Epidemiology and Risk Factors of Infection in Early Childhood

Home , Risk factor

Nadja Hawwa Vissing, MD, PhD, Bo Lund Chawes, MD, DMSc, EpidemiologyMorten Arendt Rasmussen, MSc, PhD, Hans andBisgaard, MD, Risk DMSc Factors of Infection in Early Childhood BACKGROUND: abstract

There is a large, unexplained variation in the frequency of childhood infections. METHODS: We described incidence and risk factors of infections in early childhood. Simple infections were captured during the first 3 years of life in the Copenhagen Prospective Studies on Asthma in Childhood 2000 birth cohort. Environmental exposures RESULTS: were analyzed by quasi-Poisson regression and sparse principal– component analysis. The 334 children experienced a median of 14 (range 2 43) infectious episodes at – ages 0 to 3 years. The overall rate of infections was associated with the number of children – 2 in the day care (adjusted incidence rate ratio [aIRR] 1.09 [1.2 1.16]) and the m per child in – the day care (aIRR 0.96 [0.92 0.99]). Upper respiratory infections were also associated with – 2 the number of children in the day care (aIRR 1.11 [1.03 1.20]) and the m per child in the – – day care (aIRR 0.95 [0.91 0.99]), whereas lower respiratory infections were associated with – – caesarean section (aIRR 1.49 [1.12 1.99]), maternal smoking (aIRR 1.66 [1.18 2.33]), older siblings (aIRR 1.54 [1.19 2.01]), and the age at entry to day care (aIRR 0.77 [0.65 0.91]). The sparse principal component analysis revealed a risk factor profile driven by tobacco exposure, social circumstances, and domestic pets, but could only be used to explain 8.4% CONCLUSIONS: of the infection burden. Children experienced around 14 infections during the first 3 years of life, but incidences varied greatly. Environmental exposures only explained a small fraction of the variation, suggesting host factors as major determinants of infectious burden.

WHAT’S KNOWN ON THIS SUBJECT: Children experience numerous infections during childhood Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of with a large and unexplained variation in individual Copenhagen, Copenhagen, Denmark susceptibility. Various environmental risk factors Dr Vissing handled the data, performed the descriptive statistics and the quasi-Poisson have been studied with inconsistent results, and regression analysis, and wrote the first draft of the manuscript; Dr Chawes contributed to the few researchers have comprehensively investigated analyses and interpretation of the data and critically reviewed and revised the manuscript; the entire exposome and its influence on childhood Dr Rasmussen performed the sparse principal component analysis and contributed considerably infections. to the analyses and interpretation of the data; Dr Bisgaard was the guarantor of the study as a whole, from conception and design to conduct of the study and acquisition of data, data analysis, WHAT THIS STUDY ADDS: We found that only a and interpretation of data. As the corresponding author, he had full access to the data and had minor fraction (8.4%) of the large variance in final responsibility for the decision to submit for publication; and all authors approved the final infection frequency between otherwise healthy manuscript as submitted and agree to be accountable for all aspects of the work. children can be explained by environmental risk DOI: https://doi.org/10.1542/peds.2017-0933 factors, suggesting that host factors are the major Accepted for publication Feb 28, 2018 determinants of infection susceptibility in early childhood. Address correspondence to Hans Bisgaard, MD, DMSc, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Ledreborg Alle 34, 2820 Gentofte, Denmark. E-mail: [email protected] PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright © 2018 by the American Academy of Pediatrics To cite: Vissing NH, Chawes BL, Rasmussen MA, et al. Epidemiology and Risk Factors of Infection in Early Childhood. Pediatrics. 2018;141(6):e20170933

Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018:e20170933 ARTICLE Children experience numerous particular focus on asthma, allergy, II. Lower respiratory tract simple infectious episodes, and eczema as clinical outcomes infections (LRTIs): pneumonia particularly1,2 in the first 3 years of along with a strong focus on 21 and bronchiolitis; life. ‍ Although such infections infections as explanatory variables. III. Gastrointestinal infections (GIs): are rarely fatal in industrialized Children born prematurely (<36 GI, diarrhea, and vomiting; and countries, they have a considerable weeks of gestation), with severe impact on childhood health, congenital malformation or lower IV. Isolated fever and other hospitalization rates, and quality respiratory infection during the infections. of life and are a sizeable economic first month of life were excluded. In We analyzed the overall incidence of burden to society because of health the first 3 years of life, the children infections (the 4 groups combined) 2000 care use, parental work absenteeism, attended the COPSAC research and the incidence rates of groups and secondary3,4 infections of parents clinic at 1 month of age and every I, II, and III separately. Group VI and siblings. ‍ 6 months thereafter for scheduled comprised a heterogeneous group investigations as well as for acute There is considerable variation in 21 of infections without respiratory care visits. The research clinic was the frequency of simple or gastrointestinal symptoms and de facto acting general practitioner infections between otherwise was not analyzed as a separate 2,3, 5 for the children. healthy children, ‍ ‍ but there is outcome. Details on diagnoses can be limited data on epidemiology and found in Supplemental Information The study was approved by the risk factors in early– childhood. Source(Supplemental Data Validation Table 5). Suspected risk factors include day Copenhagen Ethics Committee 1,6 8 (KF01-289/96) and the Danish Data care attendance,9,10 ‍‍ duration of breastfeeding, ‍ crowding in day Protection Agency (2008-41-1754). Source data validation was performed 11 3,12 Written and oral informed consent care, the– presence of siblings, ‍ to evaluate the quality23 of data as environmental tobacco exposure was obtained from both parents. previously published. 13 15 16 Infections Records (ETS), ‍ indoor air pollution, from the COPSAC2000 database were 17 ’ low 18socioeconomic status and male compared with information recorded sex. However, there is a lack of 19,20 by the children s general practitioner reproducibility between studies, Infection burden was captured and revealed ˃90% completeness which complicates evidence-based at the scheduled 6-monthly visit and no important influence from preventive strategies. Furthermore, from birth until age 3 years and at Risksocioeconomics Factors or concurrent asthma. the authors of most previous studies additional acute care visits, where only focus on single or few selected the COPSAC2000 pediatricians risk factors, despite the fact that interviewed the parents about A total of 84 environmental and many exposure variables are highly any illnesses, symptoms, duration, constitutional risk factors were correlated. medication, and vaccinations collected during the first 3 years Our objective with this study was sinceInternational the last visit. Classification Infections of life. These risk factors include to characterize the epidemiology of ofwere Diseases, classified 10th according Revision to information on demographics, third simple infections during the first 3 the 22 trimester exposures, maternal years of life in a longitudinal clinical ‍ and infections during pregnancy, birth cohort study with extensive stored in a designated database. newborn characteristics, neonatal assessment of early life exposures, When needed, the physician added biomarkers, postnatal exposures,– aiming to identify a risk factor additional clinical information. In day care attendance, diet, indoor21,24 29 profile associated with incidence of case of missed visits, the parents environment, and genetics ‍ ‍‍ infections. were interviewed at theInternational subsequent (SupplementalStatistics Table 6). METHODS Classificationvisit on infectious of Diseases, episodes. 10th For Revisionthe current study, the Study Cohort 22 First, we applied a quasi-Poisson ‍ diagnoses were retrieved regression model estimating and grouped as follows: unadjusted incidence rate ratios The Copenhagen Prospective Study (IRRs) for the following 18 variables on Asthma in Childhood 2000 I. Upper respiratory tract suspected to be associated18 with (COPSAC2000) is a longitudinal infections (URTIs): common infection burden: sex, birth birth cohort study of 411 children cold, tonsillitis, pharyngitis, otitis weight and length, mode30 of delivery born to asthmatic mothers with a media, and croup; (vaginal or caesarean), paternal Downloaded from www.aappublications.org/news by guest on September 29, 2021 2 VISSING et al – asthma, maternal13,14 smoking during3,5, 12 Inc, Cary, NC) version 9.3 and Median duration of all infectious pregnancy, ‍ older siblings, ‍ ‍ R version 2.12.0. The SPCA analyses episodes was 6 days (IQR 3 8), which 31 ’ – P cats or dogs in the home,17,18 were performed in MatLab version decreased with increasing age (IRR household income, the mother17s R2016b by the algorithm available34 on per year 0.94 [0.90 0.97]; = .001). occupation and level of education, http://models.life. ku.dk/sparsity.‍ The same inverse association between – 9,10 – P duration of breastfeeding, ‍ day disease duration and age was seen for 1,6 8,11 RESULTS – care attendance ‍‍ ‍ (age at URTI (IRR 0.96 [0.92 1.00]; = .057), P – Baseline Characteristics introduction,2 number of children, PLRTI (IRR 1.10 [0.87 0.94]; and m per child in the day care), = .02), and GI (IRR 0.84 [0.73 0.96]; ’ and ETS assessed by nicotine = .01), whereas the duration of – P concentration in the child32 s hair Children who missed 2 or more isolated fever increased with age at ages 1 and 3 years. These successive 6-monthly scheduled visits (IRR 1.10 [1.02 1.19]; = .02). variables were selected a priori were excluded from the analysis The median number of days with – on the basis of literature to reduce leaving 334 (81%) of the 411 children infection throughout the 3-year study the risk of multiple testing. The eligible for analysis. Children in the period was 94 days (IQR 64 132), quasi-Poisson model was used to study group were significantly more in which the majority of days were – account for overdispersion in the often born by caesarean delivery, caused by URTI (62 days [IQR data. Thereafter, we performed an exposed to a cat at home, attended 40Antibiotics97]) (Table 2). adjustedP quasi-Poisson regression more crowded day cares, and came analysis including risk factors with from families with higher household a value ˂.20, estimating adjusted incomes but were less exposed to Antibiotics were administered in incidence rate ratios (aIRRs). maternal smoking during pregnancy 24.9% of the infectious episodes In the adjusted analysis, some (Table 1). A total of 327 (98%) children (87.7% of LRTI episodes, 23.1% of variables were closely correlated were fully vaccinated according to URTI episodes, 12.5% of isolated (ie, smoking during pregnancy and the national immunization program fever episodes, and 10.2% of GI ETS, birth weight and birth length, Simple(Supplemental Infection Table Burden 7). episodes). The most frequently and socioeconomic variables), and, used drug was amoxicillin (59.4%) in these cases, only 1 variable was followed by penicillin (27.9%) (Fig 2, included for adjustment to avoid A total of 5009 infections were SupplementalQuasi-Poisson TableRegression 8). Risk collinearity (see Supplemental reported among the 334 children Factor Analysis Information for clarification). We (Fig 1A), yielding a median incidence Overall Rate of Infections performed a stratified analysis on the – rate of 14 infections per child basis of a diagnosis of asthma at any – (mean 15; range 2 43; interquartile time point before age 3 years. range (IQR) 10 18). Respiratory The only risk factors significantly Secondly, we applied an unsupervised tract infections were most frequent related to the overall incidence of with a median of 10 episodes per infections was crowding in day care, data-driven sparse principal33, 34 2 – component analysis (SPCA) ‍ to child, corresponding to 71% of all Pmeasured as m per child in the day explore common underlying patterns infections (9 episodes per child for care (aIRR per IQR: 0.96 [0.92 0.99]; from the entire set of variables from URTI and 1 episode per child for = .04), meaning that2 children in day the COPSAC2000 database without a LRTI) (Table 2). care centers with m per child in the lower quartile had 4% fewer infections priori hypotheses on association to With Figure 1B, we show the incidence than those in the upper quartile infections (ie, a total of 84 descriptive rates of infections at ages 0 to 3 years, (Table 3). Also, the total number of covariates) (Supplemental Table illustrating an incidence peak around children in the day care was associated 6). SPCA is a modification of the age 1 year with a slight decline in the – P with increased incidence of infections multivariate technique principal third year of life. Figure 1C reveals the (aIRR 1.09 [1.02 1.16]; = .01), meaning component analysis (PCA). seasonal variation in incidence rates that children attending day care with Further details can be found in the with respiratory tract infections being the number of children in the upper Supplemental Information. The SPCA more common in the winter, whereas patterns were used as input variables fever and gastroenteritis had less quartile experienced 9% more infections in a forward stepwise Poisson Durationseasonal variation. of Infections thanURTIs those in the lower quartile. regression for the prediction of incidence of infections. The quasi-Poisson regression analyses ‍Table 2 reveals the mean durations of The same associations were seen for were performed in SAS (SAS Institute, infectious episodes for each category. crowding in day care, in which the Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018 3 TABLE 1 Baseline Characteristics Study Population Excluded From Analysis Categorical: n (%); Missing Data Categorical: n (%); Missing Excluded Compared With the Numeric: Mean (IQR) Numeric: Mean (IQR) Data Study Populationa Children Boys (yes or no) 164 (49) 0 39 (51) 0 .81 Pregnancy and birth Birth wt, kg 3.51 (0.63) 0 3.52 (0.57) 0 .87 BMI at birth 12.8 (1.75) 0 12.8 (1.72) 0 .90 Caesarean delivery (yes or no) 78 (23) 0 7 (9) 0 P = .005 Early exposures Older children in household at 1 y (yes 138 (41) 0 14 (27) 26 .06 or no) Duration of breastfeeding, mo 0 16 P = .005 0–3 31 (9) 12 (20) 3–6 76 (23) 20 (33) >6 227 (67) 29 (48) Maternal smoking during pregnancy (yes 43 (13) 0 20 (26) 0 P = .004 or no) Nicotine level in hair at 1 y (ng/mg) 3.3 (2.1) 14 3.0 (4.4) 29 .73 Nicotine level in hair at 3 y (ng/mg) 2.0 (0.9) 24 2.2 (0.6) 65 .91 Day care Age at start, d 372 (186) 7 383 (189) 26 .74 No. children in day care 29 (41) 12 16.9 (11) 33 P = .002 Space per child, m2 13 (11) 14 9.7 (6.9) 34 .08 Indoor environment Cat at home at birth (yes or no) 44 (13) 3 16 (25) 12 P = .02 Dog at home at birth (yes or no) 49 (15) 3 5 (8) 12 .13 Socioeconomic variables Household annual income, 100 000 DKR 5.09 (2.4) 1 3.96 (2.0) 25 P < .001 Mother’s education 2 32 .70 College or lower level 199 (60) 26 (58) Medium length 89 (27) 11 (25) University 44 (13) 8 (18) Mother’s employment 112 (34) 0 29 .17 Nonprofessional 112 (34) 16 (33) Professional 156 (47) 18 (38) Student 37 (12) 5 (10) Unemployed 29 (9) 9 (19) Atopic disposition Father with asthma 60 (18) 9 8 (11) 3 .11 DKR, Danish kroner. a Calculated by χ2 test for categorical risk factors and student’s t test for numerical risk factors.

