Abstract
Aims
Wheezing in childhood is common and evidence is accumulating for the role of the gut microbiome in the development of atopic wheeze. Changes to the early-life gut microbiota and secretory IgA (SIgA) production have been linked to childhood disease; however, their connection to nonatopic wheeze is unknown. The objectives of the present study were to evaluate the relationships between early-life gut microbiota trajectories, SIgA and childhood nonatopic wheeze.
Methods
Early-life gut microbiota, SIgA and child outcome data were collected as part of the Canadian Healthy Infant Longitudinal Development (CHILD) cohort study on 1203 children. Gut microbiota trajectories were categorised as C1–C1, C1–C2, C2–C1 and C2–C2 based on low (cluster 1; C1) or high (cluster 2; C2) Bacteroides abundance in faecal samples collected at 3 and 12 months. SIgA was assessed in faecal samples at 3 months.
Results
The main outcome was nonatopic wheeze before age 5 years (n=105). Logistic regression analysis showed the C1–C2 trajectory, of low Bacteroides abundance at 3 months but higher Bacteroides abundance at 12 months, to be associated with increased adjusted odds ratio (aOR) for nonatopic wheeze (aOR 1.74, 95% CI 1.13–2.67). This was further increased if the child was not exclusively breastfed and had high SIgA level in combination with the C1–C2 trajectory (OR 4.10, 95% CI 1.15–14.59).
Conclusions
Nonatopic wheeze is associated with a depletion of Bacteroides in infancy, and in children not exclusively breastfed the risk is further increased among those with high endogenous SIgA levels. These results highlight the importance of the interplay between the gut microbiota and immune system development during critical periods in early life and how it is linked to nonatopic wheeze in childhood.
Shareable abstract
The development of childhood nonatopic wheeze is associated with the interplay between early-life gut microbiota trajectories and the maturation of the infant immune system https://bit.ly/4irJHYH
Introduction
One in three children experience at least one episode of wheeze before preschool age [1]. When wheezing persists, it is associated with asthma development and long-term respiratory morbidity [2]. Childhood wheeze is characterised as atopic or nonatopic according to the triggers and associated immune responses [3]. While the clinical phenotype is similar to atopic wheeze, asthma guidelines highlight the need for improved diagnosis of the nonatopic wheeze phenotype [4]. Persons with nonatopic wheeze are less likely to respond to inhaled corticosteroid treatment [3, 5] or benefit from long-term effects of inhaled corticosteroids against declining lung function [6]. Although the development of nonatopic wheeze in children is likely multifaceted, its aetiology is mostly unknown. Increasing our understanding of nonatopic childhood wheeze is important for both short- and long-term respiratory health.
The gut microbiome is a complex dynamic assembly composed of a myriad of microbial communities, and differences in structural diversity in early life have been linked to disease development, including respiratory ailments [7]. During the first few weeks of life, an infant's immune system is developing, and over the first year the gut microbiome will undergo changes that will be critical for the development and maturation of local gut immunity and the systemic immune system [8]. Early-life events and exposures are fundamental to the health of our gut microbes. During infancy when the gut microbial ecosystem is still developing, its composition is highly influenced by external factors. For example, exclusive breastfeeding supports the development of a healthy gut microbiota, and may even restore the dysbiosis of low Bacteroides abundance and elevated Enterobacteriaceae to Bacteroides abundance ratios, typically seen in infants born via caesarean section [9]. Four early-life gut microbiota trajectories have previously been established based on clusters of high and low abundance of Bacteroides and high Enterobacteriaceae to Bacteroides ratios in infants at 3 and 12 months [10]. The gut microbiota trajectory characterised by low Bacteroides abundance in both early and late infancy is linked to the development of allergic conditions, including food sensitisation [10]. However, the potential importance of early-life gut microbiota trajectories on the development of conditions not associated with allergic immune responses, such as nonatopic wheeze, is still unknown.
Secretory IgA (SIgA) is a common antibody that can be secreted by gut mucosa that provides nonspecific protection against bacterial and viral infections. In the first few weeks of life, infants rely on SIgA from breastmilk, but as the infant gut microbiome starts to mature, the child's immune system becomes more efficient at producing its own SIgA. While low levels of SIgA have been associated with atopic disease [11], an elevated endogenous production of SIgA in early life may drive premature maturation of infant gut microbiota caused by a lack of the immune-dampening protective effect when SIgA is supplied via breastmilk [12]. The presence of high infant levels of SIgA in the absence of breastfeeding may thus further modify the infant's immunity and potentially the association with nonatopic wheeze development. The purpose of this study was to first assess whether there is a link between early-life gut microbiota trajectories and nonatopic wheeze in childhood, and second, to evaluate the potential effects of high endogenous infant SIgA levels on this association according to breastfeeding status.
Methods
Study design
This prospective study was based on data collected as part of the Canadian Healthy Infant Longitudinal Development (CHILD) cohort study (https://www.childstudy.ca) [13]. The original cohort of pregnant females (n=3264) was recruited during the third trimester between 2009 and 2012. To be included in the study, the pregnancy had to be a singleton, with the child born at ≥35 weeks of gestation and with a birth weight of ≥2500 g. Children conceived by in vitro fertilisation were not included. Information on the parents, the pregnancy and home environment were collected during the pregnancy, and follow-up health data for the child were available until the age of 5 years. Based on exposure and outcome data availability, a subset of 1203 mother–child dyads were included in the analysis (supplementary figure S1).
