key: cord-0026251-q6wpi0v8
authors: Langer, Susan; Horn, Johannes; Gottschick, Cornelia; Klee, Bianca; Purschke, Oliver; Caputo, Mahrrouz; Dorendorf, Evelyn; Meyer-Schlinkmann, Kristin Maria; Raupach-Rosin, Heike; Karch, André; Rübsamen, Nicole; Aydogdu, Mustafa; Buhles, Matthias; Dressler, Frank; Eberl, Wolfgang; von Koch, Franz Edler; Frambach, Torsten; Franz, Heiko; Guthmann, Florian; Guzman, Carlos A.; Haase, Roland; Hansen, Gesine; Heselich, Valerie; Hübner, Johannes; Koch, Hans Georg; Oberhoff, Carsten; Riese, Peggy; Schild, Ralf; Seeger, Sven; Tchirikov, Michael; Trittel, Stephanie; von Kaisenberg, Constantin; Mikolajczyk, Rafael
title: Symptom Burden and Factors Associated with Acute Respiratory Infections in the First Two Years of Life—Results from the LoewenKIDS Cohort
date: 2022-01-05
journal: Microorganisms
DOI: 10.3390/microorganisms10010111
sha: 592099bce3bc23f15f4fe5ab6b649ad58b35ba02
doc_id: 26251
cord_uid: q6wpi0v8

Acute respiratory infections (ARIs) are the most common childhood illnesses worldwide whereby the reported frequency varies widely, often depending on type of assessment. Symptom diaries are a powerful tool to counteract possible under-reporting, particularly of milder infections, and thus offer the possibility to assess the full burden of ARIs. The following analyses are based on symptom diaries from participants of the German birth cohort study LoewenKIDS. Primary analyses included frequencies of ARIs and specific symptoms. Factors, which might be associated with an increased number of ARIs, were identified using the Poisson regression. A subsample of two hundred eighty-eight participants were included. On average, 13.7 ARIs (SD: 5.2 median: 14.0 IQR: 10–17) were reported in the first two years of life with an average duration of 11 days per episode (SD: 5.8, median: 9.7, IQR: 7–14). The median age for the first ARI episode was 91 days (IQR: 57–128, mean: 107, SD: 84.5). Childcare attendance and having siblings were associated with an increased frequency of ARIs, while exclusive breastfeeding for the first three months was associated with less ARIs, compared to exclusive breastfeeding for a longer period. This study provides detailed insight into the symptom burden of ARIs in German infants.

Acute respiratory infections (ARIs) continue to be the most common health problem during childhood worldwide. Although most ARIs are not severe [1] , they contribute to a high number of outpatient visits [2] , antibiotic prescriptions, hospitalizations [2, 3] , as well as to socioeconomic burden [4, 5] and absenteeism in education and work [6] . In addition, infections with respiratory viruses (e.g., human rhinovirus, enterovirus, and adenovirus) in early childhood can influence the development of chronic and immunemediated diseases such as asthma, type II diabetes, and obesity later in life [7, 8] . Respiratory infections often heal spontaneously, and, in more than 50% of the cases, there is no doctor's consultation required, and, in even less cases, hospitalization is involved [4, 5] . Therefore, it is impossible to determine the true frequency and burden of ARIs based on medical reports or hospital-based studies [9, 10] . Individual information on frequency of infections can indicate particularly high susceptibility to infections, and initiate further assessment. For this purpose, contemporary norms are necessary.

Previous observational studies [11] [12] [13] [14] used different assessment methods to determine the frequency of ARI episodes, often including a retrospective assessment. However, retrospective methods may result in under-reporting and recall problems [14] , if not only a short time period is considered [15] . It is therefore highly relevant to assess the frequency of ARIs with a real-time approach, such as daily entries into a symptom diary. Symptom diaries are an excellent method to counteract under-recording and allow a detailed description of the burden of disease. Different studies found between three to seven ARIs per year in early childhood (children up to two years old) [16] [17] [18] [19] . In Germany, frequencies of ARIs were last reported for children born in 1990 [1] . Here, only three episodes per year were recorded for children in the first year of life. Since then, no publication was published in Germany which estimated the frequencies of ARIs in children.