– P 2 m per child was inversely associated [1.19 2.01]; < .005), whereas an Children diagnosed with persistent – P older age at the introduction to day wheeze and/or asthma before age 3 with incidence of URTI (aIRR per IQR – 0.95 [0.91 1.00]; = .048), and the care wasP inversely associated with LRTI years had the same infection burden incidence (aIRR per IQR 0.77 [0.65 as the remaining children, except total number of children in the day – P – 0.91]; < .005). A subanalysis on the they had a higher number of LRTIs care was associated with increased P effect of the age of the youngest older (IRR 2.97 [2.36 3.75]; < .001). incidence of URTI (aIRR 1.11 [1.03 siblings revealed a tendency toward Excluding children with persistent 1.20]; = .01). We observed no other the sibling effect declining with the age wheeze and/or asthma yielded the significantLRTIs associations. of the youngestP older sibling; however, same risk factor profile for LRTIs this decline was not statistically (SupplementalSPCA Multiparametric Table 9). Risk Factor significant ( = .40). Male sex was Pattern Analysis Risk factors associated with incidence – P – P associated with borderline significance of LRTI were caesarean delivery (aIRR (aIRR 1.29 [1.00 1.68]; = .052). 1.49 [1.12 1.99]; = .01), maternal GIs – P The 84 environmental and smoking during pregnancy (aIRR constitutional covariates were 1.66 [1.18 2.23]; < .005), and the decomposed into 11 underlying latent presence of older siblings (aIRR 1.54 No significant associations were found. risk factor patterns called components. Downloaded from www.aappublications.org/news by guest on September 29, 2021 4 VISSING et al – P

1.15 [0.99 1.34]; = .06), meaning that higher levels of ETS increased risk of LRTI. Component 4 was significantly – P associated with LRTI (IRR per IQR 0.83 [0.70 0.98]; = .03). This component was primarily driven by breastfeeding, both the duration of solely breastfeeding and any breastfeeding. Component 8 was – Psignificantly associated with both LRTI (IRR per IQR 0.78 [0.66 0.92]; – P = .004) and GI (IRR per IQR 0.85 [0.74 0.98]; = .03). This component was primarily driven by furred pets and 17q21 polymorphisms. Excluding asthmatic children from the analysis did not alter the results. DISCUSSION

