ARTICLE |
a Danish Epidemiology Science Centre, Statens Serum Institut, Denmark
b Bandim Health Project, Bissau, Guinea-Bissau
c Paediatric Clinic 2, Rigshospitalet, Denmark
d Department of Pediatrics, Section of Infectious Diseases, University of Colorado School of Medicine, and The Children's Hospital, Denver, Colorado
| ABSTRACT |
|---|
|
|
|---|
PATIENTS AND METHODS. Relative risks for respiratory syncytial virus hospitalization were studied in a nested 1:5 case-control design using exposure information obtained from interviews with mothers of 2564 case and 12816 control children who had been followed prospectively from birth and until 18 months of age as participants in the Danish National Birth Cohort. Information on the children's ages at respiratory syncytial virus hospitalization, presentation of infrequent wheezing, recurrent wheezing, and atopic dermatitis were used to study these associations chronologically.
RESULTS. The adjusted relative risk of respiratory syncytial virus hospitalization in the offspring was 1.11 for maternal atopic dermatitis, 1.72 for maternal asthma, and 1.23 for paternal asthma. Atopic dermatitis in the child was associated with an increased risk of subsequent respiratory syncytial virus hospitalization among infants <6 months of age. Infrequent wheezing was associated with a relative risk of subsequent respiratory syncytial virus hospitalization of 2.98 and recurrent wheezing with a relative risk of 5.90. These associations were present also if infants with medical risk factors were excluded from the analysis.
CONCLUSIONS. Asthmatic disposition and wheezing were strong determinants of subsequent respiratory syncytial virus hospitalization in Danish children <18 months of age.
Key Words: acute lower respiratory tract infection atopy hospitalization respiratory syncytial virus wheezing
Abbreviations: RSV—respiratory syncytial virus LRI—lower respiratory tract infection DNBC—Danish National Birth Cohort RR—relative risk CI—confidence interval
Respiratory syncytial virus (RSV) is the most important cause of viral lower respiratory tract disease in infants and children worldwide and the most common cause of childhood hospitalization in high-income countries.1 In industrialized countries, the rates of RSV hospitalization have been increasing,2 and, likewise, an increase in atopic disease has been reported.3
In prospective cohort studies, RSV bronchiolitis during infancy4,5 or in the first 3 years of life has been found to be an important risk factor for the development of subsequent wheezing and asthma in the ensuing 7 to 11 years,6 and a causal relation between RSV hospitalization and subsequent hypersensitive airways has been suggested.7 However, it has also been suggested that a family history of atopy may dispose infants to develop more serious RSV lower respiratory tract infections (LRIs),8 and studies have shown that a parental history of asthma, especially on the maternal side, increases the risk of developing LRI in early life.9 It is still unclear, however, whether RSV LRI is causal in the development of subsequent recurrent wheezing and asthma or whether severe RSV LRI occurs on the background of an atopic predisposition and/or hypersensitive airways. The present study examined whether atopic disposition, wheezing, and atopic disorders were determinants for subsequent RSV hospitalization.
| METHODS |
|---|
|
|
|---|
The Danish National Birth Cohort
The Danish National Birth Cohort (DNBC) is a cohort of >100000 pregnant women recruited consecutively from 1997 to 2003 and, after written consent, followed throughout pregnancy until 18 months postpartum.12 Danish-speaking women who intended to carry their pregnancy to term were invited to participate at their first prenatal consultation. Participation involved 4 structured computer-assisted telephone interviews performed by trained interviewers at gestational weeks 12 and 30 (interviews 1 and 2) and postnatal when the child was 6 and 18 months old (interviews 3 and 4). The women were called
4 times for each interview. Children were included and followed-up despite missing interviews. In the present study, children were included without regard to missing information, and the percentages of missing interviews were 7.4% of the first, 7.5% of the second, 26.2% of the third, and 30.8% of the fourth interview. In the interviews, information on health, lifestyle, and socioeconomic status of the parents was collected. For the children, data regarding feeding habits, diseases, medication, vaccinations, motor and cognitive development, anthropometry, and day care attendance were obtained. The age at which an illness occurred was registered at fortnightly intervals, day 7 in that period being used as the age at onset of that illness episode.
Study Design
The study was a nested case-control study in the DNBC cohort to determine risk factors for RSV hospitalization in infants and children aged <18 months. Using the central person registry number, the DNBC was linked with the RSV database to identify individuals with
1 hospitalization because of RSV infection (the cases). For each case child, 5 control children matched by date of birth ± 1 day were selected randomly among DNBC children under risk for RSV hospitalization at the date when the case became RSV hospitalized.
The study was approved by the Danish Data Protection Board and the Danish Central Ethical Commitée.
Risk Factors Examined
From the Danish National Patient Registry, we obtained data on gestational age, birth weight, nonatopic medical risk factors in the child (intrauterine growth retardation, premature birth [gestational age <259 days], low birth weight [
2500 g], cystic fibrosis, chronic lung disease of prematurity, congenital heart disease, immune suppression, Down syndrome or other chromosomal disorders, cerebral palsy, or neuromuscular disease), and whether parents were living together at time of the birth.
Definitions of Risk Factors
Atopic Dermatitis
The definition of atopic dermatitis was based on a subgroup study of responses in the DNBC.13 The definition that resulted in the highest sensitivity (81%) and specificity (91%) for identifying atopic dermatitis diagnosed by a dermatologist was used in the present study. The age at the first episode of atopic dermatitis was used to describe whether the child had atopic dermatitis when hospitalized with RSV infection.
Infrequent Wheezing
Information on wheezing was obtained from the third and fourth interview from the Danish National Birth Cohort. In these 2 interviews, the mother/caretaker was asked whether the child had experienced any wheezing episodes and, if so, how many and whether a doctor had verified the diagnosis. Infrequent wheezing was defined as <3 caretaker-identified wheezing episodes during the follow-up period, regardless of a physician diagnosis of asthma (118 of 342 children with infrequent wheeze had a physician diagnosis of asthma).
Recurrent Wheezing
Recurrent wheezing was defined as
3 caretaker-identified wheezing episodes and a physician's diagnosis of asthma or asthmatic bronchitis in accordance with previous studies on the association between RSV and wheezing.7 The age at the first episode of wheezing (determined as the seventh day in the first half-month period in which the mother reported any wheezing) was used to describe age at onset of wheezing. Children who met the definition of recurrent wheezing were regarded as having recurrent wheezing from the age of their first wheezing episode. To test the robustness of this approach, we further reanalyzed the data with a latency period of another week until the age at onset. Results were essentially the same and have, therefore, not been shown. We also tested the risk of RSV hospitalization between 6 and 18 months of age only among children who had infrequent or recurrent wheezing in the first 6 months of life ("stringent criteria" analysis).
Maternal Asthma
Information on maternal asthma was obtained from the first interview. A combination of affirmative answers to questions about any asthma ever and a physician's diagnosis of asthma and ongoing asthma was used to define maternal asthma.
Paternal Asthma
In the first interview the mother was asked whether the father suffered from asthma. A confirmative answer defined paternal asthma.
Statistical Methods
Information on breastfeeding, day care attendance, atopic dermatitis, infrequent wheezing, and recurrent wheezing in the child was included in the multivariable analysis as time-dependent variables.
Breastfeeding
The age when breastfeeding was discontinued was used to describe the current breastfeeding status at age of RSV hospitalization.
Day Care
Information on the child's age at start of any day care attendance was obtained from the third interview when the child was 6 months old and was used to describe whether the child attended day care when hospitalized with RSV. Information on day care attendance was also collected in the fourth interview but without information on the child's age when starting day care. Hence, the study is likely to underestimate the effect of day care, because only 12.6% of the children attended day care at 6 months of age compared with 88.2% at 18 months of age.
Cox regression analysis stratified for the matched case-control design was used to obtain estimates of relative risk (RR) of RSV hospitalization associated with the different risk factors. The RRs presented are the estimates from a multivariable model including information on all of the factors under study (gender, medical risk factors, breastfeeding, day care attendance, atopic dermatitis, wheezing, recurrent wheezing, cohabitation, household income, square area of the home in meters squared, furred pets in the home, rural residence, maternal alcohol consumption, maternal smoking, maternal education, maternal occupation, maternal and paternal atopic dermatitis, maternal and paternal allergic rhinitis, maternal and paternal asthma, atopic dermatitis among siblings, and previous RSV hospitalization). Hence, all of the estimates of RR were adjusted for all of the other factors examined for a significant association with RSV hospitalization. Only the factors having a significant association with RSV hospitalization are presented in the results. Because case-control pairs were matched on date of birth, the estimates are already adjusted for age and RSV seasonality.