This study was approved by the University of Alberta Health Research Ethics Board (Pro00103296) permitting the use of data from the CHILD cohort study sites in Alberta, British Columbia, Manitoba and Ontario. Expectant parents provided informed consent prior to enrolling in the study.
Microbiota trajectories
Faecal samples were collected according to a standardised protocol during scheduled home visits when the infant was 3 months old (early infancy) and clinic visits at 12 months (late infancy) of age. The sampling process, DNA extraction, amplification, 16S rRNA amplicon sequencing and bioinformatics have been described in detail elsewhere [10]. Briefly, 80–200 mg of a faecal sample was processed for DNA extraction using the QIAamp DNA Stool Mini kit (QIAGEN, Venlo, the Netherlands) and the V4 region of the bacterial 16S rRNA gene was amplified using the V4-515f and V4-806r universal primers. All PCR reactions were performed in triplicate. Paired-end sequencing of the pooled PCR amplicons was performed on the Illumina MiSeq platform.
As per the QIIME pipeline (v.1.9.0), forward and reverse sequencing reads were assembled into a final length of 144 bp (singletons were discarded) using PandaSeq, demultiplexed and mapped against the GREENGENES reference database (v.13.8). Sequences with <60% similarity to the reference were removed. Remaining sequences were clustered with Usearch61 at 97% sequence similarity against the GREENGENES database using a closed-reference picking process, and taxonomic assignment was done using the RDP classifier. After taxonomic assignment, operational taxonomic units (OTUs) representing microbial origin were selected, and microbial OTUs with overall relative abundance below 0.01% were excluded from downstream analyses. Data were rarefied to 13 000 sequences per sample to adjust the various sequencing depths among samples. With the recommended pipeline in QIIME, relative abundance of microbial OTUs was summarised at the phylum to genus levels. Microbial alpha diversity within samples was calculated as Chao1 richness, observed OTUs, phylogenetic diversity whole-tree index, Simpson index and Shannon index. Microbial community differences between samples (beta diversity) were measured as UniFrac distance between the genus-level profiles.
The gut microbiome trajectories were based on longitudinal shifts from early to late infancy in clusters of microbiota communities based on the partitioning around medoids clustering method, as created and fully described by Tun et al. [10]. All faecal microbiota samples were divided into two distinct clusters, C1 and C2, according to weighted UniFrac distances, the Calinski–Harabasz index and Silhouette width. Based on these clusters and the early and late infancy time points, four early-life microbiota trajectories were identified: C1–C1, C1–C2, C2–C1 and C2–C2. Trajectories were characterised by microbial species abundance using the LefSe (linear discriminant effect size) method. The most dominant microbiota trajectory, C2–C2, had the highest Bacteroides abundance and lowest Enterobacteriaceae to Bacteroides abundance ratio in early and later infancy [10]. C1–C2, the next most common trajectory was characterised by a higher abundance of Bacteroides in later but not early infancy. The reverse pattern was observed for Bacteroides abundance in later versus early infancy for C2–C1. Persistently low levels of Bacteroides were characteristic of the C1–C1 trajectory. Trajectory C1–C2 had a lower Bacteroides abundance and a higher Enterobacteriaceae to Bacteroides abundance ratio (p<0.001) in early infancy compared with C2–C2 [10].
Quantification of faecal SIgA
Levels of SIgA were analysed in a subset of 528 of infants at 3 months of age using an SIgA ELISA kit (Immundiagnostik, Bensheim, Germany). The samples were quantified as milligrams of SIgA per wet weight (grams) of faeces (mg·g−1) [14] and further grouped as the highest (Q4) versus the lowest three quartiles (Q1–3) of SIgA (supplementary figure S2).
Definition of the outcome: nonatopic wheeze versus no wheeze
The main outcome of the study was nonatopic wheeze in the child at 5 years of age. Information on childhood wheezing was collected according to a standardised questionnaire by trained study staff. For the current study, childhood wheezing was considered present if the parents answered yes to one or more questions on whether the child had been wheezing within 1 year prior to their fifth birthday.
The presence of atopy was based on an allergy skin-prick test (SPT) performed by trained staff when the child was 3 years of age. A positive SPT was indicated with a wheal ≥2 mm in diameter around the prick site for at least one of 17 common inhalant and food allergens, including Alternaria tenuis, cat hair, dog epithelium, Dermatophagoides pteronyssinus, Dermatophagoides farinae, German cockroach, peanut, soybean, egg white, milk, Penicillium, Cladosporium, Aspergillus fumigatus, trees, grasses, weeds and ragweed.
Data on the presence of wheeze and atopy were combined, and the main outcome of the study, nonatopic wheeze, was defined as wheeze in the absence of atopy. Children without wheeze, regardless of atopy status, were included in the study as controls. Children with both wheeze and atopy (n=46) were excluded from the study.