Main Findings

This prospective Danish birth cohort study revealed that otherwise healthy children experienced a median of 14 simple infectious – episodes throughout the first 3 years – of life (mean 15; IQR 10 18; range 2 43) with substantial variation in frequency between individuals. Only crowding in day care had a significant but modest influence on the overall incidence of infections FIGURE 1 Burden of infections. A, Distribution of burden of infections per child. B, Burden of infections over and incidence of URTI, whereas time. C, Seasonal variation. RTI, respiratory tract infection. LRTI incidence was associated with maternal smoking, caesarean delivery, older siblings, and early day care attendance. When including Supplemental Figure 3 reveals the key explain 8.4% of the variation in 84 environmental and constitutional covariates driving each component the data. Described in component covariates with an explorative SPCA and the percentage of the total 1 was a pattern primarily driven approach, we were only able to variance that each component accounts by measures of ETS, combined describe 8.4% of the large variance for. In total, the 11 components could with socioeconomics, indoor air in infection frequency. With these only be used to explain 44% of the pollution, and pets (Supplemental findings, we suggest that host factors overall data variation. Fig 3). Unexpectedly, lower levels are the major determinants of infection ‍Table 4 reveals the association of ETS were associated with Strengthssusceptibility and in Limitationsearly childhood. between the components and increased incidence of infections. Component 1 was also associated incidence of infections. Only – P component 1 was significantly with incidence of URTI (IRR per The strength of our study is the close associated with the overall incidence IQR 0.93 [0.86 0.99]; = .03), and longitudinal prospective clinical – P of infections (IRR per IQR 0.91 there was a trend toward an inverse surveillance at the COPSAC2000 [0.86 0.97]; = .003), but the association between component 1 clinical research unit that included component could only be used to and incidence of LRTI (IRR per IQR 3 years of follow-up and a total Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018 5

(1 – 5) (3 – 8) (3 – 8) (3 – 8) (3 – 8) (4 – 9) (4 – 9) (5 – 8) (4 – 8) (5 – 9) (1 – 5) (1 – 5) (2 – 5) (2 – 7) (2 – 8) (2 – 8) (1 – 6) observation of 365730 days. The (6 – 12) (6 – 12) (6 – 10) study is a single-center study with 6 monthly assessments conducted by experienced study pediatricians 4.2 6.9 8.2 7.5 7.4 6.8 6.5 8.2 7.9 9.8 9.8 8.3 7.4 3.7 3.7 4.0 4.4 6.0 4.4 6.0 who examined the children and obtained their clinical history that was supported by daily diary cards. Through this, data capture Duration of Infections, Days was secured and the risk of 2 6 7 6 6 6 5 7 7 8 8 8 7 3 3 3 3 4 3 4

Median Mean IQR misclassification was reduced. We ’ previously validated our data with

records from23 children s general practitioners and showed good data capture with sensitivity ˃90%. (0 – 2) (0 – 0) (0 – 7) (0 – 0) (0 – 3) (0 – 7) (0 – 7) (0 – 2) (0 – 4) IQR (8 – 31) (7 – 28) (0 – 15) (3 – 15) (0 – 12) (14 – 44) (20 – 53) (14 – 40) (10 – 40) (40 – 97) Highly detailed assessment of (64 – 132) the exposome is a key feature of the COPSAC2000 study, including information on a wide range of 8 2.4 3.2 4.0 2.8 2.3 4.5 4.0 2.5 3.1 24.1 21.3 32.2 40.5 30.5 28.9 10.1 74.3 10.9 environmental and constitutional 103.2 risk factors collected prospectively,

combining interview data21 with objective measurements. No. of Days With Infection Per Child 0 5 0 0 0 9 0 2 1 4 0 0 94 19 15 26 34 25 20 62 It is an advantage that, with our study, we combine traditional Median Mean inferential statistical methods with an unsupervised data-driven analysis. The quasi-Poisson regression analysis reveals if a single (0 – 1) (1 – 4) (1 – 4) (0 – 2) (2 – 6) (3 – 8) (2 – 6) (2 – 5) (0 – 0) (0 – 1) (0 – 0) (1 – 4) (0 – 1) (0 – 2) (0 – 1) (0 – 3) (0 – 1) (0 – 1) (6 – 12) (10 – 18) risk factor is significantly associated IQR with higher infection rate but does not reveal whether a pattern comprising several risk factors carries the relevant information. 0.6 3.0 2.8 1.2 4.3 5.9 4.7 9.4 3.6 0.3 0.5 0.4 2.7 0.6 1.1 0.9 1.7 0.4 0.7 15.0 Many exposure variables are correlated, and a multiparametric approach is therefore needed, such

Incidence of Infections Per Child as the SPCA, which can be used to handle an underlying correlation 0 3 2 1 4 5 4 9 3 0 0 0 2 0 1 1 1 0 0 14 structure. Furthermore, the SPCA Median Mean approach overcomes the issue of multiple testing in unidirectional analyses. Using this method, we were able to consider the high number 196 989 949 399 110 162 127 889 213 375 301 569 139 234 5009 1451 1985 1573 3152 1214 of exposure variables available in Total No. Episodes our cohort study without a priori selection of risk factors. It is a limitation that the selection of ’ collected variables is driven by the cohort s focus on atopic diseases, and Burden of Infections: Incidence and Duration Infectious Episodes in 334 Children a cohort study in which childhood – 3 – 1 – 2 – 3 – 3 – 1 – 2 – 3 – 3 – 1 – 2 – 3 – 3 – 1 – 2 – 3 – 3 – 1 – 2 2 – 3 0 0 1 2 0 0 1 2 0 0 1 2 0 0 1 2 0 0 1 infections are primarily addressed

All infections, y URTI, y LRTI, y

Fever (isolated), y

GI, y

TABLE 2 would possibly include other

variables. However, few2,3, 5,such11 studies have been conducted, ‍ ‍ ‍ and all Downloaded from www.aappublications.org/news by guest on September 29, 2021 6 VISSING et al 39,40 infections, particularly LRTI. ‍ We found a tendency toward increased incidence of LRTIs, but there was no apparent association between sex and other types of infections. ETS, estimated both by maternal ’ smoking during pregnancy and by nicotine level in the child s hair, was associated with increased frequency of LRTIs, and, although not statistically significant, the SPCA analysis supported the positive association

between ETS and incidence of LRTI,13 FIGURE 2 which aligns with other reports. Antibiotic treatment. The percentage of infections treated with antibiotics is shown. We previously showed that children exposed to tobacco smoke in utero

