Burden of Recurrent Respiratory Tract Infections in Children

A Prospective Cohort Study

Abstract

Background: 

The burden of recurrent respiratory infections is unclear. We identified young children with recurrent respiratory infections in order to characterize the clinical manifestations, risk factors and short-term consequences.

Methods: 

In this prospective cohort study, 1089 children were followed from birth to 2 years of age for respiratory infections by a daily symptom diary. Nasal swabs taken during respiratory infections were analyzed for viruses from 714 children. Nasopharyngeal swabs collected at 2 months of age were cultured for bacteria. The 10% of children with the highest number of annual respiratory illness days were defined to have recurrent respiratory tract infections.

Results: 

The 90th percentile in the number of annual respiratory illness days was 98. Children above this limit (n = 109) had a median of 9.6 acute respiratory infections per year. Rhinovirus was detected in 58% of their infections. Of the children with recurrent infections, 60% were diagnosed with at least 3 episodes of acute otitis media, 73% received at least 3 antibiotic treatments and 21% were hospitalized for an acute respiratory infection. Tympanostomy was performed for 35% and adenoidectomy for 13% of the children. Asthma was diagnosed in 12% by 24 months of age. Older siblings increased the risk of recurrent respiratory infections. Early nasopharyngeal colonization with Streptococcus pneumoniae was common in children who later developed recurrent infections.

Conclusions: 

Children with recurrent respiratory infections frequently use health care services and antibiotics, undergo surgical procedures and are at risk for asthma in early life. Having older siblings increases the risk of recurrent infections.

Respiratory tract infections are among the most common of all illnesses. Young children have more frequent respiratory infections than older children or adults; in those younger than 2 years of age, mean number of 5 to 6 episodes per year have been reported.^1–3 The vast majority of respiratory tract infections are caused by viruses,^1,2,4–8 but antibiotics are often used for otitis media or suspected bacterial lower respiratory tract infection.⁹

As children lose maternal immunoglobulin G during the first year of life and have limited acquired immunity, they frequently have symptomatic infections caused by respiratory viruses circulating in the community.⁸ While several annual episodes of uncomplicated respiratory infections in young age are typical, particular attention should be paid to children who have unusually frequent or prolonged infections. There is no consensus on how recurrent respiratory infections in children should be defined, and only limited data are available about this condition. Frequent respiratory infections have been associated with increased exposure to microbes through close contacts in group day care,^10–12 with genetic variants of innate immunity factors, such as mannose-binding lectin, interleukins 6 and 10 and Toll-like receptor 4, and with lowered interferon production.^13–16

Previous studies have defined recurrent respiratory infections by a certain number of infection episodes per year or focused on specific diagnoses.^15,17,18 However, infections may overlap with each other or be prolonged, especially in those with recurrent infections, and focusing on a specific diagnosis does not characterize the overall disease burden affecting the children and the families. In this prospective birth cohort study, we identified children with recurrent respiratory tract infections using a percentile limit, which we set at 90%, in days with respiratory symptoms per year. We describe the early risk factors, clinical and virological characteristics and short-term consequences of recurrent respiratory tract infections in children younger than 2 years of age.

MATERIALS AND METHODS

Study Design and Conduct

In the prospective, observational birth cohort study, Steps to the Healthy Development and Well-being of Children (STEPS), families of 1827 children (30 pairs of twins) were recruited during the first trimester of pregnancy or soon after birth from a cohort of all children born in the Hospital District of Southwest Finland between January 2008 and April 2010 to their Finnish- or Swedish-speaking mothers (eligible cohort, n = 9811 mothers; n = 9936 children; Fig. 1).¹⁹ Children were followed from birth to 2 years of age for respiratory infections by daily diaries, where parents recorded symptoms, physician visits with diagnoses and treatment and absences of child from day care or parent from work. Children were invited to scheduled visits at 13 and 24 months of age. Background information was gathered by structured question forms before birth or soon after birth and at the age of 13, 18 and 24 months. Data on atopic and allergic conditions were collected at 24 months of age with qualified forms (The International Study of Asthma and Allergies in Childhood, ISAAC).