Perinatal, maternal and early-life variables
Cohort characteristics collected during pregnancy, at birth and at follow-up included biological sex of the infant at birth (male or female), mode of delivery (vaginal no intrapartum antibiotics prophylaxis (IAP), vaginal with IAP, planned caesarean section with IAP or emergency caesarean section with IAP), gestational age at birth in weeks, preterm birth (less than 37 weeks of gestation at birth), birthweight in grams, maternal ethnicity, maternal marital status at birth, maternal education level, maternal age in years, maternal history of asthma or atopy, smoking in pregnancy, vaccination status at 2 and 4 months to match the timing of SIgA measurements, respiratory infections in infancy (before 3 months of age) and antibiotics in infancy (before 1 year of age). Information on breastfeeding status was received based on self-report by the mother when the infant was 3 months of age and reported as exclusive breastfeeding, partial breastfeeding or no breastfeeding. Owing to small sample sizes, the breastfeeding variable was grouped as exclusive versus partial or no breastfeeding for parts of the analysis.
Statistical analysis
Perinatal, maternal and early-life characteristics were compared between children with and without nonatopic wheeze using a t-test for continuous variables and chi-squared test for categorical variables. The distribution of early-life gut microbiota trajectories between children with and without nonatopic wheeze was compared using the chi-squared test for a two-way measure of association. The relationship between each of the early-life gut microbiota trajectories in comparison with all other trajectories and the development of nonatopic wheeze were evaluated by calculating the odds ratios (ORs) and 95% confidence intervals (CIs) using logistic regression analysis and visualised using forest plots, while adjusting regression models for smoking in pregnancy and oral antibiotics before the age of 1 year (according to the Directed Acyclic Graph (DAG) method of selecting a minimum set of confounding factors; supplementary figure S3). As the effect modification of breastfeeding was tested, breastfeeding status was not a candidate confounding variable and not added to the DAG. Sensitivity analysis was conducted to calculate ORs for nonatopic wheeze according to each gut microbiota trajectory, excluding children with atopy from the nonwheeze control group (n=116 excluded).
Mean SIgA levels at 3 months between those who were exclusively, partially or not breastfed were evaluated using a one-way ANOVA and pairwise comparisons with Bonferroni corrections. The potential effect modification of breastfeeding status on the relationship between the C1–C2 early-life gut microbiota trajectory, SIgA levels and nonatopic wheeze was assessed. This was conducted by running stratified regression analyses according to partial/no breastfeeding versus exclusive breastfeeding status at 3 months of age, and calculating the ORs and 95% CI for a four-level trajectory/SIgA variable (gut microbiota trajectory C1–C2 versus other and SIgA as quartile 4 versus quartile 1–3). The OR and 95% CI was also calculated for the odds of nonatopic wheeze in the presence of the C1–C2 gut microbiota trajectory within strata of SIgA [15]. Missing data were not imputed. Statistical analysis was performed using Stata BE v.17 software (StataCorp, College Station, TX, USA).
Results
Cohort creation
The initial cohort was based on 3264 parent–child dyads. After only including those with complete data on microbiota trajectories over the first year, atopy status at 3 years and data on wheeze status at 5 years, and excluding those with atopic wheeze, the final cohort consisted of 1203 parents and children. Of these, 105 (8.7%) children had nonatopic wheeze within the year before turning 5 years and 1098 (91.3%) remained free from wheeze.
Characteristics of cohort
Table 1 shows the perinatal, maternal and early-life characteristics of the cohort. The distribution of sex, gestational age at birth and weight at birth were similar between those with nonatopic wheeze and those without wheeze. Mode of delivery trended towards being statistically different between groups (p=0.075), with fewer children with nonatopic wheeze being delivered vaginally without IAP (42.0%) compared with no wheeze (54.7%); planned caesarean rates were higher in children with nonatopic wheeze (13.0%) versus no wheeze (9.1%). Most mothers were of white ethnicity (78.8%). The mean age of all mothers at the time of delivery was 32.3±4.6 years. With the exception for trends of a higher prevalence of asthma/atopy, prenatal smoking, single marital status and lower education level in mothers of children with nonatopic wheeze, maternal characteristics were comparable between children with and without wheeze.
TABLE 1.
