.P.
      • Belanger K.
      • Triche E.
      • Holford T.
      • Gold D.R.
      • Kim Y.
      • et al.
      Dust mite, cockroach, cat, and dog allergen concentrations in homes of asthmatic children in the northeastern United States: impact of socioeconomic factors and population density.
      which was used as the socioeconomic variable during home selection. We incorporated smoking in the home (n = 20) in the statistical model with fungal concentration, and this did not significantly affect the results. One sample was collected per child, the follow-up period for asthma severity was 1 month long, and longer follow-up times or periodic tests throughout the year might allow for additional insights in future studies. Atopic status was used to define allergic asthma, although the presence of eosinophils was not measured. Although not all atopic asthmatic patients have allergic asthma,
      • Pearce N.
      • Pekkanen J.
      • Beasley R.
      How much asthma is really attributable to atopy?.
      this likely results in only minor misclassification bias. Observations about associations between microbial communities and atopic status are cross-sectional. Atopic status is present on a spectrum, and subjects were classified as atopic/nonatopic based on age-adjusted criteria. This division does not result in clearly defined atopic status, such as with asthma severity, in which patients with moderate asthma were excluded. The sample size for this study limited the ability for additional analysis beyond what was completed.
      The strengths of this study include the prospective determination of asthma severity in the month-long period after environmental sampling. This study was designed to provide the opportunity to consider all asthmatic children together, as well as stratified by atopic status, while also controlling for socioeconomic status, asthma maintenance medication use, and season. Use of the Global Initiative for Asthma Guidelines to define asthma severity controls for use of asthma maintenance medication.
      • Gent J.F.
      • Kezik J.M.
      • Hill M.E.
      • Tsai E.
      • Li D.W.
      • Leaderer B.P.
      Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
      Homes were stratified by level of urbanization to control for socioeconomic status. Asthma maintenance medication use is also associated with socioeconomic status.
      • Kozyrskyj A.L.
      • Mustard C.A.
      • Simons F.R.
      Socioeconomic status, drug insurance benefits, and new prescriptions for inhaled corticosteroids in schoolchildren with asthma.
      Additionally, all homes in this study were homogeneous in geographic region, and samples were included from 1 season (May-October). Homes originated from a diverse stock of single and multifamily homes in suburban and urban locations. We measured and considered both fungal and bacterial exposures in the same study. We used molecular-based tools, including next-generation DNA sequencing and qPCR, to thoroughly determine the diversity, taxa, and concentration of both fungi and bacteria in house dust. The use of qPCR-determined concentrations allows calculation of absolute concentrations from sequencing data.
      • Dannemiller K.C.
      • Lang-Yona N.
      • Yamamoto N.
      • Rudich Y.
      • Peccia J.
      Combining real-time PCR and next-generation DNA sequencing to provide quantitative comparisons of fungal aerosol populations.

       Conclusions

      Microbial exposures are associated with childhood asthma severity, and associations might be different in atopic and nonatopic children. This supports the notion that 2 subtypes of asthma, allergic and nonallergic, are affected differently by environmental exposures. Further research is needed to determine whether asthma control measures should differ based on asthma subtype. For example, measures that reduce overall fungal concentration might be important for patients with nonallergic asthma, whereas patients with allergic asthma might achieve best results through lifestyle changes and building type and operation factors that affect fungal community composition. Both subtypes of asthma might benefit from a reduction in concentration of allergenic fungal species. In the future, the results from this study and others can be coupled with information on how housing characteristics affect indoor microbial communities to update recommendations to further reduce asthma severity in children.
      Key messages
      • Asthma severity in children is associated with microbial exposures in the home.
      • Asthma severity in atopic children is associated with fungal community composition, whereas asthma severity in nonatopic children is associated with total fungal concentration.
      We thank the families for participation in this study.

      Appendix

      References

        • Douwes J.
        • Gibson P.
        • Pekkanen J.
        • Pearce N.
        Non-eosinophilic asthma: importance and possible mechanisms.
        Thorax. 2002; 57: 643-648
        • Pearce N.
        • Pekkanen J.
        • Beasley R.
        How much asthma is really attributable to atopy?.
        Thorax. 1999; 54: 268-272
        • Kim H.Y.
        • DeKruyff R.H.
        • Umetsu D.T.
        The many paths to asthma: phenotype shaped by innate and adaptive immunity.
        Nat Immunol. 2010; 11: 577-584
        • Amin K.
        • Ludviksdottir D.
        • Janson C.
        • Nettelbladt O.
        • Bjornsson E.
        • Roomans G.M.
        • et al.
        Inflammation and structural changes in the airways of patients with atopic and nonatopic asthma.
        Am J Respir Crit Care Med. 2000; 162: 2295-2301
        • Umetsu D.T.
        • McIntire J.J.
        • Akbari O.
        • Macaubas C.
        • DeKruyff R.H.
        Asthma: an epidemic of dysregulated immunity.
        Nat Immunol. 2002; 3: 715-720
        • Levine S.J.
        • Wenzel S.E.
        Narrative review: the role of Th2 immune pathway modulation in the treatment of severe asthma and its phenotypes.
        Ann Intern Med. 2010; 152: 232-237
        • Michel O.
        • Kips J.
        • Duchateau J.
        • Vertongen F.
        • Robert L.
        • Collet H.
        • et al.
        Severity of asthma is related to endotoxin in house dust.
        Am J Respir Crit Care Med. 1996; 154: 1641-1646
        • Zambelli-Weiner A.
        • Stockton M.
        • Ehrlich E.
        • Whitmer M.
        • Lewis D.
        • Olenchock S.
        • et al.
        House dust endotoxin exposure mitigates asthma severity in a tropical environment.
        J Allergy Clin Immunol. 2004; 113: S304
        • Goldman D.L.
        • Vicencio A.G.
        The chitin connection.
        mBio. 2012; 3 (e00056-12)
        • Blatter J.
        • Forno E.
        • Brehm J.
        • Acosta-Pérez E.
        • Alvarez M.
        • Colón-Semidey A.
        • et al.
        Fungal exposure, atopy, and asthma exacerbations in Puerto Rican children.
        Ann Am Thorac Soc. 2014; 11: 925-932
        • Douwes J.
        • Zuidhof A.
        • Doekes G.
        • van der Zee S.
        • Wouters I.
        • Boezen M.H.
        • et al.
        (1 → 3)- β -d-glucan and endotoxin in house dust and peak flow variability in children.
        Am J Respir Crit Care Med. 2000; 162: 1348-1354
        • Mendell M.J.
        • Mirer A.G.
        • Cheung K.
        • Tong M.
        • Douwes J.
        Respiratory and allergic health effects of dampness, mold, and dampness-related agents: a review of the epidemiologic evidence.
        Environ Health Perspect. 2011; 119: 748-756
        • Murray C.S.
        • Poletti G.
        • Kebadze T.
        • Morris J.
        • Woodcock A.
        • Johnston S.L.
        • et al.
        Study of modifiable risk factors for asthma exacerbations: virus infection and allergen exposure increase the risk of asthma hospital admissions in children.
        Thorax. 2006; 61: 376-382
        • Gent J.F.
        • Kezik J.M.
        • Hill M.E.
        • Tsai E.
        • Li D.W.
        • Leaderer B.P.
        Household mold and dust allergens: exposure, sensitization and childhood asthma morbidity.
        Environ Res. 2012; 118: 86-93
        • Belanger K.
        • Holford T.R.
        • Gent J.F.
        • Hill M.E.
        • Kezik J.M.
        • Leaderer B.P.
        Household levels of nitrogen dioxide and pediatric asthma severity.
        Epidemiology. 2013; 24: 320-330
      1. US Department of Health and Human Services. Global Initiative for Asthma, Global Strategy for Asthma Management and Prevention. 2002. Available at: http://www.ginasthma.org/Guidelines/guidelines-2002-original%3a-workshop-report%2c-global-strategy-for-asthma-management-and-prevention.html. Accessed January 1, 2011.

      2. Mayo Clinic. Immunoglobulin E (IgE) serum. 2006. Available at: http://mayomedicallaboratories.com/test-catalog/pring.pht?unit_code=8159. Accessed April 5, 2011.

        • Yamamoto N.
        • Bibby K.
        • Qian J.
        • Hospodsky D.
        • Rismani-Yazdi H.
        • Nazaroff W.W.
        • et al.
        Particle-size distributions and seasonal diversity of allergenic and pathogenic fungi in outdoor air.
        ISME J. 2012; 6: 1801-1811
        • Adams R.I.
        • Bhangar S.
        • Pasut W.
        • Arens E.A.
        • Taylor J.W.
        • Lindow S.E.
        • et al.
        Chamber bioaerosol study: outdoor air and human occupants as sources of indoor airborne microbes.
        PLoS One. 2015; 10: e0128022
      3. National Oceanic and Atmospheric Administration, National Centers for Environmental Information. 1981-2010 U.S. Climate Normals. Hartford Bradley International Airport, CT, US. Available at: http://www.ncdc.noaa.gov/cdo-web/datatools/normals. Accessed June 30, 2015.

        • Belanger K.
        • Gent J.F.
        • Triche E.W.
        • Bracken M.B.
        • Leaderer B.P.
        Association of indoor nitrogen dioxide exposure with respiratory symptoms in children with asthma.
        Am J Respir Crit Care Med. 2006; 173: 297-303
        • Belanger K.
        • Beckett W.
        • Triche E.
        • Bracken M.B.
        • Holford T.
        • Ren P.
        • et al.
        Symptoms of wheeze and persistent cough in the first year of life: associations with indoor allergens, air contaminants, and maternal history of asthma.
        Am J Epidemiol. 2003; 158: 195-202
        • Leaderer B.P.
        • Belanger K.
        • Triche E.
        • Holford T.
        • Gold D.R.
        • Kim Y.
        • et al.
        Dust mite, cockroach, cat, and dog allergen concentrations in homes of asthmatic children in the northeastern United States: impact of socioeconomic factors and population density.
        Environ Health Perspect. 2002; 110: 419-425
        • Dannemiller K.C.
        • Gent J.F.
        • Leaderer B.P.
        • Peccia J.
        Influence of housing characteristics on bacterial and fungal communities in homes of asthmatic children.
        Indoor Air. 2015; ([Epub ahead of print])https://doi.org/10.1111/ina.12205
        • Manter D.
        • Vivanco J.
        Use of the ITS primers, ITS1F and ITS4, to characterize fungal abundance and diversity in mixed-template samples by qPCR and length heterogeneity analysis.
        J Microbiol Methods. 2007; 71: 7-14
        • Larena I.
        • Salazar O.
        • Gonzalez V.
        • Julian M.C.
        • Rubio V.
        Design of a primer for ribosomal DNA internal transcribed spacer with enhanced specificity for ascomycetes.
        J Biotechnol. 1999; 75: 187-194
        • Caporaso J.G.
        • Kuczynski J.
        • Stombaugh J.
        • Bittinger K.
        • Bushman F.D.
        • Costello E.K.
        • et al.
        QIIME allows analysis of high-throughput community sequencing data.
        Nat Methods. 2010; 7: 335-336
        • Dannemiller K.C.
        • Reeves D.
        • Bibby K.
        • Yamamoto N.
        • Peccia J.
        Fungal high-throughput taxonomic identification tool for use with next-generation sequencing (FHiTINGS).
        J Basic Microb. 2013; 54: 315-321
        • Gauch H.G.
        • Whittaker R.H.
        • Wentworth T.R.
        A comparative study of reciprocal averaging and other ordination techniques.
        J Ecol. 1977; 65: 157-174
        • Horn H.S.
        Measurement of “overlap” in comparative ecological studies.
        Am Nat. 1966; 100: 419-424
        • Caporaso J.G.
        • Lauber C.L.
        • Walters W.A.
        • Berg-Lyons D.
        • Lozupone C.A.
        • Turnbaugh P.J.
        • et al.
        Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample.
        Proc Natl Acad Sci U S A. 2011; 108: 4516-4522
        • Lozupone C.
        • Knight R.
        UniFrac: a new phylogenetic method for comparing microbial communities.
        Appl Environ Microb. 2005; 71: 8228-8235
        • Dannemiller K.C.
        • Lang-Yona N.
        • Yamamoto N.
        • Rudich Y.
        • Peccia J.
        Combining real-time PCR and next-generation DNA sequencing to provide quantitative comparisons of fungal aerosol populations.
        Atmos Environ. 2014; 84: 113-121
        • Storey J.D.
        • Tibshirani R.
        Statistical significance for genomewide studies.
        Proc Natl Acad Sci U S A. 2003; 100: 9440-9445
        • Trent M.S.
        • Stead C.M.
        • Tran A.X.
        • Hankins J.V.
        Invited review: diversity of endotoxin and its impact on pathogenesis.
        J Endotoxin Res. 2006; 12: 205-223
        • Simon-Nobbe B.
        • Denk U.
        • Pöll V.
        • Rid R.
        • Breitenbach M.
        The spectrum of fungal allergy.
        Int Arch Allergy Immunol. 2008; 145: 58-86
        • Findley K.
        • Oh J.
        • Yang J.
        • Conlan S.
        • Deming C.
        • Meyer J.A.
        • et al.
        Topographic diversity of fungal and bacterial communities in human skin.
        Nature. 2013; 498: 367-370
        • Cui L.
        • Morris A.
        • Ghedin E.
        The human mycobiome in health and disease.
        Genome Med. 2013; 5: 63
        • Liu A.H.
        Endotoxin exposure in allergy and asthma: Reconciling a paradox.
        J Allergy Clin Immunol. 2002; 109: 379-392
        • Lynch S.V.
        • Wood R.A.
        • Boushey H.
        • Bacharier L.B.
        • Bloomberg G.R.
        • Kattan M.
        • et al.
        Effects of early-life exposure to allergens and bacteria on recurrent wheeze and atopy in urban children.
        J Allergy Clin Immunol. 2014; 134: 593-601.e12
        • Fujimura K.E.
        • Demoor T.
        • Rauch M.
        • Faruqi A.A.
        • Jang S.
        • Johnson C.C.
        • et al.
        House dust exposure mediates gut microbiome Lactobacillus enrichment and airway immune defense against allergens and virus infection.
        Proc Natl Acad Sci U S A. 2014; 111: 805-810
        • Dannemiller K.
        • Mendell M.J.
        • Macher J.M.
        • Kumagai K.
        • Bradman A.
        • Holland N.
        • et al.
        Next-generation DNA sequencing reveals that low fungal diversity in house dust is associated with childhood asthma development.
        Indoor Air. 2014; 24: 236-247
        • Kozyrskyj A.L.
        • Mustard C.A.
        • Simons F.R.
        Socioeconomic status, drug insurance benefits, and new prescriptions for inhaled corticosteroids in schoolchildren with asthma.
        Arch Pediat Adol Med. 2001; 155: 1219-1224