exhibit41, a42 deficit in lung function at birth, ‍ which could make them include less risk factors compared of respiratory tract infection more prone to LRTIs. There was with this study. especially2,7, 36 during the first years of no association between URTI and life, ‍ ‍ but the authors of studies The generalizability of our findings ETS in the univariate analysis, and, with longitudinal follow-up have can be questioned because of the high surprisingly, the SPCA pointed toward questioned whether the increased risk of asthma in the cohort. Children a protective effect of ETS, which symptom burden continues later at high risk of developing asthma and 8,37, 38 could be a false discovery. The lack in childhood. ‍ We confirmed a of association is in fact in line with wheezy symptoms may experience 2 modest influence from attendance to more infections than nonwheezing other prospective studies, supporting 35 crowded day care facilities, reflected children, which could lead to an the role of ETS as a trigger of lower as a 9% increase in overall incidence overestimation of population disease respiratory symptoms, such as cough of infections and an 11% increase incidence in this cohort. Still, we and wheeze, but not enhancing the in URTIs among children attending would not expect this to affect the susceptibility to infections per se. day care with crowding in the upper observed wide variation of infections quartile compared with the lower. Cesarean delivery was associated within the cohort or the influence Likewise, we saw a 4% reduction in with a 49% increased incidence of risk factors. Furthermore, a overall incidence of infections and of LRTI. The authors of a previous sensitivity analysis excluding a 5% reduction in URTIs per IQR epidemiologic birth cohort study children with persistent wheeze and/ conducted in Norway on the basis of available space per child in the or asthma yielded similar results. Interpretation day care. The effect of day care is of questionnaires filled out by presumably caused by an increased 37101 mothers were unable 30to disease transmission. find a significant association, but, although not statistically significant, With this unparalleled comprehensive Children with older siblings have all of their results pointed toward risk factor analysis, in which we been suggested to suffer from more an increased risk of recurrent LRTI use both traditional statistics and respiratory infections39 than first- within the first 3 years of life after SPCA pattern recognition analysis, born children, but it has also been caesarean delivery. The mechanism we confirm the lack of evident and proposed that children with older behind this association is unclear. reproducible associations between siblings experience12 an earlier immune It could be a reflection of other risk simple childhood infections and maturation and a subsequent factors influencing the likelihood of exposure-related risk factors. These improved resistance against infections a caesarean delivery, but it might findings reveal that host factors rather later in childhood. We found that be that children born by vaginal than the exposome account for the having siblings at home increased delivery are exposed to a diverse variation in frequency of infections in the risk of LRTIs by 54% but had no microbiological43 flora from the early childhood. influence on the incidence of overall birth canal and that subsequent infections, URTIs, or GIs. The authors of previous studies have colonization of gut and airways44 reported that day care attendance Boys are expected to have an alters immune modulatio45 n and is associated with increased risk increased risk of respiratory susceptibility to LRTI. Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018 7 P IRR GI P Crude Adjusted IRR 1.05 .720.771.07 1.01 .18 .59 .91 0.73 1.07 .11 .58 0.97 .59 0.95 .42 a a a a P .01 <.005 <.005 <.005 Adjusted a a a a IRR 0.781.00 .24 .98 1.08 .74 1.01 0.99 .85 .98 1.07 .27 1.49 1.66 1.54 0.77 LRTI a a a a a a P .04 .04 .01 <.005 <.005 <.005 Crude a a a a a a IRR 0.891.00 .24 .99 0.96 1.03 .66 .94 1.04 0.96 .61 .36 1.03 0.97 .76 .48 a a P .01 .048 a a IRR 1.11 0.95 URTI a a P .01 <.005 a a IRR 1.11 0.94 a a P .01 .04 a a IRR 1.09 0.96 Any a a P .01 <.005 Crude Adjusted Crude Adjusted a a IRR 1.07 .16 1.07 .20 1.05 .46 1.03 .61 1.27 .10 1.29 .052 0.96 .75 0.99 .93 0.99 .66 0.97 .27 1.02 .50 0.93 .17 0.89 .15 0.94 .46 0.91 .20 0.87 .07 1.04 .18 1.03 .37 1.07 .09 1.06 .13 0.91 .28 1.05 .68 1.02 .75 1.00 .97 0.891.01 .13 .89 0.92 1.01 .30 .73 0.88 0.97 .15 .66 0.93 0.98 .43 .69 1.94 1.50 1.06 .58 1.02 .88 1.08 .53 1.02 .89 0.82 .45 0.90 .64 1.22 .41 1.22 .41 0.93 .36 0.97 .50 0.98 .84 1.01 .88 0.94 .75 0.90 .60 1.00 .98 1.06 .73 0.97 .34 0.96 .16 1.00 .97 0.98 .49 0.77 0.99 .62 1.01 .59 0.99 .64 1.01 .69 1.12 .06 0.94 .39 1.04 .47 1.11 .08 1.07 .28 1.09 .18 1.04 .56 1.06 .40 1.36 1.02 .77 1.10 .33 1.10 .34 1.17 .16 1.40 .12 1.48 .10 1.11 .57 1.19 .37 1.05 .56 0.98 .83 1.13 .24 1.04 .86 0.64 0.94 .40 0.95 .50 1.00 .97 1.01 .87 0.98 .92 0.93 .71 0.80 .21 0.77 .16 0.98 .43 1.00 .96 0.97 .32 0.99 .82 1.19 0.90 .29 0.98 .80 0.92 .44 0.98 .84 1.38 .18 1.50 .09 0.98 .93 1.01 .98 1.03 .22 1.02 .61 1.02 .52 1.01 .80 0.98 .78 1.01 .88 1.01 .94 0.98 .81 1.04 .54 1.05 .50 1.00 .99 1.01 .94 1.15 .40 1.22 .23 1.10 .55 1.06 .69 1.07 .054 1.05 .20 1.06 .14 1.04 .39 0.86 .13 1.02 .85 0.99 .92 0.95 .52 0.95 .37 1.02 .74 0.99 .84 1.06 .48 1.26 .14 1.16 .41 0.93 .61 1.04 .80 1.09 0.95 – 3 mo) 000 DKR

th (yes or no) th (yes or no) th t ’ s education (as ordinal variable. step ’ s employment (reference = Association Between Risk Factors and Disease Incidence During the First 3 Years of Life

th wt, kg th length, cm per child 3 – 6 mo Student More than 6 mo Unemployed 1 – 3) Nonprofessional professional) 2 Boys (yes or no) Bir Bir Smoking during pregnancy (yes or no) Older children in household at 1 y (yes or no) Duration of breastfeeding (reference = 0 No. children in day care m Cat at home bir Nicotine level in hair at 1 y Age at star Caesarean delivery (yes or no) Dog at home bir Nicotine level in hair at 3 y Household annual income, 100 Father with asthma Mother Mother