FIGURE 1: Enrollment and follow-up of the study children. Children were recruited during the first trimester of pregnancy or soon after birth from a cohort of all children born in the Hospital District of Southwest Finland between January 2008 and April 2010 to their Finnish- or Swedish-speaking mothers.

A subgroup of 923 children were recruited without selection criteria into an intensive follow-up on respiratory infections from birth to 2 years of age. The children were examined at the study clinic by a physician during respiratory infections, and nasal samples were collected with Copan flocked nylon swabs (Copan, Brescia, Italy) from the depth of 2–3 cm from both nostrils. If the parents felt that an evaluation at the study clinic was not needed, nasal samples were collected by the parents and sent to the study clinic via standard mail as described previously.^20,21 The children in the intensive follow-up were invited to scheduled visits also at the age of 2 months, and nasal samples for virus analyses were collected at all scheduled visits by study personnel from these children. Bacterial culture of a nasopharyngeal sample collected at the age of 2 months was done for 312 children visiting on Monday or Tuesday. Data on hospitalizations of the children in the intensive follow-up were collected from electronic registries.

An episode of acute respiratory infection was defined as the presence of nasal discharge, nasal congestion or cough, with or without fever or wheezing, documented in the diary by the parents, or as a diagnosis of an acute respiratory infection by a physician. If there were repeated diagnosis of acute otitis media, wheezing illness, pneumonia, laryngitis or pharyngitis, or several virus samples during continuous respiratory symptoms, or if data on duration of symptoms were missing, diagnoses and virus detections within 14 days were calculated as 1 episode. Illness rates per year were calculated from the duration of active follow-up for each child.

Data on children with completed diary for at least 1 year were included in the analyses. The 10% of children with the highest number of days per year with respiratory symptoms were defined to have recurrent respiratory tract infections.

The Ministry of Social Affairs and Health and the Ethics Committee of the Hospital District of Southwest Finland approved the STEPS Study. Parents of participating children gave their written informed consent.

Respiratory Virus Detection and Bacterial Culture

The nasal swab specimens were stored at −80°C before analysis. Swabs were suspended in phosphate-buffered saline, and nucleic acids were extracted by NucliSENS easyMag (BioMerieux, Boxtel, the Netherlands) or MagnaPure 96 (Roche, Penzberg, Germany) automated extractor. Extracted RNA was reverse transcribed, and the cDNA was amplified using real-time, quantitative polymerase chain reaction for rhinovirus, human enteroviruses and respiratory syncytial virus (RSV) as described earlier.^22,23 Influenza polymerase chain reaction was performed for all samples collected during influenza seasons defined on the basis of the influenza antigen test results, and data from the infectious disease surveillance registry of the National Institute for Health and Welfare, Finland,²⁴ and laboratory developed antigen detection tests were performed for influenza A and B viruses, parainfluenza type 1, 2 and 3 viruses, RSV, adenovirus and human metapneumovirus for samples collected in January 2009 or later. Semiquantitative bacterial culturing and identification from nasopharyngeal swabs was done as described earlier.²⁵

Statistical Analysis

Children with recurrent respiratory tract infections were compared with the other study children. Dichotomous data were compared by using the χ² test or Fisher exact test. Normally distributed continuous data were described and compared by using means, standard deviations and Student’s t test, and skewed data with medians, interquartile ranges and Mann–Whitney U test. Continuous data, which upper quartile was zero, were presented as frequency tables. Risk for recurrent respiratory infections was analyzed by binary logistic regression analysis using first sex, siblings, mother’s educational level, living environment and breast-feeding as predictors. Then, the logistic regression analysis was performed for nasopharyngeal bacteria (Streptococcus pneumoniae and Moraxella catarrhalis), adjusted for sex, siblings, mother’s educational level, living environment, and breast-feeding. Parental smoking was inversely associated with recurrent respiratory infections in univariate analysis and was not included in the adjusted analysis because of suspected reporting bias. Odds ratios (OR) with 95% confidence intervals (CI) were determined. P values <0.05 were considered statistically significant. The data were analyzed with the use of SPSS software, version 23.0 (IBM SPSS Statistics for Macintosh, Armonk, NY: IBM Corp.), and SAS software, version 9.4 (SAS Institute Inc., Cary, NC).

