We conducted a prospective cohort study in newborns to analyse the association between past viral infections of varying severity and the development of recurrent wheezing. We requested the participation of all newborns delivered in four health care centres in the Área Sur health zone of Madrid during their first neonatal visit (7–14 days of life). The parents or guardians signed an informed consent form and completed a questionnaire of epidemiological data (history of asthma and atopy in the family, exposure to tobacco smoke, pregnancy and delivery data, number of siblings). Nasopharyngeal aspirate (NPA) specimens were collected during this visit and routine checkups performed at ages 2, 4, 6 and 12 months. Furthermore, if the infants developed respiratory symptoms, they were evaluated by a paediatrician, with collection of an additional NPA sample.

We considered children with specimens collected during routine well-child checkups healthy controls, even if they presented with isolated rhinorrhoea. We defined asymptomatic infection as detection of a virus in the NPA of one of these patients. We defined clinically significant respiratory infection as a case that met the clinical criteria for respiratory infection presented in Table 1. We defined “first detected viral infection” as PCR detecting a respiratory virus in a sample for the first time in the life of a child, whether the infection was asymptomatic, symptomatic and managed at the outpatient level, or symptomatic and requiring hospital admission.

When a child met the criteria for respiratory infection described in Table 1, the paediatrician conducted an evaluation, a clinical data form was completed, and a NPA sample collected. If the infant required admission to hospital, data for the most relevant variables were also collected in another form. Isolated rhinorrhoea was not considered a symptomatic respiratory infection. We differentiated between the following forms of respiratory infection: upper respiratory tract infection, defined as infection meeting the criteria specified in Table 1 in the absence of respiratory distress or wheezing. We defined bronchiolitis as the first episode of respiratory infection that manifested with respiratory distress and wheezing. Subsequent episodes of similar characteristics were classified as recurrent wheezing. Episodes manifesting with inspiratory dyspnoea and wheezing were classified as laryngotracheobronchitis. Laryngitis was defined as inspiratory dyspnoea without wheezing. We defined pneumonia as evidence of focal infiltrates or consolidation in chest radiography in the absence of wheezing.

We classified all infections requiring hospital admission as severe, and the rest as mild.

When the requisite criteria were met, a NPA sample was collected and kept at 4°C to await transportation to the Influenza and Respiratory Viruses Laboratory of the Centro Nacional de Microbiología (National Centre of Microbiology). In hospitalised patients, the NPA sample was collected between 7 and 8AM of the morning following admission and kept refrigerated until it was transported.

Three RT-nested PCR assays were used to test for 16 respiratory viruses. Reverse transcription (RT) and the first amplification reaction were performed with OneStep RT-PCR kits (Qiagen). Influenza A, B and C were detected by a previously described PCR assay.15 A second multiplex PCR assay was used to detect parainfluenza virus (PIV) 1 through 4, coronavirus 229E and OC43, enterovirus and RV.16 The presence of RSV-A and B, hMPV, hBoV and adenovirus (ADV) was assessed with a third RT-nested PCR assay using the BRQ method.3

We have expressed discrete data as percentages and continuous data as mean and standard deviation. We compared clinical and laboratory characteristics using the Student's t, Mann–Whitney U, chi-square and Fisher exact tests. We defined statistical significance as a P-value of less than .05 in any of the tests. The statistical analysis was performed with the Statistical Package for the Social Sciences (SPSS) version 13.0.

We recruited a total of 302 infants at birth, of which 55.3% were male (167), 80% born to Spanish parents, and 14% of Latin-American ancestry. Thirty-five percent of infants were exposed to tobacco smoke; 36% of mothers and 37% of fathers had asthma and/or atopy. Only 2 patients were born preterm at a gestational age of 33 weeks. All infants but 7 completed the yearlong followup. Those 7 patients did not attend the 12-month checkup, but we contacted their paediatricians to obtain information about their clinical outcomes, so they were ultimately included in the analysis.

A total of 1,239 samples were collected, 1,005 during routine visits and 288 when a respiratory infection was suspected. Of the samples collected during the well-child checkups, 30.8% (310) tested positive, while viruses were detected in 77.8% of respiratory infection cases (224/288), P<.001 (odds ratio, 3; 95% confidence interval, 2.4–3.8). The frequency of viral detection in asymptomatic children (healthy controls) increased proportionally with age from 20.8% in newborns aged 15 days to up to 42% in infants aged 12 months. The virus identified most frequently was RV, found in 82% (254) of cases. Other viruses were found in a small proportion of cases: ADV (9.7%), PIV (4.2%), hBoV (3.2%), hMPV (2.9%), FLU (2.2%) and RSV (1.6%) (Fig. 1).

