Day care attendance, recurrent respiratory tract infections and asthma

Transcription

1 International Epidemiological Association 1999 Printed in Great Britain International Journal of Epidemiology 1999;28: Day care attendance, recurrent respiratory tract infections and asthma Wenche Nystad, Anders Skrondal and Per Magnus Objective Design Our objective was to use a causal model for childhood asthma to determine whether the effect of day care attendance on asthma was mediated by recurrent respiratory tract infections. A cross-sectional survey among 1447 children aged 6 16 years in Oslo. Their parents completed written questionnaires. A recursive logit model was used to estimate direct effects in terms of adjusted odds ratios (aor). Results Year of birth, number of siblings and length of maternal education were significantly associated with day care attendance. Attendance at day care increased the risk of early infections, aor = 1.8 ( ), and infections were associated with asthma, aor = 4.9 ( ). The crude association between day care and asthma was cor = 1.5 ( ), whereas the estimated direct effect was small and nonsignificant, aor = 1.2 ( ). The results may be influenced by overreporting of infections among parents of children with asthma. Conclusions Our results suggest that children who attend day care have an increased risk of asthma with early infections as a mediator of risk. Keywords Day care, early infections, asthma Accepted 25 February 1999 Children in day care centres may be exposed to an indoor environment that can induce asthma. Munir et al. have reported higher levels of cat allergens in dust from Swedish day care centres than in domestic homes. 1 Furthermore, it is known that children attending day care outside home are more likely to have infections than children in home care. 2,3 It has also been shown that airway infections in early life are associated with asthma and reduced lung function, 46 although it has also been suggested that such infections may have a protective effect on later asthma. 7,8 The relationship between day care attendance and the occurrence of asthma 9 and allergy has not been well described. 10,11 Due to the considerations outlined above, it is likely that day care attendance and occurrence of infections may influence the development of childhood asthma. The objective of the present study was to use a causal model for childhood asthma to estimate the association between day care attendance before 3 years of age and later asthma, taking into account the effect of infections as an intervening variable (Figure 1). Material and Methods Study population and data collection A cross-sectional study of asthma in school children, aged 6 16 years, using the standard questionnaire of the International Reprint requests to: W Nystad, Section of Epidemiology, Department of Population Health Sciences, National Institute of Public Health, Box 4404, Torshov, N-0403 Oslo, Norway. wenche.nystad@folkehelsa.no Study of Asthma and Allergies in Childhood (ISAAC), 12 was performed in Oslo in ,14 Parents of 2188 children responded to the questionnaire, and the response rate was 85% (2188/2577). 13,14 The sample received a second questionnaire, which was distributed one year later. Non-Norwegian children (immigrants) were excluded due to language problems, and about 400 children could not be identified. Thus 1844 parents received the second questionnaire, which included demographic information, family history of asthma and other atopic diseases and early exposures from birth throughout childhood. The response rate was 85% (1563/1841). A total of 116 children were excluded from the analysis due to incomplete data. Variables The asthma questions were based on the standard ISAAC and American Thoracic Society (ATS) questionnaires. 12,15 The specific exposures throughout childhood were based on retrospective parental reports. Exposures in the age periods of 0 11 months, 1 2 years, 3 4 years and 5 6 years were enquired about by means of an exposure matrix. Asthma Lifetime asthma was defined according to the ISAAC questionnaire: Has your child ever had asthma? 12 In addition, a question about age of onset was included according to the ATS questionnaire: At what age did his/her asthma begin? 15 The asthma diagnosis in the present analyses included only parental report of asthma which began after 3 years of age. 882

