Association of rhinovirus and potentially pathogenic bacterial detections in the first 3 months of life with subsequent wheezing in childhood

Abstract Objective Airway interactions between viruses, especially rhinoviruses, and potentially pathogenic bacteria (PPB; Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis) in early infancy may increase the risk of subsequent wheezing and asthma. We evaluated the association between rhinovirus and PPB in the first 3 months of life and wheezing episodes before age 2 years and asthma at age 5–7 years. Methods An Australian community‐based birth cohort of healthy children involved parents collecting nasal swabs weekly and completing symptom diaries daily until age 2 years. In a follow‐up subset, asthma diagnosis was assessed annually until age 7 years. Swabs were analyzed by real‐time polymerase chain reaction assays. Children were included if they returned symptom diaries beyond age 3 months (wheeze) or were reviewed at age 5–7 years (asthma). Results 1440 swabs were returned by 146 children in the first 3 months of life. Wheeze and asthma outcomes were recorded for 146 and 84 children, respectively. Each additional week of rhinovirus detection increased the incidence of wheezing before age 2 years by 1.16 times (95% confidence interval [CI]: 0.99–1.35). There were no significant associations between bacteria and wheeze. Each additional week with H. influenzae increased the odds of asthma at age 5–7 years by 135% (odds ratio: 2.35, 95% CI: 0.99–5.58). No significant interaction was observed between rhinovirus and PPB for wheezing or asthma. Conclusion Early life rhinovirus infection was associated with wheezing before age 2 years and H. influenzae with asthma by age 5–7 years. Microbes may play an etiologic role in wheezing and asthma, warranting further study.

Acute lower respiratory infections (ALRIs) are the leading cause of acute wheezing episodes in young children.It is possible that ALRIs in early life may injure the developing lungs, adversely affecting airway structure and function, increasing the risk of recurrent wheezing episodes and a later diagnosis of asthma.Cohort studies have reported an association between ALRI during infancy and impaired lung function in later childhood. 1,2A population-based, data linkage study from Western Australia found a positive dose-response association between the number and duration of hospitalizations for ALRIs in the first 12 months of life and subsequent respiratoryrelated hospitalizations, including for asthma, after 3 years of age. 3 However, it is unclear whether this association is causal.In particular, it is unknown whether ALRI is an independent risk factor for later wheeze or a marker of pre-existing structural and immunological abnormalities leading to later wheeze.
8][9] There are three rhinovirus species: A, B, and C, with type C associated with more severe ALRI and asthma. 10,11In a birth cohort study conducted in the United States, rhinovirus was detected in 48% of wheezing illnesses among children up to 3 years of age.Those who had rhinovirus detected when they wheezed had a higher risk of developing asthma at age 6 years compared to children who did not wheeze when either rhinovirus or respiratory syncytial virus were detected (odds ratio [OR]: 9.8, 95% confidence interval [CI]: 4.3-22.0). 6An Australian birth cohort study reported that children with rhinovirus-related wheezing ALRI episodes in the first year of life had increased odds of persistent wheezing (OR: 2.9; 95% CI: 1.2-7.0)and asthma (OR: 2.9; 95% CI: 1.2-7.1)at age 5 years. 12lonizing upper airway bacteria may also result in aberrant programming of the host immune system and a predisposition to recurrent wheezing and asthma. 13Previous studies have reported colonization with potentially pathogenic bacteria (PPB), including bacteria most commonly detected in the upper respiratory tract of Australian infants (Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis), 14 is associated with an increased risk of ALRI 15,16 and wheeze in young children, 17,18 and asthma in later childhood. 190][21] A Finnish birth cohort study observed that early maturation of the nasopharyngeal microbiota with Moraxella genera and accompanying instability following incursions by H. influenzae and S. pneumoniae, increased the risk of developing asthma when compared with infants possessing a persistent Moraxella dominant nasopharyngeal profile. 21e present study aimed to use data from an Australian community-based, healthy birth cohort to evaluate the association between rhinoviruses and PPB (S. pneumoniae, M. catarrhalis, and H. influenzae) in the upper airway during the first 3 months of life and (i) wheeze in the first 2 years of life; and (ii) asthma at age 5-7 years.
We also evaluated whether there was evidence of an interaction between rhinoviruses and PPB.