Perinatal, maternal and early-life characteristics of study infants and distribution by wheeze outcome
| All | No wheeze | Nonatopic wheeze | p-value | |
|---|---|---|---|---|
| Infants | 1203 | 1098 | 105 | |
| Infant and perinatal factors | ||||
| Sex | 0.314 | |||
| Male | 631 (52.5) | 571 (52.0) | 60 (57.1) | |
| Female | 572 (47.5) | 527 (48.0) | 45 (42.9) | |
| Mode of delivery | 0.075 | |||
| Vaginal, no IAP | 628 (53.6) | 586 (54.7) | 42 (42.0) | |
| Vaginal, with IAP | 274 (23.4) | 243 (22.7) | 31 (31.0) | |
| C-section, planned | 111 (9.5) | 98 (9.1) | 13 (13.0) | |
| C-section, emergency | 159 (13.6) | 45 (13.5) | 14 (14.0) | |
| Gestational age at birth, weeks | 39.3±1.4 | 39.3±1.4 | 39.1±1.6 | 0.249 |
| Preterm birth (<37 weeks) | 0.058 | |||
| Yes | 50 (4.3) | 42 (3.9) | 8 (7.8) | |
| No | 1133 (95.8) | 1039 (96.1) | 94 (92.2) | |
| Birth weight, g | 0.416 | |||
| <3000 | 177 (14.7) | 156 (14.2) | 21 (20.0) | |
| 3000–3500 | 463 (38.5) | 430 (39.2) | 33 (31.4) | |
| 3500–4000 | 376 (31.3) | 343 (31.2) | 33 (31.4) | |
| >4000 | 152 (12.6) | 137 (12.5) | 15 (14.3) | |
| Maternal factors | ||||
| Ethnicity | 0.474 | |||
| White | 948 (78.8) | 865 (78.8) | 83 (79.1) | |
| Asian | 139 (11.6) | 129 (11.8) | 10 (9.5) | |
| First Nation | 60 (5.0) | 56 (5.1) | 4 (3.8) | |
| Other | 52 (4.3) | 45 (4.1) | 7 (6.7) | |
| NA | 4 (0.3) | 3 (0.3) | 1 (1.0) | |
| Marital status | 0.078 | |||
| Married or common-law | 1114 (92.6) | 1024 (93.3) | 90 (85.7) | |
| Single | 61 (5.1) | 51 (4.6) | 10 (9.5) | |
| Divorced or separated | 5 (0.4) | 4 (0.4) | 1 (1.0) | |
| Maternal education | 0.092 | |||
| High school or less | 88 (7.3) | 85 (7.7) | 3 (2.9) | |
| Some postsecondary | 359 (29.8) | 318 (29.0) | 41 (39.1) | |
| University degree | 505 (42.0) | 467 (42.5) | 38 (36.2) | |
| Postgraduate degree | 220 (18.3) | 201 (18.3) | 19 (18.1) | |
| Maternal age, years | 32.3±4.6 | 32.3±4.6 | 31.9±4.5 | 0.374 |
| Smoking in pregnancy | 0.138 | |||
| Yes | 40 (3.4) | 34 (3.2) | 6 (5.9) | |
| No | 1141 (96.6) | 1.046 (96.9) | 95 (94.01) | |
| Maternal asthma and atopy | 0.179 | |||
| Yes | 770 (65.2) | 698 (64.6) | 72 (71.3) | |
| No | 411 (34.8) | 382 (35.4) | 29 (28.7) | |
| Infant/childhood factors | ||||
| Breastfeeding at 3 months | 0.586 | |||
| Exclusive | 708 (58.9) | 652 (59.4) | 56 (53.3) | |
| Partial | 311 (25.9) | 279 (25.4) | 32 (30.5) | |
| None | 181 (15.1) | 164 (14.9) | 17 (16.2) | |
| SIgA at 3 months, mg·g−1 | 7.4±7.9 | 7.3±8.1 | 8.3±6.5 | 0.407 |
| SIgA at 3 months, quartiles | 0.212 | |||
| 1st quartile | 133 (25.0) | 124 (25.7) | 9 (18.0) | |
| 2nd quartile | 133 (25.0) | 124 (25.7) | 9 (18.0) | |
| 3rd quartile | 133 (25.0) | 116 (24.1) | 17 (34.0) | |
| 4th quartile | 133 (25.0) | 118 (24.5) | 15 (30.0) | |
| Routine vaccinations at 2 months | 0.117 | |||
| Yes | 1078 (90.9) | 980 (90.5) | 98 (95.2) | |
| No | 108 (9.1) | 103 (9.5) | 5 (4.9) | |
| Any vaccinations at 2–4 months | 0.082 | |||
| Yes | 1084 (91.5) | 986 (91.0) | 98 (96.1) | |
| No | 101 (8.5) | 97 (9.0) | 4 (3.9) | |
| Antibiotics before age 1 year | 0.194 | |||
| Yes | 236 (20.67) | 211 (20.2) | 25 (25.8) | |
| No | 906 (79.3) | 834 (79.8) | 72 (74.2) | |
| Respiratory infections at <3 months | 0.549 | |||
| Yes | 65 (5.4) | 58 (5.3) | 7 (6.7) | |
| No | 1138 (94.6) | 1040 (94.7) | 98 (93.3) | |
| Atopy at age 3 years | <0.001 | |||
| Yes | 116 (9.6) | 116 (10.6) | 0 (0) | |
| No | 1087 (90.4) | 982 (89.4) | 105 (100.0) | |
Data are presented as n, n (%) or mean±sd. IAP: intrapartum antibiotics prophylaxis; C-section: caesarean section; NA: not available; SIgA: secretory IgA.
Overall, 58.9% of the children were exclusively breastfed at 3 months of age, 25.9% were partially breastfed and 15.1% were not breastfed (table 1). There were no differences in breastfeeding status between those who developed nonatopic wheeze and those who did not develop wheeze (p=0.586). There were no statistical differences in vaccination status at 2 or 4 months between the groups. While none of the children with nonatopic wheeze had allergies, 10.6% of the children without wheeze had an SPT indicating a presence of allergies at 3 years of age.