Probands

Pregnancy and bir

Pre- and/or perinatal exposures

Indoor environment Day care

Socioeconomic variables

Atopic disposition

Indicates significant results. TABLE 3 Estimates for continuous variables are presented as IRR per IQR. DKR, Danish kroner; IQR, interquartile range. Estimates for continuous variables are presented as IRR per IQR. DKR, Danish kroner; IQR, interquartile a

Downloaded from www.aappublications.org/news by guest on September 29, 2021 8 VISSING et al TABLE 4 Results of the SPCA: Associations Between Components and Infections Component Explanatory All infections RTI LRTI URTI GI fraction, % IRR Per IQR (95% CI) 1 8.4 0.91 (0.86–0.97)a 0.95 (0.89–1.01) 1.15 (0.99–1.34) 0.93 (0.86–0.99)a 0.92 (0.80–1.07) 2 6.9 0.97 (0.91–1.05) 0.96 (0.90–1.04) 1.02 (0.84–1.24) 0.96 (0.88–1.04) 0.99 (0.84–1.17) 3 4.2 1.03 (0.96–1.10) 1.02 (0.95–1.09) 1.14 (0.95–1.37) 1.00 (0.93–1.09) 1.02 (0.87–1.20) 4 4.0 0.94 (0.88–1.00) 0.97 (0.91–1.04) 0.83 (0.70–0.98)a 0.99 (0.92–1.07) 0.87 (0.75–1.01) 5 3.7 1.02 (0.97–1.07) 1.03 (0.98–1.09) 0.91 (0.79–1.05) 1.05 (0.99–1.11) 0.97 (0.86–1.10) 6 3.6 1.02 (1.00–1.04) 1.02 (1.00–1.04) 1.02 (0.96–1.08) 1.02 (1.00–1.04) 1.01 (0.97–1.06) 7 3.2 1.02 (0.96–1.08) 1.02 (0.96–1.08) 1.18 (1.00–1.41) 1.00 (0.94–1.07) 1.03 (0.90–1.19) 8 2.9 0.97 (0.91–1.03) 0.98 (0.92–1.05) 0.78 (0.66–0.92)a 1.01 (0.94–1.09) 0.85 (0.74–0.98)a 9 2.5 1.00 (0.93–1.08) 1.04 (0.96–1.12) 1.14 (0.94–1.39) 1.02 (0.94–1.11) 0.95 (0.80–1.12) 10 2.5 1.05 (0.99–1.11) 1.06 (0.99–1.13) 1.12 (0.94–1.32) 1.05 (0.98–1.13) 1.10 (0.95–1.26) 11 2.4 1.07 (0.98–1.16) 1.06 (0.97–1.16) 1.11 (0.89–1.39) 1.05 (0.96–1.16) 0.95 (0.79–1.16) Total 42 — — — — — Estimates are presented as IRR per IQR of the component. IQR, interquartile range; —, not applicable. a Indicates significant associations.

Pets at home, namely dogs, have been reduce the incidence of childhood care, and LRTI was associated with found protective of respiratory31 tract infections in industrialized countries, day care attendance, ETS, caesarean infection in early childhood. We either through specific protection delivery, duration of breastfeeding, did not find univariate association or through nonspecific46 immune- and having older siblings at home. between pet exposure and incidence modulatory effects. However, these risk factors explained of infections, but, with the SPCA, we only a small fraction (8.4%) of the The most notable finding is that pointed toward an increased overall interindividual variation in incidence so few of so many suspected risk incidence of infections for children of infections, suggesting that host factors contributed substantially with furred pets in the home, both cat factors are the major determinants. to the variation in incidence of and dog. Furthermore, pet exposure ACKNOWLEDGMENTS infections, particularly the overall in combination with certain wheeze- incidence and incidence of URTI related genotypes (17q21) was related and GI. This lack of evident and to incidence of both LRTI and GI, but, reproducible associations between We thank the children and families in this setting, dog exposure increased 2000 simple childhood infections and of the COPSAC cohort study for the number of LRTIs and GIs, whereas various exposure-related risk factors all their support and commitment. cat exposure seemed to protect. The 5,19 is in agreement with other studies, ‍ We acknowledge and appreciate the lack of clear and reproducible results 2000 suggesting that unidentified host unique efforts of the COPSAC underlines the complexity of these factors are the major determinants research team. environmental exposures. of the highly variable infection ABBREVIATIONS The burden of infections could be burden in young children. Therefore, associated with vaccination status. authors of future studies should aIRR: adjusted incidence rate ratio However, vaccination rates were high search for alternative explanations COPSAC2000: Copenhagen in our study, and we were unable to of disease patterns presumably Prospective Studies address influence from vaccines. The through a systems biology omics on Asthma in children were vaccinated according approach including other areas such Childhood 2000 to the Danish immunization schedule as functional immunology, genetics, ETS: environmental tobacco at the time of the study (diphtheria- dietary patterns, microbiome, and Haemophilus influenzae exposure tetanus toxoids-pertussis-polio, inflammatory responses early in life. GI: gastrointestinal infection type b, CONCLUSIONS IQR: interquartile range and measles-mumps-rubella; IRR: incidence rate ratio Supplemental Table 7). Since Pneumococcus LRTI: lower respiratory tract then, vaccination for several other Children experienced a median of Meningococcus infection pathogens, including , 14 simple infections during the PCA: pr incipal component analysis , rotavirus, and first 3 years of life, with 71% being SPCA: sparse principal varicella-zoster virus are being respiratory infections. Individual component analysis implemented in many countries. It is variation in disease frequency URTI: upper respiratory tract a subject for authors of future studies for all infections and URTIs was infection to explore whether these vaccines associated with crowding in day Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018 9 FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. FUNDING: Funded by the Copenhagen Prospective Studies on Asthma in Childhood 2000, which is listed on www.copsac. com. The Lundbeck Foundation (grant R16-A1694), the Ministry of Health (grant 903516), the Danish Council for Strategic Research (grant 0603-00280B), and the Capital Region Research Foundation have provided core support to the Copenhagen Prospective Studies on Asthma in Childhood 2000 research center. POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

REFERENCES 1. Wald ER, Guerra N, Byers C. Frequency respiratory illness in infants. Am J 18. Kusel MM, de Klerk N, Holt PG, and severity of infections in day Epidemiol. 1998;147(9):863–870 Landau LI, Sly PD. Occurrence and care: three-year follow-up. J Pediatr. 10. Watkins CJ, Leeder SR, Corkhill RT. management of acute respiratory 1991;118(4, pt 1):509–514 The relationship between breast and illnesses in early childhood. J Paediatr Child Health. 2007;43(3):139 146 2. Grüber C, Keil T, Kulig M, Roll S, Wahn bottle feeding and respiratory illness – U, Wahn V; MAS-90 Study Group. in the first year of life. J Epidemiol 19. Selwyn BJ. The epidemiology of History of respiratory infections in Community Health. 1979;33(3):180–182 acute respiratory tract infection the first 12 yr among children from a in young children: comparison of 11. Koch A, Sørensen P, Homøe P, et al. birth cohort. Pediatr Allergy Immunol. Population-based study of acute findings from several developing 2008;19(6):505–512 respiratory infections in children, countries. Coordinated data group of 3. Koopman LP, Smit HA, Heijnen ML, Greenland. Emerg Infect Dis. BOSTID researchers. Rev Infect Dis. 1990;12(suppl 8):S870 S888 et al. Respiratory infections in infants: 2002;8(6):586–593 – interaction of parental allergy, child 12. Ball TM, Castro-Rodriguez JA, Griffith 20. Gudnason T, Hrafnkelsson B, Laxdal B, care, and siblings— the PIAMA study. KA, Holberg CJ, Martinez FD, Wright Kristinsson KG. Can risk factors for Pediatrics. 2001;108(4):943–948 AL. Siblings, day-care attendance, infectious illnesses in children at day care centres be identified? Scand J 4. Enserink R, Lugnér A, Suijkerbuijk A, and the risk of asthma and wheezing Bruijning-Verhagen P, Smit HA, van Pelt during childhood. N Engl J Med. Infect Dis. 2012;44(2):149–156 W. Gastrointestinal and respiratory 2000;343(8):538–543 21. Bisgaard H. The Copenhagen illness in children that do and do 13. Jones LL, Hashim A, McKeever T, Prospective Study on Asthma in not attend child day care centers: Cook DG, Britton J, Leonardi-Bee J. Childhood (COPSAC): design, rationale, a cost-of-illness study. PLoS One. Parental and household smoking and baseline data from a longitudinal 2014;9(8):e104940 and the increased risk of bronchitis, birth cohort study. Ann Allergy Asthma 5. von Linstow M-L, Holst KK, Larsen K, bronchiolitis and other lower Immunol. 2004;93(4):381–389 Koch A, Andersen PK, Høgh B. Acute respiratory infections in infancy: 22. WHO. International Classification of respiratory symptoms and general systematic review and meta-analysis. Diseases (ICD). Available at: www.who. illness during the first year of life: a Respir Res. 2011;12:5 int/classifications/icd/en/. Accessed population-based birth cohort study. 14. Fer gusson DM, Horwood LJ. Parental March 9, 2016 Pediatr Pulmonol. 2008;43(6): smoking and respiratory illness 23. Vissing NH, Jensen SM, Bisgaard 584–593 during early childhood: a six-year H. Validity of information on atopic 6. Caudri D, Wijga A, Scholtens S, longitudinal study. Pediatr Pulmonol. disease and other illness in young et al. Early daycare is associated with 1985;1(2):99–106 children reported by parents in a an increase in airway symptoms in 15. Nicolai A, Frassanito A, Nenna R, et al. prospective birth cohort study. BMC early childhood but is no protection Risk factors for virus-induced acute Med Res Methodol. 2012;12:160 against asthma or atopy at 8 respiratory tract infections in children 24. Chawes BLK, Bønnelykke K, Bisgaard years. Am J Respir Crit Care Med. younger than 3 years and recurrent H. Elevated eosinophil protein X in 2009;180(6):491–498 wheezing at 36 months follow-up urine from healthy neonates precedes 7. Nafstad P, Hagen JA, Oie L, Magnus after discharge. Pediatr Infect Dis J. development of atopy in the first 6 P, Jaakkola JJ. Day care centers 2017;36(2):179–183 years of life. Am J Respir Crit Care and respiratory health. Pediatrics. 16. Pandey MR, Boleij JS, Smith KR, Wafula Med. 2011;184(6):656–661 1999;103(4, pt 1):753 758 – EM. Indoor air pollution in developing 25. Bisgaard H, Bønnelykke K, Sleiman 8. Zutavern A, Rzehak P, Brockow I, countries and acute respiratory PMA, et al. Chromosome 17q21 gene et al; LISA Study Group. Day care in infection in children. Lancet. variants are associated with asthma relation to respiratory-tract and 1989;1(8635):427–429 and exacerbations but not atopy in gastrointestinal infections in a German early childhood. Am J Respir Crit Care 17. Thörn LKAM, Minamisava R, Nouer SS, birth cohort study. Acta Paediatr. Ribeiro LH, Andrade AL. Pneumonia Med. 2009;179(3):179–185 2007;96(10):1494 1499 – and poverty: a prospective population- 26. Raaschou-Nielsen O, Sørensen M, 9. Cushing AH, Samet JM, Lambert WE, based study among children in Brazil. Hertel O, et al. Predictors of indoor et al. Breastfeeding reduces risk of BMC Infect Dis. 2011;11:180 fine particulate matter in infants’