2 DAY CARE, INFECTIONS AND ASTHMA 883 Figure 1 A causal model for childhood asthma Day care attendance Day care attendance was scored dichotomously. Children attending full-time day care during both the two first-time periods (before 3 years of age) were compared to children in other types of day care (home care, part-time and family care) before 3 years of age. Furthermore, in order to estimate the effect of different types of day care, we divided day care before 3 years of age into four categories with home care as the reference category. The categories were: home care, other types of group care outside home (part-time and family care), full-time day care during the first or second time period and full-time day care during both first and second time period. Recurrent respiratory tract infections Parents were asked about different infections (otitis media, bronchitis, recurrent respiratory tract infections, pneumonia) diagnosed by a physician from birth throughout childhood. For our exposure variable, a strict definition of infections was employed, requiring parental retrospective report of physiciandiagnosed recurrent respiratory tract infections during both time periods (before 3 years of age). The variable included only recurrent respiratory tract infections and not otitis media, bronchitis and pneumonia. Background variables Dichotomous variables included gender, breastfeeding (no versus yes), lifetime occurrence of maternal and paternal asthma and smoking during pregnancy (no versus yes). Year of birth was classified into three groups ( , , ). The number of siblings was divided into four categories, 0, 1, 2, 3. Maternal and paternal education was scored according to three levels ( 10 years, years, 14 years). We also included birth weight and having pets (cat or dog) during the first year of life in the first analyses, but these two variables were not associated with asthma, and were not included in the final model presented here. Statistical methods Figure 1 represents a causal model of childhood asthma. If a variable represents an intermediate step between exposure and outcome it should not be treated as simply a confounder in logistic regression, but a more careful analysis taking its intermediate nature into account is required. Hence, we have employed recursive logit modelling 16,17 which has recently also been discussed within graph theory. 18 In contrast to conventional logistic regression, the recursive logit model enables us to investigate whether there are direct and/or indirect effects of the exposures on the outcomes. For instance, an exposure may have no direct causal effect on an outcome but instead an indirect effect via an intermediate variable in the causal pathway. The likelihood of recursive logit models decomposes into components with separate sets of parameters. Hence, estimation can proceed by analysing each component in turn using standard software for logistic regression. Specifically, the direct effects on an outcome variable are estimated as adjusted odds ratios (aor) in a logistic regression where all variables preceding the outcome in the causal pathway are included as predictors. In addition to the model-based estimates, crude odds ratios (cor) were also estimated to measure the unadjusted associations between the exposure variables and the outcomes. All analyses were performed using SPSS for Windows 6.0. Results The youngest children, born , and children without siblings were more likely to have been in full-time day care

3 884 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Table 1 The distributions of full-time day care before 3 years of age, respiratory tract infections (infections) before 3 years of age and asthma after 3 years of age according to other variables (in per cent) among 1447 school children in Oslo Day care Infections Asthma n Yes (%) Yes (%) Yes (%) Gender Female Male Year of birth Breastfeeding No Yes Maternal asthma No Yes Paternal asthma No Yes Smoking during pregnancy No Yes No. of siblings Paternal education (years) Maternal education (years) Full-time day care No Yes Recurrent respiratory tract infections No Yes (Table 1). Furthermore, children whose mothers had more than 14 years of education were more likely to attend day care than children whose mothers had less education. Attendance in full-time day care before 3 years of age was associated with increased relative risk of early infections (recurrent respiratory tract infections); cor = 1.8 (95% CI : ). Dividing day care into four categories with home care as the reference category revealed that the highest risk of infections (cor 2.9 [ ]) occurred for children attending full-time day care both the first and second time period. Children who attended full-time day care the first or second time period (cor 1.8 [ ]), and children in other types of group care outside home (part time and family care) had moderately elevated risks (cor 1.4 [ ]) compared to home care. A history of maternal and paternal asthma was associated with early infections (Table 1). Finally, there was a positive relationship between full-time day care attendance and asthma (cor 1.5 [ ]), and early infections and asthma (cor 5.6 [ ]). There was also a dose-response relationship between day care attendance and asthma. Children in full-time day care in both time periods had the highest risk of asthma (cor 1.9 [ ]). Children who attended full-time day care in the first or second time period (cor 1.3 [ ]) had moderately elevated risks, whereas children in other types of group care outside home (part time and family care) had no increased risk (cor 1.0 [ ]) compared to home care. The full model depicted in Figure 1 was analysed using recursive logit modelling, and the estimated direct effects are presented in Table 2. Each column in the table contains the direct effects of the predictors on the outcomes. The direct effects are estimated as aor, where it is adjusted for each variable listed with entries in the column. Following standard conventions of structural equation modelling, the model in Figure 2 includes only paths for variables, which were significant at the 5% level. Year of birth, number of siblings and length of maternal education had significant effects on day care attendance, and attending day care increased the risk of early infections. There was no direct effect between number of siblings and the occurrence of infections. Maternal and paternal asthma were, however, predictors of early infections and had strong direct effects on asthma. In addition, breastfeeding had a protective effect on asthma. The occurrence of early infections increased the risk of later asthma. The estimated direct effect of day care on developing asthma was small and non-significant when early infections was included as an intervening variable. However, there appeared to be an indirect effect of day care on asthma, via early infections. Discussion In this study we investigate full-time day care before 3 years of age in a causal model for childhood asthma. The results suggest that day care attendance is associated with the development of later asthma. The effect is explained by the increased occurrence of infections among children attending day care. The model is based on an explicit chronological ordering of the outcome variables. We specify that day care precedes infections, and infections precede asthma. For this sample this is reasonable to assume, because the children attended day care at an early age. Furthermore, it is unlikely that children with infections should attend day care more frequently than other children. In our analysis, we have also only included day care and infections before 3 years of age and asthma after 3 years to ensure that the exposure comes before the outcome. The present study has some limitations. Non-Norwegian children were excluded due to cultural differences and language problems. Some children could not be included in the second data collection due to decisions made by some of the liaison committees at the schools. However, this should not result in selection bias. Asthma was measured by a simple question

4 DAY CARE, INFECTIONS AND ASTHMA 885 Table 2 Direct effects (aor) with 95% confidence intervals for the recursive logit model Outcomes Day care before Infections before Asthma after 3 years of age 3 years of age 3 years of age aor (95% CI) aor (95% CI) aor (95% CI) Gender Female Male 0.9 ( ) 1.3 ( ) 1.0 ( ) Year of birth ( ) 0.9 ( ) 1.3 ( ) ( ) 1.0 ( ) 1.0 ( ) Breastfeeding Yes 1.0 ( ) 0.7 ( ) 0.5 ( ) Maternal asthma Yes 1.3 ( ) 2.5 ( ) 4.3 ( ) Paternal asthma Yes 1.1 ( ) 2.0 ( ) 5.6 ( ) Smoking during pregnancy Yes 1.4 ( ) 1.1 ( ) 1.1 ( ) No. of siblings ( ) 1.5 ( ) 0.8 ( ) ( ) 1.1 ( ) 0.6 ( ) ( ) 1.0 ( ) 1.0 ( ) Paternal education (years) ( ) 0.9 ( ) 0.7 ( ) ( ) 1.0 ( ) 0.7 ( ) Maternal education (years) ( ) 0.9 ( ) 1.5 ( ) ( ) 1.0 ( ) 1.5 ( ) Full-time day care No Yes 1.8 ( ) 1.2 ( ) Recurrent respiratory tract infections No 1.0 Yes 4.9 ( ) inquiring about the child ever having asthma and age of onset. Such a parental report of asthma may result in misclassification. The definition of asthma has, however, always presented problems, particularly for those involved in studying the disease in general populations. The main exposure variable, full-time day care, is such a major event, that it is not likely to be biased by misclassification. Concern arises, however, regarding the validity of retrospective parental reports of physician diagnoses of recurrent respiratory tract infections. The parental recall of infections may be incomplete and biased if parents of children with asthma were more likely to recall infections in their children. This kind of differential misclassification may lead to overestimation of the effect of infections and partly explain our main finding. In line with previous studies we have shown that day care attendance is associated with respiratory tract infections. 