RESULTS

Study Population

Recruitment and follow-up of the study children is shown in Figure 1. The symptom diary was returned by families of 1161 children, and 1089 had completed the daily diary for at least 1 year and were included in the analyses. Of 1089 children, 714 (66%) were in the intensive follow-up group and 375 (34%) in the regular follow-up group. Baseline characteristics of the study children are shown in Table 1. Rates and characteristics of acute respiratory infections in children in the intensive and regular follow-up are compared in Table, Supplemental Digital Contents 1 and 2, https://links.lww.com/INF/C526, https://links.lww.com/INF/C527. Children in the intensive follow-up visited a physician and were diagnosed with a wheezing illness more frequently than those in the regular follow-up, but there were no differences in rates of acute otitis media, pneumonia, laryngitis or antibiotic treatments for acute respiratory infections.

TABLE 1: Baseline Variables in Children With and Without Recurrent Respiratory Tract Infections

Identification of Children With Recurrent Respiratory Tract Infections

The median number of respiratory illness days per year was 44.2 (interquartile range, 24.7, 69.0) per child, and the 90th percentile was 98.0 days per year (Fig. 2). Children with more than 98 respiratory illness days per year were defined as having recurrent respiratory infections (n = 109), and they were compared with all other children (n = 980).

FIGURE 2: Distribution of the frequency of days with symptoms of respiratory infection per year in the study population. The dotted line shows the 90th percentile (98.0 days per year).

Burden of Disease

The children with recurrent respiratory tract infections documented a median of 113 days with respiratory symptoms, 9.6 acute respiratory infection episodes and 6.0 physician visits for respiratory infection per child per year (Table 2). All documented diagnoses and outcomes except laryngitis and allergic rhinitis were significantly more frequent in children with recurrent respiratory infections than in other children. The number of respiratory illness days per year was highest from the age of 6 to 23 months, but substantial already before 6 months of age. Of the children with recurrent respiratory infections, 60% had at least 3 episodes of acute otitis media, and 73% were treated with antibiotics at least 3 times. During the follow-up, 33% were diagnosed with an acute wheezing illness, 16% had recurrent wheezing and 21% were hospitalized for an acute respiratory infection. Use of analgesics, inhaled asthma medications and systemic corticosteroids was common among children with recurrent infections. Tympanostomy tubes were inserted to 35% of the children, and adenoidectomy was performed for 13%. By 24 months of age, 12% of the children with recurrent respiratory infections were diagnosed with asthma.

TABLE 2: Disease Burden Caused by Respiratory Infections in Children With and Without Recurrent Respiratory Tract Infections

Of the overall respiratory disease burden, 25% of all respiratory illness days, 21% of physician visits and 32% of hospitalizations occurred to the children with recurrent respiratory infections, and 19% of acute otitis media episodes and 28% of wheezing illnesses were diagnosed in these children. Of all antibiotic treatments prescribed for respiratory infections in this cohort, 20% were prescribed to children with recurrent infections, and 23% of all tympanostomy tube placements were made to these children (Table, Supplemental Digital Content 3, https://links.lww.com/INF/C528).

The first acute respiratory infection occurred before 3 months of age in 84 (77%) and in 427 (45%) children (P < 0.001), and acute otitis media before 6 months of age in 35 (32%) and in 104 (11%) children with and without recurrent infections, respectively (P < 0.001).