Figure 1.

Viral detection in respiratory infections and controls. hRV, human rhinovirus; RSV, respiratory syncytial virus; PIV, parainfluenza virus; hMPV, human metapneumovirus; AD, adenovirus; FLU, virus influenza; hBoV: human bocavirus.

(0.11MB).

Of the 302 infants included in the study, 239 (79%) had at least one positive viral detection in the first year of life (Fig. 2). The virus detected most frequently was RV (181 samples, 71%), followed by PIV (6.3%), ADV (5%) and RSV (6%). Coinfections were only detected in 21 cases (8.8%). The mean age at detection of the first viral infection was 3.4 months (SD, 2.8). The first detected viral infection was associated with upper respiratory tract infections in 45 patients (20%), bronchiolitis in 19 (8.5%) while remaining 159 samples (70%) corresponded to asymptomatic patients and had been collected during routine well-child checkups.

Figure 2.

First positive viral detection and associated clinical diagnosis.

(0.1MB).

We analysed the outcomes of infants that had at least one positive specimen, and found that 27 (11%) developed recurrent wheezing in the first year of life (mean±SD, 2.9±1.2 episodes; range, 2–5) (Fig. 3). Rhinovirus was the first virus detected in 18 patients (66%), of who 14 (52%) were asymptomatic at the time of the specimen collection (routine checkup). The mean age of infants that developed recurrent wheezing with their first positive viral detection was 2 months (SD, 1.3; all but three were at least 3 months old). Ten of the 27 infants with episodes of bronchiolitis required admission (37%). The most frequently involved virus was RV, both in mild bronchiolitis cases managed at the outpatient level (11/17, 69%) and in severe cases that required hospitalisation (7/10, 70%).

Figure 3.

Severity of disease associated to the first positive viral test result in infants that developed recurrent wheezing.

(0.06MB).

Recurrent wheezing episodes (at least one following the initial bronchiolitis) occurred in 58.3% of infants in who the first viral detection corresponded to a severe infection that required admission, compared to 8.6% of infants whose first detection corresponded to a mild infection managed at the outpatient level or an asymptomatic infection (P<.001; odds ratio, 2.18; 95% CI, 1.05–4.5). A total of 5 infants received montelukast or budesonide during the followup (18%, 3 admitted to hospital and 2 managed at outpatient level at first positive viral detection). There were no differences in the family history of asthma or atopy between children with and without recurrent wheezing.

There were 288 episodes of infection managed in outpatient settings that corresponded to 160 infants (53% of the total sample) who had between 1 and 6 episodes each (mean±SD, 1.6±1). The most frequent symptoms were: fever in 90 cases (38%), rhinorrhoea in 224 (96%), cough in 223 (96%), respiratory distress in 51 (21%) and nonspecific symptoms in 35 (15%). The most frequent diagnoses were upper respiratory tract infection (147, 51%) and bronchiolitis (48, 16%), with a lesser incidence of recurrent wheezing, laryngitis and pneumonia. Only 2 patients had infiltrates on radiologic examination, and both required hospital admission during the followup. The viruses detected in these infections were RV in 127 cases (55.8%), followed by PIV (13%), RSV (12.5%), hMPV (8.9%) and FLU (8%) (Fig. 2). The causative agent of upper respiratory tract infection was RV in 61% of cases, followed by PIV (13%) and ADV (11%). When it came to bronchiolitis, RV was detected in 53%, RSV in 25% and hMPV in 8% of cases.

We analysed a total of 19 hospital admissions corresponding to 14 infants (14/302; 4.6% of the cohort; 5.8% of infants with a positive viral detection). Five patients were admitted more than twice in the first year of life. Fifty-eight percent of admitted patients (11) were female. The mean age of admission was 3.6 months (SD, 2.5 months; range, 6 days–8 months). Forty-two percent presented with fever (mean, 38.6°C; SD, 0.9°C) lasting 3.5 days (SD, 2). Sixty-eight percent (13/19 episodes) required oxygen supplementation for a mean of 5 days (SD, 3). Twenty-six percent (5/19) had infiltrates detected by radiology. The diagnoses were bronchiolitis in 12 cases (63%), wheezing episode in 3 (15%), upper respiratory tract infection in 3 and pneumonia in 1. Only 3 infants received antibiotics. Respiratory syncytial virus was detected in 42% of the cases and RV in 52%. Viral coinfection was detected in 4 cases.