2,3,19,20 The occurrence of early respiratory tract infections may seem low but this may be due to the strict definition of infections. We have also detected a dose-response relationship between type of day care and infections. Our finding is supported by a study among Finnish children, which revealed that care in day care centres was a determinant of acute respiratory infections in children under 2 years old, whereas family care did not essentially increase the risk. 3 The present study does not address specific information of factors like indoor air quality, which may influence the occurrence of infections. Hours spent in day care and group size may, however, be contributing factors, which may partly explain the increased morbidity. Previous published studies, which estimated the association between day care attendance and asthma, 9 and the association between day care and atopy, 10,11 have neither described type of day care, group size or hours in day care. The present study adds information by defining day care attendance more precisely. Our finding may also have implications for investigations of the association between day care attendance and atopy. Our results agree with previous findings that respiratory tract illness increases the risk of asthma. 6,21,22 Most wheezing in pre-school children is associated with viral infections. 23 Both clinical experience and epidemiological studies suggest that respiratory infections sometimes result in airway obstruction diagnosed as asthma, while at other times they do not. 21,22,24 It is also known that some children who subsequently develop asthma in early childhood will have more severe symptoms during respiratory tract infections and are more prone to develop infections than non-asthmatic children. 22,25,26 Early attendance in full-time day care may thus contribute to the development of childhood asthma in these children. The results are not consistent with the infection hypothesis, which suggests that early infections may protect against later development of atopic disorders. 7,8 Asthma is, however, not synonymous with atopy, and the heterogeneity of the asthma cases may explain our results. More specifically, some children may have asthma due to symptoms of wheeze and chest tightness alone, and others may have airway narrowing in response to provoking stimuli such as allergen exposure. In addition, it might be that upper airway infections do not influence T-cell maturation and prevent the development of allergic sensitization. 8,27 Our findings may have implications for the interpretation of the large variability in the prevalence of asthma and respiratory symptoms reported by the ISAAC Steering Committee. 28,29 The causes for this variation, both within and between countries, are not well understood, and may reflect variation in morbidity and/or sources of bias. One possible explanation may be day care attendance, which may vary systematically between regions. The worldwide variations in reports of wheezing may also reflect both infections and asthma. Specifically in small children it is a problem to distinguish between asthma and other wheezing disorders with similar respiratory symptoms. 22,30 In conclusion, the results suggest that children who attend day care appear to have an increased risk of asthma with early infections as a mediator of risk.