Early-life gut microbiota trajectories and childhood nonatopic wheeze
Figure 1 shows the distribution of early gut microbiota trajectories among children who did and did not develop nonatopic wheeze. While the C2–C2 gut microbiota trajectory was the most prevalent trajectory (47.1%) among those who did not develop wheeze, the most prevalent early-life gut microbiota trajectory among children with nonatopic wheeze, the focus of this study, was the C1–C2 trajectory (47.6%). As such, the C1–C2 trajectory was the trajectory of interest for the remainder of the analysis. The C1–C2 gut microbiota trajectory was also associated with high rates of caesarean section delivery, preterm births, and Caucasian maternal ethnicity, and low rates of atopy (supplementary table S1). Compared with the other early-life gut microbiota trajectories, the C1–C2 trajectory was associated with a 1.74-fold increased adjusted odds ratio (aOR) of nonatopic wheeze (95% CI 1.13–2.67) and the C2–C2 trajectory was indicative of a reduced likelihood of nonatopic wheeze (aOR 0.52, 95% CI 0.33–0.82; figure 2). The C1–C1 and C2–C1 trajectories were less prevalent in both groups (figure 1) and not associated with nonatopic wheeze (figure 2). In a subset of children, excluding controls with atopy (n=116), the relationship between the C1–C2 trajectory and nonatopic wheeze remained significant (OR 1.64, 95% CI 1.07–2.53). The OR for nonatopic wheeze with the C1–C2 trajectory compared with all other trajectories was 1.611 (95% CI 0.890–2.918) for children who were not or partially breastfed and 1.609 (95% CI 0.930–2.783) for those who were exclusively breastfed at 3 months.
FIGURE 1.
Distribution of early-life microbiota trajectories among children with and without nonatopic wheeze at age 5 years. C1: cluster 1; C2: cluster 2.
FIGURE 2.
Associations between each of the early-life gut microbiota trajectories compared with all other trajectories (reference group) and nonatopic wheeze before age 5 years. C1: cluster 1; C2: cluster 2. Associations are adjusted for smoking in pregnancy and oral antibiotics before age 1 year.
Impact of elevated SIgA and early-life gut microbiota trajectories on nonatopic wheeze development stratified by breastfeeding status
Mean SIgA levels and the distribution of SIgA quartiles at 3 months of age were similar between wheeze groups (p=0.407 and p=0.212, respectively; table 1). Mean SIgA levels were the highest among those exclusively breastfed (10.5±9.6 mg·g−1), followed by partially breastfed (6.6±6.8 mg·g−1) and not breastfed (3.5±3.2 mg·g−1, p<0.001).
The combination of having SIgA levels in the highest quartile (range 10.0–60.0 mg·g−1) at 3 months and the C1–C2 early-life gut microbiota trajectory was associated with a four-fold increased risk of nonatopic wheeze among children who were partially or not breastfed (OR 4.10, 95% CI 1.15–14.59) but not in those who were exclusively breastfed (OR 3.08, 0.72–13.18; figure 3) compared with children with lower SIgA and/or gut microbiota trajectories other than C1–C2.
FIGURE 3.
Test for effect modification: odds of nonatopic wheeze. Association between early-life gut microbiota trajectories (C1–C2 versus all other), secretory IgA (SIgA) quartiles at 3 months (Q4 versus Q1–3) and nonatopic wheeze at 5 years (nonatopic wheeze versus no wheeze), a and b) according to breastfeeding status at 3 months and c and d) within strata of SIgA. b) Increased odds of nonatopic wheeze are seen when comparing children partially or not breastfed with the C1–C2 trajectory and with high SIgA (Q4) compared with children with all other trajectories and low SIgA (Q1–3). d) However, the effect is gone when only looking at the odds among children with high SIgA (within Q4 strata), which is why SIgA is considered an effect modifier. Odds ratios are not adjusted. C1: cluster 1; C2: cluster 2.
Discussion
The results from this prospective cohort study of 1203 infants show that an early-life gut microbiota trajectory characterised by low Bacteroides abundance at 3 months after birth is associated with the development of nonatopic wheeze, despite the recovery in abundance of these gut microbiota by the age of 1 year. Furthermore, among the children with this initial depletion of Bacteroides, having elevated gut SIgA levels in the partial or complete absence of breastmilk SIgA is associated with a 4.10-fold (95% CI 1.15–14.59) increased risk of nonatopic wheeze. This study thus highlights the importance of a critical period during an infant's first year of life when changes to the gut microbiome and immune system can have long-term health implications, seen here as an increased risk of nonatopic wheeze. This project is one of the first studies conducted in humans linking elevated gut SIgA in infancy to respiratory disease development in childhood.