Missing values from the interview were coded as separate strata for each variable and included in the multivariable model; none of these strata were significantly associated with the outcome. Because of the higher percentage of missing interviews in the contact with the mother after delivery (third and fourth interview), information was more often missing for child characteristics than for parental characteristics, which were mostly obtained in the first and second interviews.
To be able to study the interaction between age and possible determinants for RSV hospitalization, data were analyzed divided in age groups: 0 to 5 months, 6 to 11 months, and 12 to 18 months of age with allowance for differences between the age groups in the effect of the factors under study.
| RESULTS |
|---|
|
|
|---|
The distribution of background factors significantly associated with RSV hospitalization is presented in Table 1; after adjustment, male gender, medical nonatopic risk factors, the presence of other children <12 years of age in the home, day care attendance, and maternal smoking were associated with an increased risk of RSV hospitalization. Maternal unemployment or unskilled work tended to be associated with an increased risk of RSV hospitalization.
|
After dividing data into 3 age groups (0–5, 6–11, and 12–18 months) and analyzing the association between background factors and RSV hospitalization in each age group, it was observed that the associations with male gender and the presence of other children <12 years of age in the home were particularly strong or present only among infants <6 months of age (Table 1). Breastfeeding was protective against RSV hospitalization in infancy. Among infants aged 6 to 11 months, there was an increased risk of RSV hospitalization among infants attending day care. Furthermore, the association with maternal smoking may have increased with increasing age of the child.
Atopic and Asthmatic Disposition
Asthmatic disposition increased the risk of RSV hospitalization (Fig 1). If the mother reported physician-diagnosed ongoing asthma, the risk of RSV hospitalization was increased by 72% (95% confidence interval [CI]: 44%–106%). If the mother reported paternal asthma, the risk was increased by 23% (95% CI: 4%–45%). Maternal atopic dermatitis tended to increase the risk of RSV hospitalization in the offspring. A history of atopic dermatitis in the father or siblings and parental allergic rhinitis were not associated with an increased risk of RSV hospitalization (data not shown).
|
|
3 wheezing episodes and a doctors diagnosis of wheezing (recurrent wheezing) before 6 months of age. Because this information was obtained from the third interview only, all of the infants without the third interview counted as missing. In this analysis, the age at the first episode of wheezing was set as late as possible to 6 months of age, and the risk of subsequent RSV hospitalization between 6 and 18 months of age was estimated. This analysis yielded an adjusted estimate of the RR of RSV hospitalization among infants with infrequent wheezing of 2.70 (95% CI: 1.87–3.91) between 6 and 12 months of age and 2.15 (95% CI: 1.37–3.39) between 12 and 18 months of age. Among infants with recurrent wheezing, these estimates were 12.56 (95% CI: 8.23–19.15) between 6 and 12 months of age and 6.99 (95% CI: 4.34–11.23) between 12 and 18 months of age (Table 2, stringent criteria). | DISCUSSION |
|---|
|
|
|---|
That the risk of severe RSV infection requiring hospitalization may be increased in infants and children with a genetically determined airway hypersensitivity and/or atopic disposition is further suggested by our observation of the strong association between maternal asthma and, to a lesser extent, paternal asthma and RSV hospitalization in the offspring. The difference in the stringency of the asthma definitions in the parents might partly explain the differences in the size of the associations with RSV hospitalization in the offspring found in the present study where maternal asthma was defined by current, physician-diagnosed asthma, whereas paternal asthma was any paternal asthma according to the mother. An asthmatic or atopic disposition has been shown to cause an increase in frequency of wheezing and nonwheezing LRI in the first 18 months of life14 and more severe RSV LRI.8 Recent studies show that maternal history of asthma increased the risk of LRIs in infancy9 and that infants who had impaired pulmonary function at 1 month of age were prone to develop later wheezing episodes and asthma.15 The in utero immunologic environment and early life exposures may confer additional susceptibility to what is genetically determined.16 These recent findings underline the plausibility of the observations in the present study. Hence, the results of the present study suggest that many infants who get severely ill with RSV were already disposed to wheeze.