5 886 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY Figure 2 The estimated causal model for childhood asthma. The model includes only paths which were significant Acknowledgements The contributions from Dr Jennifer Harris, Dr Kai-Håkon Carlsen, school children, their parents and teachers in Oslo are gratefully acknowledged. Funding was provided by the Norwegian Research Council. References 1 Munir AKM, Einarsson R, Dreborg SKG. Mite (Der p 1, Der f I), cat (Fel d 1) and dog (Can f 1) allergens in dust from Swedish day-care centers. Clin Exp Allergy 1995;25: Holberg CJ, Wright AL, Martinez FD, Morgan WJ, Taussig LM. Child day care, smoking by caregivers, and lower respiratory tract illness in the first 3 years of life. Pediatrics 1993;91: Louhiala PJ, Jaakkola N, Ruostsalainen R, Jaakkola JJK. Forms of day care and respiratory infections among Finnish children. Am J Public Health 1995;85: Samet JM, Tager I, Speizer F. The relation between respiratory illness in childhood and chronic air-flow obstruction in adulthood. Am Rev Respir Dis 1983;127: Shaheen SO, Barker JP, Shiell AW, Crocker FJ, Wield GA, Holgate ST. The relationship between pneumonia in early childhood and impaired lung function in late adult life. Am J Respir Crit Care Med 1994;149: Strope GL, Stewart PW, Henderson FW, Ivins SS, Stedman HC, Henry MM. Lung function in school-age children who had mild lower respiratory illnesses in early childhood. Am Rev Respir Dis 1991;144: Jarvis D, Chinn S, Luczynska C, Burney P. The association of family size with atopy and atopic disease. Clin Exp Allergy 1997;27: Cookson WOCM, Moffatt MF. Asthma: an epidemic in the absence of infection? Science 1997;275: Forsberg B, Pekkanen J, Clench-Aas J et al. Childhood asthma in four regions in Scandinavia: risk factors and avoidance effects. Int J Epidemiol 1997;26: Strachan DP, Harkins LS, Johnston IDA, Anderson HR. Childhood antecedents of allergic sensitization in young British adults. J Allergy Clin Immunol 1997;99: Forastiere F, Agabiti D, Corbo GM et al. Socioeconomic status, number of siblings, and respiratory infections in early life as determinants of atopy in children. Epidemiology 1997;8: Asher MI, Keil U, Anderson HR et al. International study of asthma and allergies in childhood (ISAAC): rationale and methods. Eur Respir J 1995;8: Nystad W, Magnus P, Røksund O, Svidal B, Hetlevik Ø. The prevalence of respiratory symptoms and asthma among school children in three different areas of Norway. Pediatr Allergy Immunol 1997;8: Nystad W, Magnus P, Gulsvik A, Skarpaas I, Carlsen KH. Changing prevalence of asthma in school children. Evidence for diagnostic changes in asthma in two surveys 13 years apart. Eur Respir J 1997; 10: British Medical Research Council. Recommended Respiratory Disease Questionnaires for Use with Adults and Children in Epidemiological Research. London: British Medical Research Council, 1978, pp.7 55.

6 DAY CARE, INFECTIONS AND ASTHMA Goodman LA. The analysis of multidimensional contingency tables when some variables are posterior to others: a modified path analysis approach. Biometrika 1973;60: Fienberg SE. The Analysis of Cross-classified Categorical Data. Cambridge, USA: MIT Press, Cox D, Wermuth N. Multivariate Dependencies Models, Analysis and Interpretation. London: Chapman & Hall, Nafstad P, Hagen JA, Botten G, Jaakkola JJK. Lower respiratory tract infections among Norwegian infants with siblings in day care. Am J Public Health 1996;86: Marbury MC, Maldonado G, Waller L. Lower respiratory illness, recurrent wheezing and day care attendance. Am J Respir Crit Care 1997;155: Sigurs N, Bjarnason R, Sigurbergsson F, Kjellman B, Bjørksten B. Asthma and Immunoglobulin E antibodies after respiratory syncytial virus bronchiolitis: a prospective cohort study with matched controls. Pediatrics 1997;95: Martinez FD, Wright AL, Taussig LM, Holberg CJ, Halonen M, Morgan MD. Asthma and wheezing in the first six years of life. N Engl J Med 1995;332: Wreight AL, Taussig LM, Ray G, Harrison RH, Holberg CJ. The Tuscon Children s Respiratory Study II. Lower respiratory tract illness in the first year of life. Am J Epidemiol 1989;129: Strachan DP. The prevalence and natural history of wheezing in early childhood. J Royal Coll Gen Prac 1985;35: Duff A, Pomeranz E, Gelber L et al. Risk factors for acute wheezing in infants and children: viruses, passive smoke, and IgE antibodies to inhalant allergens. Pediatrics 1993;92: Johnston S, Pattemore P, Sanderson G et al. Community study of role of viral infections in exacerbations of asthma in 9 11 year old children. Br Med J 1995;310: Martinez FD. The role of viral infections in the inception of asthma and allergies during childhood: could they be protective? Thorax 1997;49: The International Study of Asthma and Allergies in Childhood (ISAAC) Steering Committee. Worldwide variation in prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. Lancet 1998;351: The International Study of Asthma and Allergies in Childhood (ISAAC) Steering Committee. Worldwide variations in the prevalence of asthma symptoms: the International Study of Asthma and Allergies in Childhood (ISAAC). Eur Respir J 1998;12: Silverman M. Out of the mouths of babes and sucklings: lessons from early childhood asthma. Thorax 1993;48: