OBJECTIVES:

Although the airway microbiota is a highly dynamic ecology, the role of longitudinal changes in airway microbiota during early childhood in asthma development is unclear. We aimed to investigate the association of longitudinal changes in early nasal microbiota with the risk of developing asthma.

METHODS:

In this prospective, population-based birth cohort study, we followed children from birth to age 7 years. The nasal microbiota was tested by using 16S ribosomal RNA gene sequencing at ages 2, 13, and 24 months. We applied an unsupervised machine learning approach to identify longitudinal nasal microbiota profiles during age 2 to 13 months (the primary exposure) and during age 2 to 24 months (the secondary exposure) and examined the association of these profiles with the risk of physician-diagnosed asthma at age 7 years.

RESULTS:

Of the analytic cohort of 704 children, 57 (8%) later developed asthma. We identified 4 distinct longitudinal nasal microbiota profiles during age 2 to 13 months. In the multivariable analysis, compared with the persistent Moraxella dominance profile during age 2 to 13 months, the persistent Moraxella sparsity profile was associated with a significantly higher risk of asthma (adjusted odds ratio, 2.74; 95% confidence interval, 1.20–6.27). Similar associations were observed between the longitudinal changes in nasal microbiota during age 2 to 24 months and risk of asthma.

CONCLUSIONS:

Children with an altered longitudinal pattern in the nasal microbiota during early childhood had a high risk of developing asthma. Our data guide the development of primary prevention strategies (eg, early identification of children at high risk and modification of microbiota) for childhood asthma. These observations present a new avenue for risk modification for asthma (eg, microbiota modification).

What’s Known on This Subject:

Airway microbiota modulates immune responses in the airways and may contribute to the risk of asthma. Although the airway microbiota is a highly dynamic ecology, the role of longitudinal changes in early airway microbiota in asthma development is unclear.

What This Study Adds:

In this birth cohort of 704 children, we identified distinct longitudinal nasal microbiota profiles in early life that were associated with differential risks of developing asthma. These observations present a new avenue for risk modification for asthma (eg, microbiota modification).

The prevalence of childhood asthma remains at historically high levels, affecting ∼9% of children in the United States1  and Western Europe.2,3  Childhood asthma is thought to result from a complex interplay between genetic and environmental factors, including early acute respiratory infections (ARIs).4,5  However, the underlying mechanisms linking these factors remain to be elucidated. These major knowledge gaps have hindered efforts to prevent childhood asthma.6  The development of primary prevention strategies for asthma requires the identification of modifiable risk factors in early childhood: a critical window of lung development.7 

Emerging evidence shows that the airway microbiota modulates local immune responses,8  contributing not only to the risk and severity of ARIs912  but also to chronic morbidity. Recent cross-sectional studies have revealed that composition of the airway microbiota differs between healthy controls and children with asthma13,14  and between different asthma phenotypes.15,16  In cohort studies in children, researchers have also suggested that the early airway microbiota may play a role in the development of chronic wheezing10,17  and asthma.18,19  However, the early airway microbiota is a highly dynamic ecology during early childhood,11,17  and data on its longitudinal changes remain sparse. Furthermore, the relations of longitudinal changes in the early airway microbiota with the risk of developing childhood asthma are unclear.

To address this knowledge gap, we conducted a population-based birth cohort study to test the hypothesis that longitudinal patterns of nasal microbiota during the first years of life are associated with a risk of developing asthma. A better understanding the role of airway microbiota and its longitudinal changes in early life (as a potentially modifiable risk factor) would inform the development of primary prevention strategies for childhood asthma.

In this prospective, population-based birth cohort study, the Steps to the Healthy Development and Well-being of Children (STEPS) Study, families of Finnish children born in the Hospital District of Southwest Finland between January 2008 and April 2010 were enrolled during pregnancy or soon after birth.20  The details of the study design, participants, and testing are described in the Supplemental Information. In brief, as part of the STEPS Study, children were enrolled in a follow-up for ARIs from birth to the age of 24 months21  and followed for the development of asthma until the age of 7.5 years.4  No selection criteria other than language (Finnish- or Swedish-speaking family) were applied to recruiting the families in the STEPS Study or in the subcohort. The Ministry of Social Affairs and Health and the Ethics Committee of the Hospital District of Southwest Finland approved the study. The parents of participating children gave their written, informed consent. The study complies with the Declaration of Helsinki. The results of the study were reported according to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting cohort studies (Supplemental Table 1).

Children were followed for ARIs from birth to age 24 months, with daily symptom diaries and study clinic visits. Data on emergency department visits, hospitalizations, and hospital outpatient visits for ARIs during age 13 to 24 months and physician-diagnosis of asthma at age 6.5 to 7.5 years were retrieved from medical records of the Hospital District of Southwest Finland.4  Asthma-medication use at age 6.5 to 7.5 years was retrieved from electronic prescription records. Patient demographics, family history, perinatal history, and environmental information were collected from the National Birth Registry and by structured questionnaires. Nasal swab specimens were collected by study personnel using flocked nylon swabs (COPAN, Brescia, Italy) at scheduled participant visits at age 2, 13, and 24 months during healthy state. Nasopharyngeal swab samples were collected after nasal swab sample collections.

The nasal swab samples were stored at −80°C after the collection. The 16S ribosomal RNA (rRNA) gene sequencing methods were adapted from those developed for the National Institutes of Health Human Microbiome Project.22,23  Bacterial genomic DNA was extracted from the nasal samples with automated MagNA Pure 96 System (Roche Diagnostics, Basel, Switzerland). The 16S ribosomal DNA V4 region was amplified by using polymerase chain reaction and sequenced on the MiSeq platform (Illumina, Inc, San Diego, CA) by using the 2 × 250 bp paired-end protocol. Sequencing reads were merged by using USEARCH v7.0.1090.24  16S rRNA gene sequences were clustered into operational taxonomic units (OTUs) at a similarity cutoff value of 97% by using UPARSE.25  OTUs were determined by mapping the centroids to the SILVA database.26  Rarefaction curves of bacterial OTUs were constructed by using sequence data for each sample to ensure coverage of the bacterial diversity present. Analyses were conducted at the genus-level by using bacterial relative abundances. Nasopharyngeal swab samples were analyzed by using a semiquantitative culture method.

The primary outcome was physician-diagnosed asthma, defined as a diagnosis of asthma in the medical records at age 6.5 to 7.5 years (age 7 years) with or without an electronic prescription of inhaled corticosteroids for asthma at the same age. The secondary outcome was ARIs during age 13 to 24 months. An ARI was defined as the presence of rhinitis or cough (with or without fever or wheezing) documented in the symptom diary by the parents or as any physician-diagnosed ARI.

Children with a qualified baseline nasal sample at age 2 months and at least 1 follow-up sample were included in the analysis. To identify mutually exclusive profiles of longitudinal changes in the nasal airway microbiota during early life, we applied an unsupervised clustering (longitudinal k-means clustering) approach27  based on the correlation distance28  to individual longitudinal trajectories based on log2-transformed relative abundances of the 100 most common genera. The 100 most common genera were chosen because they accounted for 99% of the overall abundance. Because of the importance of infancy, a critical period of airway development,29  clustering was first applied to the longitudinal change in microbiota during age 2 to 13 months (the primary exposure). Then, the clustering approach was applied to the longitudinal change during age 2 to 24 months (the secondary exposure), separately. The number of profiles was chosen on the basis of Calinski-Harabasz methods and clinical plausibility (Supplemental Fig 1).27  The relative abundances of genera in each sampling age were compared between the profiles by using analysis of variance, adjusting for multiple comparisons with the use of the Benjamini-Hochberg false discovery rate method. To examine the association of the longitudinal profiles with the risk of developing asthma (primary outcome) and incidence rate of ARIs (secondary outcome), multivariable logistic regression and negative binomial (with natural logarithm of the follow-up time as an offset) models were constructed. The profile with persistent Moraxella dominance was used as the reference group in the analysis (1) because it was the largest profile and (2) because of literature demonstrating the dominance of the nasal microbiota by the Moraxella genus.11,17  The models adjusted for 4 potential confounders (sex, household siblings, parental asthma, and child’s eczema at age 13 months) that were selected on the basis of a priori knowledge.10,11,30  In the sensitivity analysis, first, the models additionally adjusted for delivery mode and previous antibiotic exposures. Second, the sensitivity analysis was performed, excluding children with early wheezing before age 13 months. Third, to examine the robustness of the clustering approach, we performed sensitivity analyses, repeating longitudinal k-means clustering based on the 50 and 150 most common genera during age 2 to 13 months and using a different dissimilarity measure to individual longitudinal trajectories. Finally, in the exploratory analysis, the longitudinal trajectories of major and clinically important genera (Moraxella, Dolosigranulum, Streptococcus, Staphylococcus, Haemophilus, and Lactobacillus9,10,17,18,31,32 ) were individually examined. Data were analyzed by using R version 3.5.2.

Overall, the cohort was composed of 923 children (Supplemental Fig 2). Medical records and electronic prescription data were available for 910 (99%) of these children at age 7 years. Altogether 2261 nasal swab samples were collected from 907 (98%) children at age 2, 13, and 24 months. A total of 46 441 397 high-quality merged sequences were obtained by 16S rRNA gene sequencing of the nasal airway samples, of which 45 854 654 (99%) were mapped to 16S reference data. Of the samples, 2172 (96%) met the quality control requirements and had sufficient sequence depth for 16S rRNA gene sequencing (rarefaction cutoff: 2023 reads per sample). A total of 704 children had a qualified nasal sample at age 2 months and at least 1 follow-up sample and were included in the current analysis. Children in the analytical cohort and those without qualified microbiota data were similar in most baseline and clinical characteristics (Supplemental Table 2).

Across all samples, the dominant phyla were Proteobacteria (50%), Firmicutes (39%), and Actinobacteria (8%; Fig 1). Particularly, the nasal airway microbiota was dominated by the Moraxella genus at age 2, 13, and 24 months, which increased with age. The abundance of Staphylococcus and Corynebacteriaceae genera were relatively high at age 2 months and decreased with age. During age 13 to 24 months, the most prevalent genera besides Moraxella were Dolosigranulum and Streptococcus. The abundance of major genera determined with 16S rRNA gene sequencing were associated with species identified by bacterial culture of simultaneously collected nasopharyngeal samples (Supplemental Fig 3).

FIGURE 1

Longitudinal development of nasal airway microbiota in healthy infants during age 2 to 24 months. Proportions of 10 most abundant genera are shown with the other genera categorized into the 5 most abundant phylum groups. Color codes of genera are based on taxonomic annotation at the phylum level.

FIGURE 1

Longitudinal development of nasal airway microbiota in healthy infants during age 2 to 24 months. Proportions of 10 most abundant genera are shown with the other genera categorized into the 5 most abundant phylum groups. Color codes of genera are based on taxonomic annotation at the phylum level.

Close modal

Longitudinal clustering of nasal microbiota data during age 2 to 13 months identified 4 early longitudinal microbiota profiles, which were named descriptively: (A) persistent Moraxella dominance profile, (B) Streptococcus-to-Moraxella transition profile, (C) persistent Dolosigranulum and Corynebacteriaceae dominance profile, and (D) persistent Moraxella sparsity profile (Fig 2A). The persistent Moraxella dominance profile (A) was also characterized by high Dolosigranulum and low Streptococcus and Staphylococcus abundances. In contrast, the persistent Moraxella sparsity profile (D) had low Dolosigranulum and persistently high Streptococcus (all P < .001) and high Haemophilus at age 13 months (P = .03; Supplemental Table 3). Children with a persistent Moraxella dominance profile were more likely to have household siblings (P < .001; Table 1).

FIGURE 2

Longitudinal clustering and composition of nasal microbiota. A, Nasal microbiota during age 2 to 13 months. Four longitudinal nasal microbiota profiles were identified by using the longitudinal k-means clustering method: (A) persistent Moraxella dominance profile that also had high Dolosigranulum and low Streptococcus and Staphylococcus abundances, n = 313 (48%); (B) Streptococcus-to-Moraxella transition profile, n = 87 (13%); (C) persistent Dolosigranulum and Corynebacteriaceae dominance profile, n = 156 (24%); and (D) persistent Moraxella sparsity profile that also had persistently high Streptococcus and high Haemophilus at age 13 months as well as low Dolosigranulum abundances, n = 92 (14%). Proportions of 10 most abundant genera are shown with the other genera categorized into the 5 most abundant phylum groups. Color codes of genera are based on taxonomic annotation at the phylum level. B, Nasal microbiota during age 2 to 24 months. Five longitudinal nasal microbiota profiles were identified by using the longitudinal k-means clustering method: (A) profile A with persistent Moraxella dominance with high Dolosigranulum and low Streptococcus and Staphylococcus abundances, n = 280 (40%); (B) profile B with Streptococcus-to-Moraxella transition, n = 85 (12%); (C) profile C with early Dolosigranulum and Corynebacteriaceae dominances, n = 142 (20%); (D) profile D with early Moraxella sparsity with its subsequent rise as well as persistently high Streptococcus abundances, n = 101 (14%); and (E) profile E with mixed longitudinal patterns, n = 93 (13%).

FIGURE 2

Longitudinal clustering and composition of nasal microbiota. A, Nasal microbiota during age 2 to 13 months. Four longitudinal nasal microbiota profiles were identified by using the longitudinal k-means clustering method: (A) persistent Moraxella dominance profile that also had high Dolosigranulum and low Streptococcus and Staphylococcus abundances, n = 313 (48%); (B) Streptococcus-to-Moraxella transition profile, n = 87 (13%); (C) persistent Dolosigranulum and Corynebacteriaceae dominance profile, n = 156 (24%); and (D) persistent Moraxella sparsity profile that also had persistently high Streptococcus and high Haemophilus at age 13 months as well as low Dolosigranulum abundances, n = 92 (14%). Proportions of 10 most abundant genera are shown with the other genera categorized into the 5 most abundant phylum groups. Color codes of genera are based on taxonomic annotation at the phylum level. B, Nasal microbiota during age 2 to 24 months. Five longitudinal nasal microbiota profiles were identified by using the longitudinal k-means clustering method: (A) profile A with persistent Moraxella dominance with high Dolosigranulum and low Streptococcus and Staphylococcus abundances, n = 280 (40%); (B) profile B with Streptococcus-to-Moraxella transition, n = 85 (12%); (C) profile C with early Dolosigranulum and Corynebacteriaceae dominances, n = 142 (20%); (D) profile D with early Moraxella sparsity with its subsequent rise as well as persistently high Streptococcus abundances, n = 101 (14%); and (E) profile E with mixed longitudinal patterns, n = 93 (13%).

Close modal
TABLE 1

Baseline Characteristics of the Cohort Children According to the Longitudinal Nasal Microbiota Profiles During Age 2–13 Months