There are several factors which might influence the frequency of ARIs in the first two years of life. Previous studies already found out, that older age (compared to the first six months), cold seasons, childcare attendance [18, 19] , having older siblings, maternal smoking [20] , and male sex [21] are associated with a higher number of ARIs, while full breastfeeding [22] is associated with a lower frequency of ARIs. With societal changes, the role of these factors might be changing. Therefore, we investigated the frequency, the full burden of symptoms, as well as factors associated with ARIs in the first two years of life based on symptom diary data of the German population based on the prospective birth cohort study LoewenKIDS.

A detailed description of the study design, methods of recruitment, and data collection is provided elsewhere [14] . Briefly, the LoewenKIDS-study is an ongoing population-based observational birth cohort study, which recruited 782 newborns between November 2014 and February 2018 in five study regions in Germany (Clinicaltrials.Gov Identifier: NCT02654210 (Accessed on: 1 January 2021)). Participants were recruited antenatal and postpartum until the age of three months and are followed up until the age of 15 years. In 2020, all study participants were two years old or older.

Parents were invited to keep a daily symptom diary in the first six years of life of their child. They recorded all the child's symptoms, symptom-free days, doctor consultations, diagnoses, medication, and absence from work or childcare on a daily basis. Participants could choose between a paper-based diary, an online version, or an app. Changes between the different modes were allowed. Symptoms such as fever, wheezing, chills, sore throat, runny/congested nose, increased need to sleep, and increased attachment were included in the symptom diary, as well as severity of the aforementioned symptoms. The symptom diary was developed on the basis of the symptom diary used by the birth cohort ORChID [15] and adapted after a feasibility study [16] .

Parents filled in questionnaires at the birth of their child and at the age of six months, one year, and then annually until the age of 15 years. Questionnaires contain information on social and health characteristics, pregnancy, and birth, as well as on selected diseases and environmental factors.

We adapted the ARI definition proposed by Lambert et al. [23] [24] [25] . We classified ARIs by distinguishing between A-and B-symptoms. An A-symptom was defined as fever, wheezing, wet cough, and doctor diagnosed pneumonia or otitis media, whereby B-symptoms included dry cough, chills, sore throat, runny or blocked nose, increased need to sleep, loss of appetite, and increased attachment. We defined the beginning of an ARI episode as the occurrence of at least one A-symptom or a day with two B-symptoms. If there were no symptoms for three consecutive days, the episode ended and a new episode could begin. The occurrence of single/isolated B-symptoms were considered within an episode but not as the start of an episode.

Data analysis was performed using R, v. 4.0.5 for Windows. Descriptive analysis included calculating frequencies and duration of ARI episodes by age, sex, and seasonality. Classification into A-and B-symptoms, as well the generation of acute respiratory episodes and the calculation of outcome variables were carried in the R-package lkstaR [26] . Summary statistics are presented as mean (standard deviation, SD) or median (interquartile range, IQR) for continuous variables and frequency (percentage) for categorical variables. We compared different strata according to the number of ARIs using t-Test.

ARI frequencies and associations between participant characteristics were estimated using the Poisson regression. Multivariable analysis included duration of exclusive breastfeeding, time of entry in daycare attendance, type of delivery, birth term, sex, and having older siblings. Multivariable models included all the above-mentioned associated factors. Effect estimates and their corresponding 95% confidence intervals (95% CI) are presented. This analysis is based on data collected from 2014 to February 2020.

The parents of all children participating in the study provided informed written consent. The respective Ethics Committees of the Martin-Luther-University Halle-Wittenberg, Medizinische Hochschule Hannover and Ludwig-Maximilians-Universität Munich, Germany approved the research protocol.

Out of the 782 enrolled children in the LoewenKIDS study, the parents of 732 (93.6%) participants submitted daily symptom diaries. The parents of 433 (55.4%) participants provided entries for 80% of days, however, in order not to miss any potential infection events, we restricted the sample for this analysis to 288 participants (37%), who completed symptom diary on 98% of the days during the first two years of life. The present sample of 288 participants does not differ much in terms of sociodemographic factors from the 732 participants in the overall sample. Characteristics of the study population show that 85% were born at term, 70% spontaneously, 48% were male, 30% had one or more siblings, 85% attended daycare, and 65% were exclusively breastfed for at least four until six months (Table 1) . 

In total, 206,001 child-days with diary entries were available for analysis of the included participants. Observed symptoms included cough, wheeze, sore throat, chills, fever, attachment, high need for sleep, loss of appetite, and runny or blocked nose. One or more of these symptoms were reported on 44,441 days (21.6%), corresponding to a mean of 154.3 (IQR: 76.2-216) days with ARI symptom per child ( Figure 1A ). Symptoms occurred in the first six months of life on average for 19.4 days, at 7-12 months for 41.8 days, at 13-18 months for 49.5 days, and at 19-24 months for 43.4 days ( Figure 1B , Table 2 ). 

In the next step, we aggregated the reported symptoms to ARI episodes based on the applied definition. Among the 288 children, a total of 3911 ARIs were reported in the first two years of life ( Figure 2A ). On average, 13.7 ARI episodes (IQR: 10-17, SD: 5.2, 10th percentile: 7 ARIs, 90th percentile: 20 ARI)) were reported in the first two years of life ( Figure 2A) . The cumulative distribution of ARI frequency shows that about 25% of children have less than 10 ARI episodes and 25% show more than 17 ARI episodes in the first two years of life independent of sex ( Figure 2B ). The median age at first ARI episode was 91 days (IQR: 57-128, mean: 107, SD: 84.5) after birth. The mean duration of ARIs was 11 days (SD: 5.8, median 9.7, IQR: 7-14). The proportion of children with ARIs at a given day increased markedly with age ( Figure 2C ). The frequency of ARI episodes in the first year Figure 1C shows the cumulative distribution of days with specific symptoms in the first two years of life in percentiles. The most common symptoms were runny or blocked nose with an average of 125 days (median: 115.5, IQR: 58-176) and cough in various forms with 76.8 days (median: 60.5; IQR: 30-115). In contrast, rare symptoms such as chills occurred on average 0.5 days (median: 0, IQR: 0), sore throat 3.2 days (median: 0; IQR: 0-3), and wheezing 11.7 days (median: 3.5; IQR: 0-13) on average (Table 2, Figure 1C ).

In the next step, we aggregated the reported symptoms to ARI episodes based on the applied definition. Among the 288 children, a total of 3911 ARIs were reported in the first two years of life (Figure 2A ). On average, 13.7 ARI episodes (IQR: 10-17, SD: 5.2, 10th percentile: 7 ARIs, 90th percentile: 20 ARI)) were reported in the first two years of life (Figure 2A) . The cumulative distribution of ARI frequency shows that about 25% of children have less than 10 ARI episodes and 25% show more than 17 ARI episodes in the first two years of life independent of sex ( Figure 2B ). The median age at first ARI episode was 91 days (IQR: 57-128, mean: 107, SD: 84.5) after birth. The mean duration of ARIs was 11 days (SD: 5.8, median 9.7, IQR: [7] [8] [9] [10] [11] [12] [13] [14] . The proportion of children with ARIs at a given day increased markedly with age ( Figure 2C ). The frequency of ARI episodes in the first year was slightly lower with a mean of 6.0 ARI episodes compared to the second year with a mean of 7.7 ARI episodes ( Table 3) .

Microorganisms 2022, 10, x FOR PEER REVIEW 7 of 13 was slightly lower with a mean of 6.0 ARI episodes compared to the second year with a mean of 7.7 ARI episodes ( Table 3) . ARI episodes were more common in the winter months showing a well-known seasonal variation of respiratory tract infections in the northern hemisphere ( Figure 2D 

The frequency of ARIs strongly depends on age and seasonality ( Figure 2C,D) . With increasing age, participants show a marked increase in ARI frequency in the first six months of age ( Figure 2C ). Furthermore, an increased ARI frequency was observed in winter months compared to the summer months ( Figure 2D ). In addition, the results from the multivariable analysis in Table 4 show that factors associated with a substantially in- ARI episodes were more common in the winter months showing a well-known seasonal variation of respiratory tract infections in the northern hemisphere ( Figure 2D ).

The frequency of ARIs strongly depends on age and seasonality ( Figure 2C,D) . With increasing age, participants show a marked increase in ARI frequency in the first six months of age ( Figure 2C ). Furthermore, an increased ARI frequency was observed in winter months compared to the summer months ( Figure 2D ). In addition, the results from the multivariable analysis in Table 4 show that factors associated with a substantially increased risk of ARIs in the first two years of life are any childcare attendance and having any number of older siblings. However, neither the time point of first childcare attendance nor the exact number of siblings seem to be important for the cumulative number of infection episodes at the age of two. In contrast, the analysis shows that short-term exclusive breastfeeding (less than four months) is associated with a lower risk of ARIs 0.78 [95% CI 0.69; 0.89] within the first two years compared to exclusive breastfeeding for four to six months (Table 4 ). Children with a longer exclusive breastfeeding of more than six months had the same risk as those with the reference group of four to six months of exclusive breastfeeding. We did not observe any association between birth mode, birth term, or sex of the child. 