Risk Factors

Older siblings and living in the rural area associated with recurrent respiratory infections in the unadjusted analysis (Table 3). After adjustment, only older siblings remained as a significant risk factor [adjusted OR (aOR), 3.03; 95% CI: 1.94–4.74].

TABLE 3: Risk Factors for Recurrent Respiratory Tract Infections

A nasopharyngeal sample was cultured for bacteria from 312 children at the age of 2 months. S. pneumoniae was found in 38 (12%), M. catarrhalis in 72 (23%), Haemophilus influenzae in 3 (1%) and other bacteria in 265 (85%) children. Early nasopharyngeal colonization with S. pneumoniae associated with later recurrent respiratory tract infections (27% vs. 10%; OR, 3.26; 95% CI: 1.38–7.68). This association was not significant in the adjusted analysis (adjusted OR, 2.44; 95% CI: 0.93–6.39).

Detected Viruses

Nasal samples were collected during 4324 of 8272 (52%) acute respiratory infections in the intensive follow-up group. Viruses detected in children with and without recurrent respiratory infections are compared in Table 4. Rhinovirus was the most common virus in both groups (58% vs. 59% in children with and without recurrent respiratory infections), and any other virus was found in <13% of the infections in each group.

TABLE 4: Detected Viruses During Acute Respiratory Infections in Children With and Without Recurrent Respiratory Tract Infections

Nasal samples were collected at 1957 visits scheduled at 2, 13 or 24 months of age (Table 5). Of these samples, 399 (20%) were virus positive, and 354 (18%) were positive for rhinovirus. Rhinovirus was found in 113 (8%) asymptomatic children and in 232 (42%) symptomatic children at scheduled visits (P < 0.001). In children with recurrent respiratory infections, who were asymptomatic at the time of visit, rhinovirus was detected in none at 2 or 13 months of age, and in 8 (27%) at 24 months of age. In contrast, in the group without recurrent respiratory infections, rhinovirus was constantly found in 8%–9% of asymptomatic children at 2, 13 or 24 months of age (P = 0.05, P = 0.06 and P = 0.001, respectively).

TABLE 5: Detected Viruses at Scheduled Visits in Children With and Without Recurrent Respiratory Tract Infections

DISCUSSION

We identified children with recurrent respiratory tract infections on the basis of the number of reported respiratory illness days per year, with the 90th percentile (98 annual illness days) as the limit. With this case definition, we included even mild illnesses not always necessitating a physician visit, and we document a high median number of infection episodes per year (9.6) in this group of children. Clustering of different upper and lower respiratory tract infections and related morbidity in these children was evident. The most common diagnoses were uncomplicated upper respiratory tract infection and acute otitis media, but many children also had bronchiolitis or pneumonia. Children with recurrent respiratory infections had frequent physician office and emergency room visits, they often use antibiotics, and a substantial proportion of them underwent surgical operations such as tympanostomy tube placements, or were hospitalized because of respiratory infections. Twelve percent were diagnosed with asthma by the age of 2 years. Almost 80% had the first acute respiratory infection before the age of 3 months, and one-third had the first acute otitis media before that age. The morbidity in children with recurrent respiratory tract infections reflects in multiple ways on their parents and other family members, documented in this study in the increased number of absences from work by parents.

Setting the limit of recurrent respiratory tract infections at the highest 10% is debatable. The considerable morbidity that we document in children identified using this limit supports its use. Alternatively, another percentage limit could have been used or a limit at a certain number of infection episodes, as in previous studies.¹⁸ Our principal argument against using the number of infection episodes in identification of children with recurrent respiratory infections was that, when documented in detail in daily diaries, infection episodes overlap with each other, particularly in the case of frequently recurring infections. The duration of respiratory symptoms in a rhinovirus infection has been reported as 1 to 2 weeks, but in can be longer, and different rhinoviruses circulate simultaneously and cause infections overlapping with each other.^26–28