For many children the third or fourth interview was missing. However, there was no difference in the rates of missing data in cases and controls. Furthermore, both case and control children were included in the study without regard to interviews. Finally, missing values were included as separate categories in each analysis, and none of these categories were associated with RSV hospitalization. The strengths of the study lie in the large overall and age-specific sample sizes, the nested case-control design, and our unique ability to study the time sequence between exposures and outcome.
In accordance with other studies on determinants for severe RSV infection,17–29 the present study found that male gender, medical risk factors, the presence of other children <12 years of age in the home, maternal smoking during pregnancy and lactation, and day care attendance were risk factors for RSV hospitalization, the strongest associations being for infants with siblings and for children with medical risk factors.
The examination by age groups revealed some interesting results. Thus, male gender was a risk factor only <6 months of age, the period of life with the presumable closest mother-child contact. We have earlier shown that cross-gender transmission from mother to son might at least partly explain this increased risk among the male infants.30 However, other studies have pointed at a relative smaller size,31,32 increased airway tone,33 and/or a greater lability34,35 of the airways in boys, thus increasing their risk of dyspnea when infected. The increased risk associated with the presence of siblings was also present only in infancy and was strongest among the youngest infants. The majority of Danish infants start day care attendance between 6 and 12 months of age, and this presumably diminishes the relative role of the family in the transmission of infections. As expected, breastfeeding was protective only in infancy, because few Danish children are breastfed >12 months. The seemingly protective effect of maternal alcohol consumption is likely because of social confounding. Thus, another study based on the DNBC showed that moderate alcohol consumption is associated with higher social class among Danish women.36 The risk of RSV hospitalization in infants of maternal smokers seemed to increase with age of the child, maybe because of ongoing exposure to tobacco smoke. It seems likely that mothers who smoked during pregnancy and lactation would be prone to continue smoking.
Several studies4–7,37–39 have reported an increased incidence of wheezing, recurrent wheezing, or asthma after RSV LRI or hospitalization in infancy and early childhood. However, the causal relationship between RSV and subsequent recurrent wheezing has not been established in children without underlying medical risk factors. Given the strength of the associations found in this study for disposition to atopic disorders and asthma and RSV hospitalization and for wheezing in the child and subsequent RSV hospitalization, it would require new study designs adjusting for atopic disposition and wheezing history to be able to document an independent effect of RSV LRI on the occurrence of subsequent wheezing. Such studies are warranted, because it is critical from a public health perspective to know whether the prevention of severe RSV will reduce the prevalence of subsequent wheezing. Our study strongly supports the hypothesis that a combination of host and environmental factors, operating in utero and during early infancy, interact to increase the risk for both hypersensitive airways and severe respiratory infection within the subject.
| ACKNOWLEDGMENTS |
|---|
The Danish RSV Data Network includes: Brita Bruun, Hillerød Sygehus; Henrik Westh, H:S Hvidovre Hospital; Henrik Friis, Slagelse Sygehus; Ole Heltberg, Centralsygehuset i Næstved; Allan Hornsleth, H:S Rigshospitalet; Åse Houmann, Thisted Sygehus; Jens Otto Jarløv, Herlev Sygehus; Inge Panum Jensen, Statens Serum Institut; Poul Kjældgaard, Sønderborg Sygehus; Steen Lomborg, Herning Centralsygehus; Kåre Mølbak, Statens Serum Institut; Jens Møller, Skejby Sygehus; Henrik C. Schønheyder, Aalborg Sygehus Syd; Curt Steen, Bornholm Centralsygehus; Steffen Strøbæk. Sydvestjysk Sygehus; and Birgitte Tønning, Viborg-Kjellerup Sygehus.
| FOOTNOTES |
|---|
Address correspondence to Lone Graff Stensballe, MD, PhD, Bandim Health Project, Danish Epidemiology Science Centre, Statens Serum Institut, Artillerivej 5, 2300 Copenhagen S, Denmark. E-mail: lgn{at}ssi.dk
Drs Stensballe, Kristensen, Simoes, Jensen, and Aaby planned and initiated the study. Dr Stensballe established the RSV database. Drs Stensballe and Jensen supervised the identification of cases and controls and the data withdrawal from the Danish National Birth Cohort. Dr Benn established the definition of atopic dermatitis. Drs Stensballe, Jensen, and Nielsen carried out the analyses. Dr Stensballe wrote the first draft of the article, and all of the authors contributed to the final version.
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||