Longitudinal Nasal Microbiota Profile During Age 2–13 moP
A: Persistent Moraxella Dominance Profile (n = 313 [48.3%])B: Streptococcus-to-Moraxella Transition Profile (n = 87 [13.4%])C: Persistent Dolosigranulum and Corynebacteriaceae Dominance Profile (n = 156 [24.1%])D: Persistent Moraxella Sparsity Profile (n = 92 [14.2%])
Age at scheduled visits, months, median (IQR)      
 2-mo 2.5 (2.4–2.7) 2.5 (2.4–2.7) 2.5 (2.4–2.7) 2.6 (2.4–2.8) .87 
 13-mo 13.5 (13.2–13.9) 13.6 (13.1–13.9) 13.6 (13.2–13.9) 13.5 (13.1–13.9) .93 
Male sex, n (%) 167 (53.4) 44 (50.6) 82 (52.6) 45 (48.9) .88 
Household sibling, n (%) 165 (52.7) 26 (29.9) 54 (34.6) 35 (38.0) <.001 
Maternal asthma, n (%) 23 (7.3) 4 (4.6) 12 (7.7) 6 (6.5) .80 
Parental asthma, n (%) 38 (12.1) 8 (9.2) 16 (10.3) 14 (15.2) .57 
Maternal smoking during pregnancy, n (%) 12 (3.8) 5 (5.7) 7 (4.5) 7 (7.6) .44 
Birth by cesarean delivery, n (%) 30 (9.6) 13 (14.9) 22 (14.1) 15 (16.3) .21 
Prematurity (<37 wk), n (%) 12 (3.8) 3 (3.4) 7 (4.5) 5 (5.4) .89 
Low birth wt (<2500 g), n (%) 9 (2.9) 3 (3.4) 7 (4.5) 1 (1.1) .51 
Small for gestational age, n (%) 5 (1.6) 2 (2.3) 5 (3.2) 2 (2.2) .63 
Intrapartum antibiotics, n (%) 35 (11.2) 8 (9.2) 25 (16.0) 12 (13.0) .37 
Breastfed during the first 2 mo of life, n (%)a 255 (81.5) 73 (83.9) 119 (76.3) 71 (77.2) .04 
Breastfed during the first 6 mo of life, n (%)a 183 (58.5) 41 (47.1) 80 (51.3) 53 (57.6) .18 
Systemic antibiotic use before age 2-mo nasal sample, n (%) 43 (13.7) 22 (25.3) 28 (17.9) 14 (15.2) .07 
Eczema by age 13 mo, n (%) 53 (16.9) 12 (13.8) 21 (13.5) 15 (16.3) .73 
Outside home day care at age 13 mo, n (%) 71 (22.7) 21 (24.1) 33 (21.2) 19 (20.7) .89 
Parental smoking at child’s age 24 mo, n (%)b 42 (13.4) 7 (8.0) 21 (13.5) 13 (14.1) .65 
Longitudinal Nasal Microbiota Profile During Age 2–13 moP
A: Persistent Moraxella Dominance Profile (n = 313 [48.3%])B: Streptococcus-to-Moraxella Transition Profile (n = 87 [13.4%])C: Persistent Dolosigranulum and Corynebacteriaceae Dominance Profile (n = 156 [24.1%])D: Persistent Moraxella Sparsity Profile (n = 92 [14.2%])
Age at scheduled visits, months, median (IQR)      
 2-mo 2.5 (2.4–2.7) 2.5 (2.4–2.7) 2.5 (2.4–2.7) 2.6 (2.4–2.8) .87 
 13-mo 13.5 (13.2–13.9) 13.6 (13.1–13.9) 13.6 (13.2–13.9) 13.5 (13.1–13.9) .93 
Male sex, n (%) 167 (53.4) 44 (50.6) 82 (52.6) 45 (48.9) .88 
Household sibling, n (%) 165 (52.7) 26 (29.9) 54 (34.6) 35 (38.0) <.001 
Maternal asthma, n (%) 23 (7.3) 4 (4.6) 12 (7.7) 6 (6.5) .80 
Parental asthma, n (%) 38 (12.1) 8 (9.2) 16 (10.3) 14 (15.2) .57 
Maternal smoking during pregnancy, n (%) 12 (3.8) 5 (5.7) 7 (4.5) 7 (7.6) .44 
Birth by cesarean delivery, n (%) 30 (9.6) 13 (14.9) 22 (14.1) 15 (16.3) .21 
Prematurity (<37 wk), n (%) 12 (3.8) 3 (3.4) 7 (4.5) 5 (5.4) .89 
Low birth wt (<2500 g), n (%) 9 (2.9) 3 (3.4) 7 (4.5) 1 (1.1) .51 
Small for gestational age, n (%) 5 (1.6) 2 (2.3) 5 (3.2) 2 (2.2) .63 
Intrapartum antibiotics, n (%) 35 (11.2) 8 (9.2) 25 (16.0) 12 (13.0) .37 
Breastfed during the first 2 mo of life, n (%)a 255 (81.5) 73 (83.9) 119 (76.3) 71 (77.2) .04 
Breastfed during the first 6 mo of life, n (%)a 183 (58.5) 41 (47.1) 80 (51.3) 53 (57.6) .18 
Systemic antibiotic use before age 2-mo nasal sample, n (%) 43 (13.7) 22 (25.3) 28 (17.9) 14 (15.2) .07 
Eczema by age 13 mo, n (%) 53 (16.9) 12 (13.8) 21 (13.5) 15 (16.3) .73 
Outside home day care at age 13 mo, n (%) 71 (22.7) 21 (24.1) 33 (21.2) 19 (20.7) .89 
Parental smoking at child’s age 24 mo, n (%)b 42 (13.4) 7 (8.0) 21 (13.5) 13 (14.1) .65 

Percentages may not equal 100% because of missingness. Using longitudinal clustering of nasal microbiota during age 2–13 mo, we initially identified 5 distinct profiles. Of these, 1 profile included only 2 children and was excluded from the analysis, resulting in 4 distinct early longitudinal microbiota profiles. Baseline characteristics of the entire cohort are presented in Supplemental Table 2. IQR, interquartile range.

a

Data on breastfeeding at age 2 mo were available for 585 (90.3%) children.

b

Data on parental smoking at child’s age 24 mo were available for 511 (78.9%) children.

In the analysis using nasal microbiota data during age 2 to 24 months, longitudinal clustering identified 5 longitudinal microbiota profiles: (1) profile A with persistent Moraxella dominance, (2) profile B with Streptococcus-to-Moraxella transition, (3) profile C with early Dolosigranulum and Corynebacteriaceae dominances, (4) profile D with early Moraxella sparsity, with its subsequent increase, and (5) new profile E with mixed longitudinal patterns (Fig 2B; Supplemental Tables 4 and 5). When we compared different clustering approaches using microbiota data during age 2 to 13 months and microbiota data during age 2 to 24 months, most children (76%–88%) were shown to have had similar longitudinal microbiota profiles (Supplemental Fig 4).

Overall, 57 (8%) children had physician-diagnosed asthma at age 7 years, of which 93% also had a documented prescription for inhaled corticosteroids for asthma at age 7 years. α and β diversity measures did not significantly differ between children who developed asthma by age 7 years and those who did not (Supplemental Table 6). In contrast, the relative abundance of Haemophilus was significantly higher at age 13 months in children who later developed asthma than in those who did not (false discovery rate = 0.03). In the examination of early longitudinal profiles (age 2–13 months), compared with children with a persistent Moraxella dominance (A), those with a persistent Moraxella sparsity (D) profile had significantly higher risks of asthma (odds ratio [OR], 2.47; 95% confidence interval [CI], 1.14–5.35). In the multivariable model, the association remained significant (OR, 2.74; 95% CI, 1.20–6.27; Fig 3; Supplemental Table 7). Likewise, with the use of age 2 to 24 month longitudinal profiles, compared with the profile A with persistent Moraxella dominance, the profile D with early Moraxella sparsity was associated with significantly higher risks of asthma (OR, 2.75; 95% CI, 1.25–6.05; Fig 3; Supplemental Table 8). In several sensitivity analyses, the results were similar (Supplemental Tables 9 and 10), and, when examining the robustness of the clustering approach, different methods yielded rather similar clusters that were associated with the risk of asthma (Supplemental Tables 11 and 12).