Previously documented environmental risk factors for respiratory infections include close contact with other children, male sex, tobacco smoke exposure, lower socioeconomic status and lack of breast-feeding,^{10–12,29–32} but also higher educational level of the family has been reported to associate with an increased infection rate.³³ In this study, older siblings were independently associated with a risk for recurrent respiratory tract infections. Nasopharyngeal colonization with S. pneumoniae at the age of 2 months was more common in children with recurrent respiratory infections, but the association did not remain significant after adjustment. This was probably because of older siblings transmitting pneumococci early to the infant.³⁴ Nevertheless, this finding could reflect the role of S. pneumoniae in the increased rate of otitis media and other infections in young children with close child contacts. Previous studies have linked early nasopharyngeal bacterial colonization to acute otitis media, recurrent wheezing and asthma.^35,36

Children attending day care at the age of 13 or 18 months were not overrepresented in the group with recurrent respiratory infections in this study. However, cross-sectional data on day-care attendance may have left its effect unclear, as the age when children started day care varied notably, and our definition of recurrent infections was based on the number of respiratory illness days during the whole follow-up period. More detailed analyses on the influence of the day care in the rates of respiratory infections are warranted. Parental smoking seemed to be associated with a decreased risk of recurrent respiratory tract infections in unadjusted analysis, but we assume that this could be due to more accurate documentation of respiratory symptoms in children by the nonsmoking parents from higher socioeconomic class, similar to an earlier study.³³

In our study, rhinovirus was the salient cause of recurrent respiratory tract infections. There was no difference in the distribution of viruses between children with or without recurrent infections. RSV and influenza A and B viruses, which are important causes of hospitalizations in children, had minor roles compared with rhinovirus, but these and other respiratory viruses added on the disease burden caused by rhinovirus. Notably, a large portion of the study children were immunized against influenza. Our results agree with the earlier data of rhinovirus being the most common cause of respiratory tract infections in children.^2,4,5,18 In addition to sampling at the study clinic, we used home nasal sampling by the parents, which has been found feasible and adds data on mild infections that do not necessitate a physician visit.²⁰

An unexpected but interesting finding was that none of the children with recurrent respiratory infections were positive for rhinovirus, or any other virus, at asymptomatic state at 2 or 13 months of age. In our comparison group, 8%–9% of asymptomatic children were rhinovirus positive at these ages, and other studies have reported high rates of rhinovirus positivity in asymptomatic children.³⁷ It could be hypothesized that the absence of asymptomatic rhinovirus infections at young age in children with recurrent infections would be due to aberrant innate immune responses. At 24 months of age, when children have more developed immunity, children with recurrent infections had even more asymptomatic rhinovirus findings than the other children. However, these findings should be interpreted cautiously because of low numbers of asymptomatic subjects in the group with recurrent respiratory infections.

There are limitations in this study. Our study children had less frequently older siblings and were more often from families with higher occupational class than those who did not participate. This fact, and the drop-out of part of the children during the follow-up, may limit the generalizability of our results. Our study design with a study clinic available for acute illness visits for children in the intensive follow-up might have arbitrarily increased the numbers of diagnoses. Indeed, these children visited a physician and were diagnosed with a wheezing illness more frequently than children in the regular follow-up. However, the rates of other diagnoses and treatments did not differ.

CONCLUSION

We present here data that should help in identification of children with recurrent respiratory tract infections. We report that these children frequently use healthcare services and antibiotics, undergo surgical procedures and are at risk of developing asthma. Rhinovirus is the principal causative agent of their infections. The present data should be considered in future studies, aiming to develop prevention and treatment of recurrent respiratory tract infections in children.

ACKNOWLEDGMENTS

We thank all 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; study nurses Niina Lukkarla, Petra Rajala and Mira Katajamäki for their assistance in the study clinic; Tiina Ylinen for her technical assistance in virus detection; Päivi Haaranen for her technical assistance in bacterial culture; Anne Kaljonen for her assistance with data handling; and Jaakko Matomäki for his assistance with statistical analyses.