| Recording of sociodemographic characteristics and illness episodes
At enrollment, parents provided sociodemographic and health characteristics, including pregnancy and birth details. 24 They received a digital thermometer and diary cards to complete daily, listing predefined respiratory symptoms and diagnoses in tick-box format.
Parents were taught to recognize respiratory symptoms, including wheezing and shortness of breath.When symptoms occurred, parents recorded healthcare visits in a separate illness-burden diary.Both diaries were returned to the research team monthly by mail.
During the ORChID study, parents were interviewed by telephone every 3 months to update information on feeding practices and childcare arrangements.Exclusive breastfeeding occurs when the TAKASHIMA ET AL.
| 3429 child was breastfed without taking milk formula or solids. 22Childcare was categorized as formal (regulated care outside the child's home) and informal (nonregulated care by family or friends).Vaccination data were captured from the Australian Immunization Register.

| Respiratory specimen collection and testing
In the ORChID study, bilateral anterior nasal swabs were collected at birth and thereafter weekly by parents using a single swab regardless of symptoms.All swabs were collected using a plasticshaft, rayon-budded swab and inserted into a transport tube with a foam pad reservoir soaked with viral transport medium (Virocult MW950, Medical Wire & Equipment).The median interval between swabs was 7 days (interquartile range [IQR]: 7-12).
Collected swabs were surface-mailed to the study laboratory (received median [IQR] 3 days [2-3] after collection), where they were stored at −80°C.Swabs were batch-tested for 17 respiratory viruses, as well as S. pneumoniae, M. catarrhalis, and H. influenzae by previously validated real-time polymerase chain reaction (PCR) assays. 9,14All virus and bacterial detections with PCR cycle threshold (Ct) values < 40 were considered positive.Specimen quality was assessed by testing for a marker of human genomic DNA, endogenous retrovirus-3 (ERV-3).As rhinovirus was by far the predominant virus detected in the first 3 months of life, 9 analyses of viruses in this study were restricted to rhinovirus detections.Rhinovirus genotyping was achieved by amplifying variable region VP4/VP2 genes.The PCR products were purified using the QIAquick PCR purification kit (Qiagen) and were then submitted for DNA sequencing to the Australian Genome Research Facility (The University of Queensland).Phylogenetic analysis was performed on a 230-bp section of the sequenced VP4/VP2 region (see Supporting Information: Methods).

| Wheeze and asthma
The primary outcomes were wheezing episodes reported between 3 and 24 months of life, and asthma at age 5-7 years.The presence of wheeze was extracted from the parent-completed daily symptom and illness-burden diaries.Children who supplied diary data beyond the age of 3 months were included in this analysis.At ages 5-7 years, parents were asked whether their child had ever received a diagnosis of asthma from a doctor, or if they had used an inhaled beta-2 agonist or inhaled corticosteroid asthma medication in the previous 12 months. 25

| Analysis
The association between rhinovirus and PPB detections in the first 3 months of life and wheeze in the first 2 years of life was investigated using Poisson regression models offset by the natural logarithm of the number of diary days returned.The association between rhinovirus and PPB detections and asthma at age 5-7 years was analyzed using logistic regression models.First, we constructed a single regression analysis with four main effects (rhinovirus, S. pneumoniae, M. catarrhalis, and H. influenzae).This model, called the "Rhinovirus and Bacteria Model," describes the contribution of each variable to the outcome after adjusting for the other three covariables.Second, the "Full Adjusted Model" included rhinovirus and the three PPBs as main effects, and potentially confounding variables as covariables.The covariables selected were based upon the results of a recent systemic review, 26 expert clinician opinion, and directed acyclic graphs (Supporting Information: E-images 1 and 2).
These included: season of birth, childcare attendance, tobacco smoke exposure at birth, maternal (for wheeze outcome only) and parental (for asthma outcome only) history of asthma, gestational age at birth, exclusive breastfeeding during the first 3 months of life (for wheeze outcome only), cesarean birth and an older child in the household at birth.To examine the sensitivity of models, we constructed "Single Pathogen Models," which are individual models each with a single main effect, and "Single Pathogen Adjusted Models," which are the individual models adjusted for the covariables listed above.
Possible interaction effects were examined by adding the interaction term, assessing the statistical significance (alpha set at 0.05) and comparing the magnitude of the main exposure effect sizes with and without the interaction term.We examined the sensitivity of the timing of exposure by re-running models examining rhinovirus and PPB detections in the first 4 weeks and first 6 months of life. 19ditional analyses considering each of the three rhinovirus species in the first 3 months were also performed. 4Missing values were not imputed.All analyses were conducted using Stata statistical software v13 (StataCorp).