Downloaded from www.aappublications.org/news by guest on September 29, 2021 10 VISSING et al bedrooms in Denmark. Environ Res. 33. Witten DM, Tibshirani R, Hastie T. A a seven year follow up study. Thorax. 2011;111(1):87–93 penalized matrix decomposition, 1999;54(8):664–669 with applications to sparse 27. Vissing NH, Larsen JM, Rasmussen 40. Leeder SR, Corkhill R, Irwig LM, principal components and canonical MA, et al. Susceptibility to lower Holland WW, Colley JR. Influence of correlation analysis. Biostatistics. respiratory infections in childhood family factors on the incidence of 2009;10(3):515 534 is associated with perturbation of – lower respiratory illness during the the cytokine response to pathogenic 34. Rasmussen MA, Bro R. A tutorial first year of life. Br J Prev Soc Med. airway bacteria. Pediatr Infect Dis J. on the Lasso approach to sparse 1976;30(4):203–212 2016;35(5):561 566 modeling. Chemom Intell Lab Syst. – 41. Bisgaard H, Loland L, Holst KK, Pipper 2012;119:21 31 28. Giwercman C, Halkjaer LB, Jensen SM, – CB. Prenatal determinants of neonatal Bønnelykke K, Lauritzen L, Bisgaard H. 35. Svanes C, Jarvis D, Chinn S, Omenaas lung function in high-risk newborns. Increased risk of eczema but reduced E, Gulsvik A, Burney P; European J Allergy Clin Immunol. 2009;123(3): risk of early wheezy disorder from Community Respiratory Health 651–657.e4 exclusive breast-feeding in high- Survey. Early exposure to children risk infants. J Allergy Clin Immunol. in family and day care as related to 42. Hanrahan JP, Tager IB, Segal MR, et al. The effect of maternal smoking 2010;125(4):866–871 adult asthma and hay fever: results from the European Community during pregnancy on early infant 29. Sørensen M, Bisgaard H, Stage M, lung function. Am Rev Respir Dis. Loft S. Biomarkers of exposure to Respiratory Health Survey. Thorax. 1992;145(5):1129–1135 environmental tobacco smoke in 2002;57(11):945–950 infants. Biomarkers. 2007;12(1): 36. Schwar tz B, Giebink GS, Henderson 43. Stokholm J, Thorsen J, Chawes BL, et al. Cesarean section changes 38–46 FW, Reichler MR, Jereb J, Collet JP. neonatal gut colonization. J Allergy Clin 30. Magnus MC, H berg SE, Stigum H, Respiratory infections in day care. å Immunol. 2016;138(3):881–889.e2 et al. Delivery by cesarean section Pediatrics. 1994;94(6, pt 2):1018–1020 and early childhood respiratory 37. Côté SM, Petitclerc A, Raynault M-F, 44. Følsgaard NV, Schjørring S, Chawes BL, symptoms and disorders: the et al. Short- and long-term risk of et al. Pathogenic bacteria colonizing Norwegian mother and child infections as a function of group child the airways in asymptomatic neonates cohort study. Am J Epidemiol. care attendance: an 8-year population- stimulates topical inflammatory 2011;174(11):1275–1285 based study. Arch Pediatr Adolesc mediator release. Am J Respir Crit Care Med. 2013;187(6):589 595 31. Ber groth E, Remes S, Pekkanen Med. 2010;164(12):1132–1137 – J, Kauppila T, Büchele G, Keski- 38. Ball TM, Holberg CJ, Aldous MB, 45. Vissing NH, Chawes BLK, Bisgaard Nisula L. Respiratory tract illnesses Martinez FD, Wright AL. Influence of H. Increased risk of pneumonia during the first year of life: effect attendance at day care on the common and bronchiolitis after bacterial of dog and cat contacts. Pediatrics. cold from birth through 13 years colonization of the airways as 2012;130(2):211–220 of age. Arch Pediatr Adolesc Med. neonates. Am J Respir Crit Care Med. 2013;188(10):1246 1252 32. Al-Delaimy WK, Crane J, Woodward A. Is 2002;156(2):121–126 – the hair nicotine level a more accurate 39. Ponsonby A-L, Couper D, Dwyer T, 46. Pollard AJ, Finn A, Curtis N. Non- biomarker of environmental tobacco Carmichael A, Kemp A. Relationship specific effects of vaccines: plausible smoke exposure than urine cotinine? between early life respiratory and potentially important, but J Epidemiol Community Health. illness, family size over time, and the implications uncertain. Arch Dis Child. 2002;56(1):66–71 development of asthma and hay fever: 2017;102(11):1077–1081

Downloaded from www.aappublications.org/news by guest on September 29, 2021 PEDIATRICS Volume 141, number 6, June 2018 11 Epidemiology and Risk Factors of Infection in Early Childhood Nadja Hawwa Vissing, Bo Lund Chawes, Morten Arendt Rasmussen and Hans Bisgaard Pediatrics originally published online May 24, 2018;

Updated Information & including high resolution figures, can be found at: Services http://pediatrics.aappublications.org/content/early/2018/05/22/peds.2 017-0933 References This article cites 45 articles, 9 of which you can access for free at: http://pediatrics.aappublications.org/content/early/2018/05/22/peds.2 017-0933#BIBL Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Infectious Disease http://www.aappublications.org/cgi/collection/infectious_diseases_su b Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.aappublications.org/site/misc/Permissions.xhtml Reprints Information about ordering reprints can be found online: http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on September 29, 2021 Epidemiology and Risk Factors of Infection in Early Childhood Nadja Hawwa Vissing, Bo Lund Chawes, Morten Arendt Rasmussen and Hans Bisgaard Pediatrics originally published online May 24, 2018;

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://pediatrics.aappublications.org/content/early/2018/05/22/peds.2017-0933

Data Supplement at: http://pediatrics.aappublications.org/content/suppl/2018/05/23/peds.2017-0933.DCSupplemental

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. Pediatrics is owned, published, and trademarked by the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2018 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on September 29, 2021