FIGURE 3

Multivariable-adjusted associations of longitudinal nasal microbiota profiles during age 2 to 13 months and age 2 to 24 months with the risk of asthma. To examine the association of longitudinal microbiota profiles and risk of asthma, multivariable logistic regression models with adjustment for potential confounders (sex, household siblings, parental asthma, and child’s eczema by age 13 months) were fit. Full results of the analysis are shown in Supplemental Tables 6 and 7.

FIGURE 3

Multivariable-adjusted associations of longitudinal nasal microbiota profiles during age 2 to 13 months and age 2 to 24 months with the risk of asthma. To examine the association of longitudinal microbiota profiles and risk of asthma, multivariable logistic regression models with adjustment for potential confounders (sex, household siblings, parental asthma, and child’s eczema by age 13 months) were fit. Full results of the analysis are shown in Supplemental Tables 6 and 7.

Close modal

A total of 3182 episodes of ARI were documented during age 13 to 24 months, with an incidence rate of 7.2 (95% CI, 6.8–7.5) per child-year. Compared with the persistent Moraxella dominance (A) profile, the persistent Moraxella sparsity (D) profile had significantly higher incidence rates of ARIs (incidence rate ratio, 1.16; 95% CI, 1.01–1.32; Fig 4; Supplemental Table 13).

FIGURE 4

Multivariable-adjusted association of longitudinal nasal microbiota profiles during age 2 to 13 months with the rate of ARIs during age 13 to 24 months. To examine the association of longitudinal microbiota profiles with the incident rate of ARIs, a multivariable negative binomial regression model with adjustment for potential confounders (sex, household siblings, parental asthma, and child’s eczema by age 13 months) was fit. Full results of the analysis are shown in Supplemental Table 13.

FIGURE 4

Multivariable-adjusted association of longitudinal nasal microbiota profiles during age 2 to 13 months with the rate of ARIs during age 13 to 24 months. To examine the association of longitudinal microbiota profiles with the incident rate of ARIs, a multivariable negative binomial regression model with adjustment for potential confounders (sex, household siblings, parental asthma, and child’s eczema by age 13 months) was fit. Full results of the analysis are shown in Supplemental Table 13.

Close modal

In genus-level analysis, longitudinal trajectory with increasing Haemophilus over age 2 to 13 months was associated with higher risks of asthma (OR, 2.03; 95% CI, 1.02–4.04). In addition, longitudinal trajectory with high Lactobacillus at age 2 months was associated with lower risks of asthma (OR, 0.19; 95% CI, 0.05–0.75). In contrast, there was no significant association of longitudinal trajectories of Moraxella, Dolosigranulum, Streptococcus, or Staphylococcus during age 2 to 13 months with risks of asthma (all P ≥ .20; Supplemental Fig 5; Supplemental Table 14).

In this prospective, population-based birth cohort study of 704 children with longitudinal nasal airway microbiota testing, we identified 4 distinct longitudinal microbiota profiles in early life. Specifically, compared with children with a persistent Moraxella dominance profile, those with a persistent Moraxella sparsity profile during age 2 to 13 months had a significantly higher risk of asthma. This high-risk profile was also characterized by low Dolosigranulum, high Streptococcus, and increasing Haemophilus. Additional investigation of longitudinal microbiota profiles during age 2 to 24 months demonstrated similar association with the risk of asthma. Our results indicate that, in children at high risk for developing asthma, altered longitudinal patterns in the nasal microbiome (considered together as an ecology) were already present during the first years of life, which is a critical period for lung development and potential window for primary prevention of childhood asthma. To the best of our knowledge, this is the first study that has demonstrated the association of longitudinal changes in early airway microbiota with the development of childhood asthma.

Previous studies have reported seemingly inconsistent relationships between the airway microbiota (particularly Moraxella) and risks of ARI and asthma in children. For example, using a culture-dependent approach, in a cohort study of 260 Danish children born to mothers with asthma, researchers reported that hypopharyngeal colonization with Moraxella catarrhalis, S pneumoniae, or Haemophilus influenzae at age 1 month was associated with a higher risk of asthma at age 5 years.18  Additionally, by applying 16S rRNA gene sequencing methods to 244 Australian infants with a high risk of atopy, frequent nasopharyngeal colonization with Moraxella, Streptococcus, or Haemophilus genera during age 0 to 2 years was associated with a higher risk of chronic wheeze (ie, not asthma) at age 5 years in children with early allergic sensitization.17  However, in a subset of 160 infants from the same cohort, the Moraxella-dominant nasopharyngeal profile before age 9 weeks was not associated with the risk of chronic wheeze.10  Likewise, in earlier studies, researchers have linked airway Moraxella with lower9,12,33,34  and higher10,11,21,35  risk and severity of ARIs and asthma exacerbations. The current study, with longitudinal microbiota testing, demonstrated that children with an early Moraxella sparsity profile during age 2 to 13 months (the profile at the highest risk for asthma and ARIs), subsequently, had an increase in Moraxella during age 13 to 24 months, indicating not only the complexity of longitudinal changes in airway microbiota but also the importance of early (≤13 months) microbiota, with regard to the risk of asthma. Additionally, our data demonstrated that this high-risk profile also was characterized by persistently high Streptococcus, which is in line with earlier reports in which researchers link it to ARIs,9,12  chronic wheeze,10  asthma exacerbations,36  and lower pulmonary function in children with asthma.13  However, the longitudinal change in each of these major genera (Moraxella and Streptococcus) individually was not associated with the risk of asthma. Although there are many possible explanations for the apparent inconsistencies across studies (eg, differences in cross-sectional versus longitudinal study designs, study populations, sample size, and approach to microbial testing), we believe our findings underscore the concept of the microbiome as a highly-functional and evolving ecology, and highlight the importance of studying longitudinal changes in microbiome patterns for understanding the impact on lung health. The validity of our results is buttressed by longitudinal testing of nasal microbiota using 16S rRNA gene sequencing in a population-based birth cohort. Our data build on previous studies10,1719,34  and extend them by demonstrating the association of longitudinal changes in early nasal microbiota with risk of ARIs and childhood asthma.

There are several potential mechanisms for the observed association between the airway microbiota and risk of asthma. First, it is plausible that the airway microbiota alters immune responses in the airway, increasing susceptibility to the development of asthma, through complex interactions between bacteria and the host. For example, airway colonization with H influenzae (bacteria also abundant in the persistent Moraxella sparsity profile) in healthy infants was reported to induce mixed T helper cell (Th) type 1, 2, and 17 inflammatory responses.37  Consistently, we observed that increasing Haemophilus was also individually associated with high risks of asthma, and, in previous studies, researchers have linked it to lower respiratory infections,9,12  wheeze17,18  and asthma in children.18,38  Conversely, another potential mechanism is depletion of specific bacteria in the airway that protect against respiratory morbidities (resilience microbiota39  [such as Moraxella],9,12,33,34 Dolosigranulum,11,12,33,34  and Lactobacillus31,32 ). Indeed, a previous study revealed that M catarrhalis is associated with augmented type Th1 inflammation,40  thereby, potentially, shifting the immune response toward a Th1-dominance state. In addition, a recent mouse model study revealed that intranasal administration of Lactobacillus rhamnosus GG decreased bronchoalveolar lavage eosinophils, interleukin 5 and interleukin 13 levels, and airway hyperreactivity.32  Alternatively, specific airway microbiota profiles in early childhood might simply be a marker of children who are more prone to develop asthma. Furthermore, these potential mechanisms are not mutually exclusive. Notwithstanding the complexity, identification of the early longitudinal nasal microbiota profiles associated with an increased risk of childhood asthma is an important finding. Our observations should advance research not only into the complex interrelations of the airway microbiome, host immune responses, and asthma pathogenesis but also into the development of primary prevention strategies for childhood asthma.

Our study has several potential limitations. First, although asthma is a disease of lower airways, we examined nasal airway microbiota. However, lower airway sampling would be invasive and ethically unacceptable in young infants. Additionally, previous studies have revealed a reliable correlation between upper and lower airway microbiota.12,41  Second, the STEPS cohort did not collect nasal swabs shortly after birth. Regardless, the current study that used the age 2-, 13-, and 24-month samples successfully revealed longitudinal microbiota profiles and their association with the risk of developing childhood asthma. Third, because we used nasal samples of healthy infants with a potentially low quantity of bacterial DNA, part of the samples was excluded from analysis because of insufficient sequence depth. However, the analytical cohort did not differ from the nonanalytical cohort, arguing against a substantial selection bias. Fourth, because we used 16S rRNA gene sequencing methods, identification at the species-level was not attainable in the current study. Yet, using culture methods in a subset of children, we showed associations between 16S rRNA gene sequencing and culture results (eg, children with high Moraxella genus DNA were colonized with culturable M catarrhalis). Fifth, we were limited in the confirmation of the exact mechanisms for the interrelations between the succession of the airway microbial ecosystem, host response, and development of asthma. To address this important question, the application of multiomics approaches (eg, integration with metabolomics) is a next focus for our study group. Sixth, one may surmise that the use of the persistent Moraxella dominance profile as the reference group is arbitrary. Lastly, although this is a large population-based study of healthy children, our inferences may not be generalizable to other populations and settings. However, the composition of the airway microbiota in these Finnish children was similar to that of Dutch children,11  and the prevalence of asthma (8%) is also comparable to that in the United States and other Western countries.13 

In this population-based birth cohort study of 704 children, we identified distinct early longitudinal nasal microbiota profiles that were associated with differential risks of developing childhood asthma. Specifically, compared with children with a persistent Moraxella dominance profile, those with a persistent Moraxella sparsity profile during age 2 to 13 months had a significantly higher risk of developing asthma. Our results highlight that children have distinct and dynamic longitudinal nasal microbiota profiles during the first years of life and suggest that not only certain bacteria but the microbiome as an ecology plays an important role in the development of childhood asthma. These data guide the development of primary prevention strategies (eg, early identification of children at high risk and modification of microbiota32 ) for childhood asthma.

We thank all the families who participated in this study, the midwives for their help in recruiting the families, and the whole STEPS Study team for assistance with data collection; Hanna Lagström, PhD, for her many contributions to the STEPS Study; Tamara Teros-Jaakkola, MD, for her assistance with the data collection; and Anne Kaljonen, MS, for her assistance with data handling. We thank the Finnish Functional Genomics Centre, supported by the University of Turku, Åbo Akademi University, and Biocenter Finland, for DNA extractions and Juho Vuononvirta, PhD (University of Turku, Finland), for the bacterial cultures. We also thank Nadim A. Ajami, PhD (Alkek Center for Metagenomics and Microbiome Research, Department of Molecular Virology and Microbiology at Baylor College of Medicine) for 16S rRNA gene sequencing analysis.

Dr Toivonen collected the data, conducted the statistical analysis, and drafted the initial manuscript; Drs Karppinen and Schuez-Havupalo collected the data and critically reviewed and revised the manuscript; Dr Waris contributed to the conception and design of the study, conducted the DNA extractions, and critically reviewed and revised the manuscript; Dr He contributed to the conception and design of the study, conducted the bacterial cultures, and critically reviewed and revised the manuscript; Drs Hoffman and Petrosino generated the microbiome data, conducted the initial statistical analysis, and critically reviewed and revised the manuscript; Dr Dumas assisted with the statistical analysis and critically reviewed and revised the initial manuscript; Dr Camargo contributed to the design of the study and analysis of the data and critically reviewed and revised the initial manuscript; Dr Hasegawa designed and conducted the statistical analysis and critically reviewed and revised the initial manuscript; Dr Peltola conceptualized and designed the study, supervised the conduct of the study, contributed to the analysis of the data, and critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: Supported by the University of Turku; the Åbo Akademi University; the Turku University Hospital; the Academy of Finland (grants 123571, 140251, 277535, and 324926); the Finnish Medical Foundation; the Päivikki and Sakari Sohlberg Foundation; the Foundation for Pediatric Research; the Emil Aaltonen Foundation; the Paulo Foundation; the Allergy Research Foundation; the Juho Vainio Foundation; the Tampere Tuberculosis Foundation; the Finnish Anti-Tuberculosis Association; the Orion Research Foundation; the Maud Kuistila Memorial Foundation; Research Funds from Specified Government Transfers, Hospital District of Southwest Finland; and the Väinö and Laina Kivi Foundation. The funding sources had no involvement in the study design, collection, analysis, interpretation of data, writing of the report, or the decision to submit the article for publication.

ARI

acute respiratory infection

CI

confidence interval

OR

odds ratio

OTU

operational taxonomic unit

rRNA

ribosomal RNA

STEPS

Steps to the Healthy Development and Well-being of Children

Th

T helper cell

1
Moorman
JE
,
Akinbami
LJ
,
Bailey
CM
, et al
.
National surveillance of asthma: United States, 2001-2010
.
Vital Health Stat 3
.
2012
;(
35
):
1
58
2
Lai
CKW
,
Beasley
R
,
Crane
J
,
Foliaki
S
,
Shah
J
,
Weiland
S
;
International Study of Asthma and Allergies in Childhood Phase Three Study Group
.
Global variation in the prevalence and severity of asthma symptoms: phase three of the International Study of Asthma and Allergies in Childhood (ISAAC)
.
Thorax
.
2009
;
64
(
6
):
476
483
3
Stokholm
J
,
Blaser
MJ
,
Thorsen
J
, et al
.
Maturation of the gut microbiome and risk of asthma in childhood
.
Nat Commun
.
2018
;
9
(
1
):
141
4
Toivonen
L
,
Forsström
V
,
Waris
M
,
Peltola
V
.
Acute respiratory infections in early childhood and risk of asthma at age 7 years
.
J Allergy Clin Immunol
.
2019
;
143
(
1
):
407
410.e6
5
van Meel
ER
,
den Dekker
HT
,
Elbert
NJ
, et al
.
A population-based prospective cohort study examining the influence of early-life respiratory tract infections on school-age lung function and asthma
.
Thorax
.
2018
;
73
(
2
):
167
173
6
Hasegawa
K
,
Mansbach
JM
,
Camargo
CA
 Jr.
.
Infectious pathogens and bronchiolitis outcomes
.
Expert Rev Anti Infect Ther
.
2014
;
12
(
7
):
817
828
7
Levy
BD
,
Noel
PJ
,
Freemer
MM
, et al
.
Future research directions in asthma. An NHLBI working group report
.
Am J Respir Crit Care Med
.
2015
;
192
(
11
):
1366
1372
8
Wypych
TP
,
Wickramasinghe
LC
,
Marsland
BJ
.
The influence of the microbiome on respiratory health
.
Nat Immunol
.
2019
;
20
(
10
):
1279
1290
9
Hasegawa
K
,
Mansbach
JM
,
Ajami
NJ
, et al;
the MARC-35 Investigators
.
Association of nasopharyngeal microbiota profiles with bronchiolitis severity in infants hospitalised for bronchiolitis
.
Eur Respir J
.
2016
;
48
(
5
):
1329
1339
10
Teo
SM
,
Mok
D
,
Pham
K
, et al
.
The infant nasopharyngeal microbiome impacts severity of lower respiratory infection and risk of asthma development
.
Cell Host Microbe
.
2015
;
17
(
5
):
704
715
11
Bosch
AATM
,
de Steenhuijsen Piters
WAA
,
van Houten
MA
, et al
.
Maturation of the infant respiratory microbiota, environmental drivers, and health consequences. A prospective cohort study
.
Am J Respir Crit Care Med
.
2017
;
196
(
12
):
1582
1590
12
Man
WH
,
van Houten
MA
,
Mérelle
ME
, et al
.
Bacterial and viral respiratory tract microbiota and host characteristics in children with lower respiratory tract infections: a matched case-control study
.
Lancet Respir Med
.
2019
;
7
(
5
):
417
426
13
Kim
B-S
,
Lee
E
,
Lee
M-J
, et al
.
Different functional genes of upper airway microbiome associated with natural course of childhood asthma
.
Allergy
.
2018
;
73
(
3
):
644
652
14
Depner
M
,
Ege
MJ
,
Cox
MJ
, et al
.
Bacterial microbiota of the upper respiratory tract and childhood asthma
.
J Allergy Clin Immunol
.
2017
;
139
(
3
):
826
834.e13
15
Pérez-Losada
M
,
Authelet
KJ
,
Hoptay
CE
,
Kwak
C
,
Crandall
KA
,
Freishtat
RJ
.
Pediatric asthma comprises different phenotypic clusters with unique nasal microbiotas
.
Microbiome
.
2018
;
6
(
1
):
179
16
Taylor
SL
,
Leong
LEX
,
Choo
JM
, et al
.
Inflammatory phenotypes in patients with severe asthma are associated with distinct airway microbiology
.
J Allergy Clin Immunol
.
2018
;
141
(
1
):
94
103.e15
17
Teo
SM
,
Tang
HHF
,
Mok
D
, et al
.
Airway microbiota dynamics uncover a critical window for interplay of pathogenic bacteria and allergy in childhood respiratory disease
.
Cell Host Microbe
.
2018
;
24
(
3
):
341
352.e5
18
Bisgaard
H
,
Hermansen
MN
,
Buchvald
F
, et al
.
Childhood asthma after bacterial colonization of the airway in neonates
.
N Engl J Med
.
2007
;
357
(
15
):
1487
1495
19
Thorsen
J
,
Rasmussen
MA
,
Waage
J
, et al
.
Infant airway microbiota and topical immune perturbations in the origins of childhood asthma
.
Nat Commun
.
2019
;
10
(
1
):
5001
20
Lagström
H
,
Rautava
P
,
Kaljonen
A
, et al
.
Cohort profile: steps to the healthy development and well-being of children (the STEPS Study)
.
Int J Epidemiol
.
2013
;
42
(
5
):
1273
1284
21
Toivonen
L
,
Hasegawa
K
,
Waris
M
, et al
.
Early nasal microbiota and acute respiratory infections during the first years of life
.
Thorax
.
2019
;
74
(
6
):
592
599
22
Human Microbiome Project Consortium
.
A framework for human microbiome research
.
Nature
.
2012
;
486
(
7402
):
215
221
23
Human Microbiome Project Consortium
.
Structure, function and diversity of the healthy human microbiome
.
Nature
.
2012
;
486
(
7402
):
207
214
24
Edgar
RC
.
Search and clustering orders of magnitude faster than BLAST
.
Bioinformatics
.
2010
;
26
(
19
):
2460
2461
25
Edgar
RC
.
UPARSE: highly accurate OTU sequences from microbial amplicon reads
.
Nat Methods
.
2013
;
10
(
10
):
996
998
26
Quast
C
,
Pruesse
E
,
Yilmaz
P
, et al
.
The SILVA ribosomal RNA gene database project: improved data processing and web-based tools
.
Nucleic Acids Res
.
2013
;
41
(
database issue
):
D590
D596
27
Genolini
C
,
Alacoque
X
,
Sentenac
M
,
Arnaud
C
.
kml and kml3d: R packages to cluster longitudinal data
.
J Stat Softw
.
2015
;
65
(
4
):
1
34
28
James
G
,
Witten
D
,
Hastie
T
,
Tibshirani
R
.
An Introduction to Statistical Learning: With Applications in R
.
New York, NY
:
Springer
;
2017
29
Man
WH
,
de Steenhuijsen Piters
WAA
,
Bogaert
D
.
The microbiota of the respiratory tract: gatekeeper to respiratory health
.
Nat Rev Microbiol
.
2017
;
15
(
5
):
259
270
30
Papi
A
,
Brightling
C
,
Pedersen
SE
,
Reddel
HK
.
Asthma
.
Lancet
.
2018
;
391
(
10122
):
783
800
31
Rosas-Salazar
C
,
Shilts
MH
,
Tovchigrechko
A
, et al
.
Nasopharyngeal Lactobacillus is associated with a reduced risk of childhood wheezing illnesses following acute respiratory syncytial virus infection in infancy
.
J Allergy Clin Immunol
.
2018
;
142
(
5
):
1447
1456.e9
32
Spacova
I
,
Petrova
MI
,
Fremau
A
, et al
.
Intranasal administration of probiotic Lactobacillus rhamnosus GG prevents birch pollen-induced allergic asthma in a murine model
.
Allergy
.
2019
;
74
(
1
):
100
110
33
Hasegawa
K
,
Linnemann
RW
,
Mansbach
JM
, et al
.
Nasal airway microbiota profile and severe bronchiolitis in infants: a case-control study
.
Pediatr Infect Dis J
.
2017
;
36
(
11
):
1044
1051
34
Biesbroek
G
,
Tsivtsivadze
E
,
Sanders
EAM
, et al
.
Early respiratory microbiota composition determines bacterial succession patterns and respiratory health in children
.
Am J Respir Crit Care Med
.
2014
;
190
(
11
):
1283
1292
35
McCauley
K
,
Durack
J
,
Valladares
R
, et al;
National Institute of Allergy and Infectious Diseases–sponsored Inner-City Asthma Consortium
.
Distinct nasal airway bacterial microbiotas differentially relate to exacerbation in pediatric patients with asthma
.
J Allergy Clin Immunol
.
2019
;
144
(
5
):
1187
1197
36
Kloepfer
KM
,
Lee
WM
,
Pappas
TE
, et al
.
Detection of pathogenic bacteria during rhinovirus infection is associated with increased respiratory symptoms and asthma exacerbations
.
J Allergy Clin Immunol
.
2014
;
133
(
5
):
1301
1307.e3
37
Følsgaard
NV
,
Schjørring
S
,
Chawes
BL
, et al
.
Pathogenic bacteria colonizing the airways in asymptomatic neonates stimulates topical inflammatory mediator release
.
Am J Respir Crit Care Med
.
2013
;
187
(
6
):
589
595
38
Hilty
M
,
Burke
C
,
Pedro
H
, et al
.
Disordered microbial communities in asthmatic airways
.
PLoS One
.
2010
;
5
(
1
):
e8578
39
Hasegawa
K
,
Camargo
CA
 Jr.
.
Airway microbiota and acute respiratory infection in children
.
Expert Rev Clin Immunol
.
2015
;
11
(
7
):
789
792
40
Amedei
A
,
Della Bella
C
,
Niccolai
E
, et al
.
Moraxella catarrhalis-specific Th1 cells in BAL fluids of chronic obstructive pulmonary disease patients
.
Int J Immunopathol Pharmacol
.
2009
;
22
(
4
):
979
990
41
Marsh
RL
,
Kaestli
M
,
Chang
AB
, et al
.
The microbiota in bronchoalveolar lavage from young children with chronic lung disease includes taxa present in both the oropharynx and nasopharynx
.
Microbiome
.
2016
;
4
(
1
):
37

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data