| Study population
Overall, 158 children were included in the full ORChID study, with 146 supplying diary data for the wheeze outcome and 84 providing asthma data at age 5-7 years.Characteristics of children in the full ORChID cohort were generally similar to those included in these analyses, except for a higher prevalence of tobacco smoke exposure at birth in excluded children (5/12 [41.7%]) compared to children included in the analysis (14/144 [9.7%]) for wheeze outcome (Supporting Information: E-Table 1).1).

Sociodemographic and clinical characteristics are described in
The 146 children who provided data for either outcome returned 1774 swabs during their first 3 months of life (Figure 1).Of 1400 swabs with an ERV3 value, 463 (33.1%) recorded a virus and/or PPB detections, compared with 54/344 (15.7%) of swabs with no ERV3 value recorded.The number of children returning swabs each week for their first 6 months of life (6 months chosen because of the sensitivity analysis) is displayed in Supporting Information: E-Image 3.
Rhinovirus and PPB were identified progressively from the first week of life (Supporting Information: E-Image 4).Rhinovirus was detected at least once in 71/146 (48.6%) children who provided diary data in the first 3 months of life (Supporting Information: E-Table 2).
When examining the sensitivity of results to the exposure period, the low frequency of pathogen detection in the first 4 weeks of life led to imprecise associations with wide CIs (Supporting Information: E-Table 6).When detections in the first 6 months of life were considered, findings were similar to the main analysis but with more precision (IRR: 1.13; 95% CI: 1.04-1.24).When rhinovirus species  E-Table 7).
Sensitivity analyses considering detections during the first 6-months showed similar effect sizes to the main results for rhinovirus, S. pneumoniae and M. catarrhalis (Supporting Information: E-Table 9).However, the association with H. influenzae was  While many studies have reported the associations between rhinovirus-induced wheezing illness and recurrent wheezing [27][28][29] and asthma, 6,12,29 it is not clear whether any rhinovirus infections (symptomatic or asymptomatic) in early life are associated with wheezing in the first 2 years of life. 30A potential rationale for the positive dose-response relationship with wheezing in the first 2 years of life in this study is that children who encounter a higher number of rhinovirus infections during the period from birth to the first 3 months of life may have an increased likelihood of subsequent rhinovirus infections until 2 years of age. 31Consequently, repeated rhinovirus infections may result in a higher incidence of wheezing episodes induced by rhinovirus, which also results in a higher incidence of childhood asthma.Abbreviations: CI, confidence interval; IRR, incident rate ratio; SD, standard deviation.a Mean number of weeks a pathogen was detected in the first 3 months of life.
b A Poisson regression analysis with four main effects (rhinovirus, S. pneumoniae, M. catarrhalis, and H. influenzae).
c Adjusted for season of birth, maternal asthma history, gestational age, delivery method, tobacco smoke exposure at birth, exclusive breastfeeding during the first 3 months of life, and older child in the household at birth.Childcare attendance at 3 months was omitted as only four children were attending any form of childcare.
T A B L E 3 Rhinovirus and potentially pathogenic bacterial detections in the first 3 months of life and asthma (ever) at ages 5-7 years (N = 84).Abbreviations: CI, confidence interval; OR, odds ratio; SD, standard deviation.a Mean number of weeks a pathogen was detected in the first 3 months of life.
b A single logistic regression analysis with four main effects (rhinovirus, S. pneumoniae, M. catarrhalis, and H. influenzae).
c Adjusted for season of birth, parental asthma history, gestational age, delivery method, tobacco smoke exposure at birth, exclusive breastfeeding during the first 3 months of life, and older child in the household at birth.Childcare attendance at 3 months was omitted as only four children were attending any form of childcare.
TAKASHIMA ET AL.
| 3433 three immunophenotypes, suggesting a heterogeneous immune response. 32Those at the greatest risk of asthma had a history of eczema, bronchiolitis associated with rhinovirus, higher peripheral blood eosinophil counts, and an upper airway microbiome dominated by H. influenzae and M. catarrhalis, suggesting an altered mucosal immune response to these PPB.The presence of these PPBs could be part of a more complex host-pathogen relationship and immunophenotype associated with asthma.
Birth cohort studies have reported mixed results on the early detection of PPB and its relationship to subsequent asthma.In the of life was significantly associated with chronic wheeze at age 5 years (OR: 3.8, 95% CI: 1.3-12.0). 16The authors reported that Haemophilus was very rare in nasopharyngeal samples from healthy infants, and Moraxella colonization was established later during infancy, which the authors suggested may be due to Perth's warm Mediterranean-like climate, and contrasts with the results in healthy infants from the warm humid subtropical climate of South-East Queensland.