Although wheezing in childhood is common [1], the aetiology of the nonatopic wheeze phenotype is known to a lesser degree than its atopic wheeze counterpart. Following birth and in early childhood, infant lungs continue to undergo rapid development to adapt to the outside world, and disruptions to both the internal and external environment can shape their developmental trajectory. Insufficiencies in the composition of the gut microbiota have been linked to the development of respiratory disease [7]. The infant's gut microbiota following caesarean delivery is generally characterised by a depletion of beneficial microbiota due to the lack of mother-to-offspring vertical transmission [9] and is believed to have long-term health implications for the child [16]. In the current study, we did not see a statistical difference in caesarean section rates associated with nonatopic wheeze. However, the C1–C2 early-life gut microbiota trajectory, characterised by low abundance of Bacteroides and a high Enterobacteriaceae to Bacteroides ratio, which is typically seen following caesarean delivery [10, 17], was the most prevalent trajectory among infants with nonatopic wheeze. While the importance of specific microbes remains to be examined, the results from this study point to critical periods in infancy when compositional changes to gut microbiota may lead to increased risk of nonatopic wheeze. It is also worth mentioning that gut microbial enrichment with Enterobacteriaceae during infancy is found to be positively correlated with faecal levels of SIgA [18].
Immunoglobulin A is crucial for a well-functioning immune system in early life because it acts as a first line of defence against a myriad of pathogens. Most of what we know about IgA comes from the study of serum IgA rather than intestinal SIgA quantified in faecal samples, and of serum IgA deficiency in disease [19]. Contrary to our hypothesis, the study found infants with elevated endogenously produced SIgA were at greatest risk for nonatopic wheeze. Newborn infants start producing their own intestinal SIgA within the first week of life but rely on SIgA in breastmilk for protection and enhanced immunity [20, 21]. Evidence for the role of breastfeeding in preventing nonatopic wheeze is conflicting [22]. We observed lower but not statistically significant rates in breastfeeding exclusivity among infants who developed nonatopic wheeze. More consistently, early-life lower respiratory tract infections are strong predictors of nonatopic wheeze in schoolchildren [22] and respiratory viruses can trigger intestinal SIgA production [23, 24]. Viral respiratory infections are more common following caesarean delivery [25]. While planned caesarean birth most often preceded the C1–C2 gut microbiota trajectory in our study, there were no microbiota trajectory or wheeze status differences in respiratory infection rates at 3 months of age, when faecal SIgA levels were measured.
We found that infants who were already at risk for nonatopic wheeze due to their C1–C2 gut microbiota trajectory, had a four-fold higher likelihood of nonatopic wheeze if they had both high faecal SIgA levels and were partially or not breastfed at 3 months of age. Thus, the combination of high endogenous SIgA production in early infancy with a deficiency of breastmilk SIgA may further predispose infants to nonatopic wheeze. This theory is supported by work in murine models suggesting that endogenous SIgA production by intestinal immune cells and lack of the immune-dampening effect of maternal IgA from breastmilk can lead to gut microbial dysbiosis, lack of immune tolerance and gut inflammation in offspring [12]. Together with a small intestine bloom of segmented filamentous bacteria and Th17 cells, lack of breast milk SIgA but intact endogenous SIgA production also led to Th17 cell elevation in the lungs of murine offspring [26].
The data for this study came from a well-established birth cohort and its large sample size allowed for in-depth characterisation of gut microbiota trajectories and statistical modelling; however, there was a low number of children who possessed both the C1–C2 early-life gut microbiota trajectory and were not breastfed, resulting in an inadequate sample size for a stratified analysis. To allow for the assessment of high SIgA levels in lieu of IgA being supplied predominantly via breastmilk among children in the C1–C2 trajectory and the risk of nonatopic wheeze, children who were partially breastfed at 3 months were grouped together with the ones who were not breastfed. This increased the sample size but created a less distinct group where lack of transfer of maternal IgA would likely be minimal but not guaranteed absent. The assessment of the impact of high SIgA on nonatopic wheeze in a sample completely free from any contamination with maternal IgA might potentially have provided an even stronger association. Another limitation of the study was the relatively low rates of wheeze, which are inconsistent with reports from population-based studies [1]. It is unknown whether this discrepancy is due to sampling bias during the recruitment process, voluntary response bias among those who agreed to participate, or recall bias during the follow-up sessions, but it should be considered when generalising the results to broader populations. The cohort consists of predominately white mothers of high socioeconomic status, and generalisation of the results of the current study may be less appropriate in other populations. Finally, the characterisation of the gut microbiota trajectories may have been improved if amplicon sequence variant (ASV)-based rather than OTU-based clustering methods had been used when creating the gut microbiota clusters; however, few differences were found between OTU and ASV classification in the current cohort.
In conclusion, a gut microbiota characterised by a low abundance of Bacteroides in early infancy is associated with an increased risk of nonatopic wheeze by preschool age, despite the recovery of Bacteroides abundance by the age of 1 year. The risk of nonatopic wheeze with this microbiota trajectory is four-fold greater in formula-supplemented infants with high gut SIgA levels. As elevated SIgA levels were found to be a risk factor in the absence or reduction of maternal IgA transfer from breastmilk, our results might be indicative of dysregulated production of infants' own SIgA against a backdrop of diminished protection from milk SIgA. This study highlights a critical period in early life during which changes to the microbiota and immune system may lead to increased risk of nonatopic wheeze in childhood.