The viral and bacterial interaction on recurrent wheezing was not detected in two other cohort studies.In the COPSAC birth cohort, upper airway viruses (OR: 2.8, 95% CI: 1.7-4.4)and bacteria (OR: 2.9, 95% CI: 1.9-4.3)detected by PCR assays and conventional culture methods, respectively, were associated with wheezing episodes, but the associations of viruses and bacteria were independent of one another. 17Another prospective cohort study compared oropharyngeal swabs from 109 children with an acute wheezing illness and 75 non-wheezing children attending a tertiary pediatric hospital in Perth, Australia. 34In this study, rhinoviruses did not have a significant impact upon bacterial community composition determined by 16S rRNA gene sequencing, and wheezing and viruses were not related to the bacterial community.These results implied there was no interaction between viruses and bacteria.
There are two possible, nonexclusive theories describing susceptibility to wheezing and asthma after early virus and PPB detections in young children.Firstly, host immune programming at birth is skewed toward Th2 responses and developing asthma.Infants with poor innate antiviral defenses develop more severe illnesses during infections with respiratory viruses. 35Early and more severe infections could lead to increased damage to the lower airways during a critical growth phase of the lungs, leading to changes in airway structure and function that promote asthma.In the COPSAC birth cohort, children who eventually developed asthma had an aberrant early life immune response evident by increased IL-5, IL-13, IL-17, and IL-10 production, which might predispose to persistent lower airway resident PPB colonization and result in chronic airway inflammation progressing to asthma. 13In addition, rhinovirus-related hospitalization rates are especially high for infants and children with asthma, which suggests viral infections and their associated induced host inflammatory response might directly injure the lower airways during acute infections. 36Alternatively, susceptible children may constitutionally have smaller airways and thus are predisposed to increased airway obstruction leading to recurrent wheezing during ALRIs and subsequently to asthma.This latter theory suggests that children who are born with small airways are also vulnerable to respiratory viral infections like rhinovirus and respiratory syncytial virus, which heighten the risk of ALRI and wheezing.
Strengths of the ORChID study include its longitudinal design and weekly sampling of nasal swabs in a healthy, communitybased birth cohort allowing the analysis of cumulative temporal rhinovirus and PPB detections in the first 3 and 6 months of life.
There are, however, several limitations.First, despite sensitive PCR assays, suboptimal swabbing techniques may have missed virus detections.Nevertheless, we have shown that parentcollected nasal swabs have similar virus detection rates to those obtained by health personnel when employing PCR assays. 37cond, the number of ALRI episodes in the first 3 months of life was low in the ORChID cohort, and thus their impact on subsequent wheezing episodes and asthma may have been underestimated.Third, our statistical analysis had suboptimal power for some comparisons due to the low incidence of H. influenzae detections.Fourth, we did not serotype the S.This study was conducted before the European guidelines were published, and we followed the Australian guidelines that recommended clinical assessment as the most reliable means of diagnosing asthma in this age group. 39Seventh, mainly because of low viral loads, one third of rhinoviruses were unable to be sequenced.[42] Eighth, we did not test the atopic status of children.Future studies should consider the relationship between atopy and early sensitization and pathogen detections.Finally, as often occurs with intense studies of this nature, the results may not generalize to children from other backgrounds and environmental settings.
Our cohort had a high rate of parental asthma which may reduce the generalizability of our findings.
In conclusion, early rhinovirus infections were associated with increased odds of wheezing in the first 2 years of life, and early H.
influenzae colonization was associated with increased odds of asthma between 5 and 7 years of age.The underlying mechanisms for developing recurrent wheezing and asthma after rhinovirus or PPB detections in young children remain unclear.There may be different immune responses to these microbes between atopic children, those with constitutional small airways, and otherwise healthy children.
Future research should explore whether recurrent wheezing and asthma are from inherited structural airway abnormalities, a complex interplay between rhinoviruses and abnormal host immune responses to airway colonization by PPB or viral infections, or all these factors. 43