Acknowledgements
The authors would like to acknowledge all the participants in this study.
Footnotes
Provenance: Submitted article, peer reviewed.
Ethics statement: This study was approved by the University of Alberta Health Research Ethics Board (Pro00103296) permitting the use of data from the CHILD cohort study sites in Alberta, British Columbia, Manitoba and Ontario.
This article has an editorial commentary: https://doi.org/10.1183/23120541.00437-2025
Conflict of interest: L.E. Moore reports support for the present manuscript from an Innovation Grant from the Women and Children's Health Research Institute (WCHRI), Edmonton, AB, Canada. C.J. Field reports a CIHR operating grant, CRC Tier I – personal funding, a NSERC Discovery Grant, a University of Alberta WCHRI innovation grant, consultancy fees from Dairy Farmers of Canada and Byheart, and leadership roles with ILSI North American and Canada, and IFANS. P.J. Mandhane reports grants from the Canadian Institute for Health Research, the Women and Children's Health Research Institute (University of Alberta), Public Health Agency of Canada (completed) and Alberta Health (completed); support for attending meetings from the European Society of Clinical Microbiology and Infectious Disease; and a leadership role with Alberta Health Services. E. Simons reports grants from the Public Health Agency of Canada (PHAC), Canadian Institutes of Health Research (CIHR), Children's Hospital Foundation and Children's Hospital Research Institute of Manitoba (CHF and CHRIM); payment or honoraria for lectures, presentations, speakers' bureaus, manuscript writing or educational events from Child Health Research Day at CHRIM; support for attending meetings from Doctors Manitoba, Public Health Agency of Canada, and the CHILD Cohort Study; participation on a data safety monitoring board or advisory board for DIVA Study; and a leadership role with the American Academy of Allergy, Asthma, and Immunology Credentialing Committee. T.J. Moraes reports grants from CF Canada, PSI and CIHR, unrelated to the current project. P. Subbarao holds a Tier 1 Canada Research Chair in Pediatric Asthma and Lung Health, and reports participation on a data safety monitoring board for NIH Primero Observational. A. Hicks reports leadership roles with the Canadian Paediatric Society Section of Environmental Health and the Canadian Thoracic Society Asthma Steering Committee. A.L. Kozyrskyj reports support for the present manuscript from an Innovation Grant from the Women and Children's Health Research Institute (WCHRI), Edmonton, AB, Canada. H.M. Tun, S.E. Turvey and M. Hicks have nothing to disclose.
Support statement: This research was funded by a project grant from the Women and Children's Health Research Institute. Funding information for this article has been deposited with the Open Funder Registry.
Supplementary material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
00240-2025.SUPPLEMENT
References
- 1.Alvarez-Alvarez I, Niu H, Guillen-Grima F, et al. Meta-analysis of prevalence of wheezing and recurrent wheezing in infants. Allergol Immunopathol (Madr) 2018; 46: 210–217. doi: 10.1016/j.aller.2016.08.011 [DOI] [PubMed] [Google Scholar]
- 2.Lodge CJ, Lowe AJ, Allen KJ, et al. Childhood wheeze phenotypes show less than expected growth in FEV1 across adolescence. Am J Respir Crit Care Med 2014; 189: 1351–1358. doi: 10.1164/rccm.201308-1487OC [DOI] [PubMed] [Google Scholar]
- 3.Just J, Gouvis-Echraghi R, Couderc R, et al. Novel severe wheezy young children phenotypes: boys atopic multiple-trigger and girls nonatopic uncontrolled wheeze. J Allergy Clin Immunol 2012; 130: 103–110.e8. doi: 10.1016/j.jaci.2012.02.041 [DOI] [PubMed] [Google Scholar]
- 4.Gaillard EA, Kuehni CE, Turner S, et al. European Respiratory Society clinical practice guidelines for the diagnosis of asthma in children aged 5–16 years. Eur Respir J 2021; 58: 2004173. doi: 10.1183/13993003.04173-2020 [DOI] [PubMed] [Google Scholar]
- 5.Berry M, Morgan A, Shaw DE, et al. Pathological features and inhaled corticosteroid response of eosinophilic and non-eosinophilic asthma. Thorax 2007; 62: 1043–1049. doi: 10.1136/thx.2006.073429 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Marcon A, Marchetti P, Antó JM, et al. Atopy modifies the association between inhaled corticosteroid use and lung function decline in patients with asthma. J Allergy Clin Immunol Pract 2020; 8: 980–988.e10. doi: 10.1016/j.jaip.2019.10.023 [DOI] [PubMed] [Google Scholar]
- 7.Arrieta MC, Stiemsma LT, Dimitriu PA, et al. Early infancy microbial and metabolic alterations affect risk of childhood asthma. Sci Transl Med 2015; 7: 307ra152. doi: 10.1126/scitranslmed.aab2271 [DOI] [PubMed] [Google Scholar]