2 | METHODS 2 . 1 |
Study subjects and settingThe Observational Research in Childhood Infectious Diseases (ORChID) study (clinicaltrials.gov:NCT01304914) was a community-based, birth cohort of acute respiratory infections in healthy children during the first 2 years of life.22ORChID study participants lived in the subtropical city of Brisbane, Australia.Mothers were recruited from antenatal clinics at one of two metropolitan hospitals (one private and one government-funded) between September 2010 and October 2012.Healthy children born at 36-42 weeks' gestation without congenital abnormalities or underlying chronic disorders were enrolled in the study.Parents provided informed consent for their child's participation shortly after birth.During the ORChID study, parents kept a daily symptom diary for their child, and collected weekly nasal swab specimens.Children exited the study when their parents stopped returning study material, or at their second birthday, whichever occurred earlier.At the end of their involvement with the ORChID project, children and their parents/caregivers were invited to participate in an extension study, the Early Life Lung Function (ELLF) study. 23This required an annual review by research staff between ages 3-7 years and included completion of a standardized respiratory health questionnaire.The Royal Brisbane and Women's Hospital (HREC/10/QRBW125) Human Research Ethics Committee (HREC) approved the ORChID study.The Children's Health Queensland (HREC/10/QRCH/16 and HREC/13/QRCH/156) and The University of Queensland (2010/ HE00820 and 2013/HE001291) HRECs approved the ORChID and ELLF studies.

3. 2 |
Rhinovirus and potentially pathogenic bacterial detections in the first 3 months and wheeze in the first 2 years of life There were 48/146 (32.9%) children whose parents reported they wheezed between ages 3 and 24 months.Each extra week of rhinovirus infection in the first 3 months of life led to a 1.16 times increase in wheezing within the first 2 years of life, after adjusting for the presence of PPB (incident rate ratio [IRR]: 1.16; 95% CI: 0.99-1.35;Table

T A B L E 1
Sociodemographic characteristics of ORChID (N = 146) and ELLF (N = 84) children providing wheeze or asthma outcome data.

4 |
DISCUSSIONIn this community-based birth cohort of healthy Australian children, each additional week of rhinovirus detections in the first 3 months of life was associated with an increased risk of reporting wheezing episodes in the first 2 years of life.However, rhinovirus infections in the first 3 months of life were not associated with asthma at ages 5-7-years.Although PPB did not have strong associations with wheezing, H. influenzae detections in early life increased asthma OR point estimate at age 5-7 years by 135%.No interaction effects between rhinovirus and PPB were observed for either wheezing or asthma.

H
. influenzae was associated with an increased risk of asthma, however, due to the small size of our sample, our findings should be interpreted with caution.Nevertheless, recent research using the clustering approach in a 17-center United States cohort study of 921 infants hospitalized with bronchiolitis identified T A B L E 2 Rhinovirus and potentially pathogenic bacterial detections in the first 3 months of life and risk of wheezing (ever) in the first 2 years of life (N = 146).
pneumoniae isolates.Fifth, unbiased next-generation sequencing to examine alterations in microbial community profiles during the first 3 months of life was not performed.However, low bacterial DNA loads in nasal swabs limit these studies to 16S rRNA gene sequencing and often organization taxonomic unit discrimination does not go beyond the genus level.Instead, we included all three PPBs in the "rhinovirus and bacteria model" to mimic the nature of co-existing bacteria found consistently in the upper airways by conventional culture-based methods.Sixth, we used an epidemiologic definition for the diagnosis of asthma, and it relied upon assessment by a physician and was not supplemented by at least two objective measures, such as fractional exhaled nitric oxide values, demonstration of airway hyperresponsiveness or a positive bronchodilator response as recommended by the European Respiratory Society in children aged 5-16 years.38

1
Flow chart of children and nasal swabs in the Observational Research in Childhood Infectious Diseases study.