- 8.Zhang H, Zhang Z, Liao Y, et al. The complex link and disease between the gut microbiome and the immune system in infants. Front Cell Infect Microbiol 2022; 12: 924119. doi: 10.3389/fcimb.2022.924119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chen YY, Tun HM, Field CJ, et al. Impact of cesarean delivery and breastfeeding on secretory immunoglobulin A in the infant gut is mediated by gut microbiota and metabolites. Metabolites 2023; 13: 148. doi: 10.3390/metabo13020148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tun HM, Peng Y, Chen B, et al. Ethnicity associations with food sensitization are mediated by gut microbiota development in the first year of life. Gastroenterology 2021; 161: 94–106. doi: 10.1053/j.gastro.2021.03.016 [DOI] [PubMed] [Google Scholar]
- 11.Dzidic M, Abrahamsson TR, Artacho A, et al. Aberrant IgA responses to the gut microbiota during infancy precede asthma and allergy development. J Allergy Clin Immunol 2017; 139: 1017–1025.e14. doi: 10.1016/j.jaci.2016.06.047 [DOI] [PubMed] [Google Scholar]
- 12.Donald K, Petersen C, Turvey SE, et al. Secretory IgA: linking microbes, maternal health, and infant health through human milk. Cell Host Microbe 2022; 30: 650–659. doi: 10.1016/j.chom.2022.02.005 [DOI] [PubMed] [Google Scholar]
- 13.Subbarao P, Anand SS, Becker AB, et al. The Canadian Healthy Infant Longitudinal Development (CHILD) study: examining developmental origins of allergy and asthma. Thorax 2015; 70: 998–1000. doi: 10.1136/thoraxjnl-2015-207246 [DOI] [PubMed] [Google Scholar]
- 14.Drall KM, Tun HM, Morales-Lizcano NP, et al. Clostridioides difficile colonization is differentially associated with gut microbiome profiles by infant feeding modality at 3–4 months of age. Front Immunol 2019; 10: 2866. doi: 10.3389/fimmu.2019.02866 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Knol MJ, VanderWeele TJ. Recommendations for presenting analyses of effect modification and interaction. Int J Epidemiol 2012; 41: 514–520. doi: 10.1093/ije/dyr218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Arboleya S, Suárez M, Fernández N, et al. C-section and the neonatal gut microbiome acquisition: consequences for future health. Ann Nutr Metab 2018; 73: Suppl. 3, 17–23. doi: 10.1159/000490843 [DOI] [PubMed] [Google Scholar]
- 17.Yasmin F, Tun HM, Konya TB, et al. Cesarean section, formula feeding, and infant antibiotic exposure: separate and combined impacts on gut microbial changes in later infancy. Front Pediatr 2017; 5: 200. doi: 10.3389/fped.2017.00200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Janzon A, Goodrich JK, Koren O, et al. Interactions between the gut microbiome and mucosal immunoglobulins A, M, and G in the developing infant gut. mSystems 2019; 4: e00612-19. doi: 10.1128/msystems.00612-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yazdani R, Azizi G, Abolhassani H, et al. Selective IgA deficiency: epidemiology, pathogenesis, clinical phenotype, diagnosis, prognosis and management. Scand J Immunol 2017; 85: 3–12. doi: 10.1111/sji.12499 [DOI] [PubMed] [Google Scholar]
- 20.Koutras AK, Vigorita VJ. Fecal secretory immunoglobulin A in breast milk versus formula feeding in early infancy. J Pediatr Gastroenterol Nutr 1989; 9: 58–61. doi: 10.1002/j.1536-4801.1989.tb09821.x [DOI] [PubMed] [Google Scholar]
- 21.Maruyama K, Hida M, Kohgo T, et al. Changes in salivary and fecal secretory IgA in infants under different feeding regimens. Pediatr Int 2009; 51: 342–345. doi: 10.1111/j.1442-200X.2008.02748.x [DOI] [PubMed] [Google Scholar]
- 22.Strina A, Barreto ML, Cooper PJ, et al. Risk factors for non-atopic asthma/wheeze in children and adolescents: a systematic review. Emerg Themes Epidemiol 2014; 11: 5. doi: 10.1186/1742-7622-11-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Aoki Y, Ikeda T, Tani N, et al. Evaluation of the relationships between intestinal regional lymph nodes and immune responses in viral infections in children. Int J Mol Sci 2021; 23: 318. doi: 10.3390/ijms23010318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rodríguez A, Tjärnlund A, Ivanji J, et al. Role of IgA in the defense against respiratory infections. Vaccine 2005; 23: 2565–2572. doi: 10.1016/j.vaccine.2004.11.032 [DOI] [PubMed] [Google Scholar]
- 25.Kristensen K, Fisker N, Haerskjold A, et al. Caesarean section and hospitalization for respiratory syncytial virus infection: a population-based study. Pediatr Infect Dis J 2015; 34: 145–148. doi: 10.1097/INF.0000000000000552 [DOI] [PubMed] [Google Scholar]
- 26.Donald K, Serapio-Palacios A, Gerbec Z, et al. Secretory IgA in breast milk protects against asthma through modulation of the gut microbiota. Cell Rep 2024; 43: 114835. doi: 10.1016/j.celrep.2024.114835 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
00240-2025.SUPPLEMENT