In the LoewenKIDS birth cohort study, we found that children show an average of 13.7 ARI episodes (first year 6.0 ARIs, second year 7.7 ARIs) with a median duration of 11 days in the first two years of life. ARIs increase with age and occur more frequently during the winter months compared to the summer months. Within the 13-to 18-months lifespan, children in our cohort showed the highest frequency of ARIs, days with symptoms, and occurrence of specific symptoms, such as runny nose or cough during the first two years of life. Attendance at daycare and the presence of siblings in the same household were associated with an increased risk for a higher frequency of ARIs, while exclusive short-term breastfeeding (less than four months) was associated with less ARIs compared to exclusive breastfeeding for four to six months.

The use of symptom diaries to study ARIs is rarely reported. The frequency varies widely between studies, ranging from three to seven ARIs per year in early childhood. Our results are similar to birth cohort studies in Australia [16, 18] , Scandinavia [27, 28] , and Canada [29] . However, some studies report lower frequencies such as the Perth study with 4.0 ARIs [17]/4.2 ARIs [5] , the Dutch Whistler study with 4.2 [19] , or the German Mas-90 Study [1] with the lowest of 3.1 ARIs in the first year of life. It should be noted that all studies took place at different times and under different conditions.

To our knowledge, there is only one study about the frequency of ARIs in early childhood in Germany using symptom diaries, which was published 30 years ago [1] . The Mas-90 study examined children born in 1990 and found an ARI frequency of 3.1 in the first year of life and 3.2 in the second year, which are considerably lower than our results. This is probably because symptoms were recorded only in a kind of symptom diary, rather retrospectively (personal consultation), and additionally, half of the participants with incomplete data were included in the analysis [1] . Therefore, underestimation may be possible. In addition, the number of daycare places for children under three years has increased considerably in Germany since 1990 [30] . It is well known that children attending daycare centers have a higher risk of ARIs [13, [18] [19] [20] 31, 32] , which is in line with the findings of this study. Children in our cohort who entered daycare at the age of 13-26 months show a 1.26-fold risk of developing ARIs (RR: 1.26; CI [1.15; 1.39]) compared with children who did not attend daycare until the age of two years. A cohort study from Pittsburgh [33] as early as 1990 showed that the risk of infection increases with the number of contacts in different care settings. They showed that children in childcare with a group size of at least seven children as well as children with a care time of at least 20 h per week contracted considerably more infections than children who were in home care or in group care with three to six children. Similarly, children having siblings are more likely to have an increased number of infections, which was shown in our cohort as well as in other studies before [1, 16, 17, 20, 34] . We could not find any relationship when considering sex, birth mode, or birth term, and we were unable to measure an association between smoking exposure and ARI frequencies because an insufficient number of parents in our cohort reported smoking.

Our cohort showed that short-term exclusive breastfeeding (less than four months) without other nutritional support is associated with a lower risk for ARIs, 0.78 (95% CI 0.69; 0.89), within the first two years compared to exclusive breastfeeding for four to six months. Children with longer exclusive breastfeeding more than six months had the same risk as those with the reference group four to six months of exclusive breastfeeding. There are some studies which showed an association between breastfeeding and a lower risk of ARIs compared to no breastfeeding in early life [20, 22, 35, 36] . However, the results are inconsistent. Cushing et al. showed a protective effect only for lower respiratory infections and no association with upper respiratory infections [22] . Frank et al. showed a protective effect between exclusive breastfeeding in the age period of three to six months compared with no breastfeeding only for ear infections and ARIs with fever and no association for ARIs in general [36] . In contrast, Wright et al. [35] also showed a protective effect of breastfeeding and a reduction in upper respiratory tract infections (URTIs), but only in the first four months of life, and Vissing et al. also showed inconclusive results [20] . On the other hand, there are studies with no evidence for a positive association [37] or even with negative effects for long breastfeeding [1, 16, 36] . The authors [1, 16] assume that the negative effects are due to possible selection effects by, for example, over-reporting of episodes in participating families with higher social status, longer breastfeeding and less smoking.

A birth cohort study from Copenhagen [27] indicated that children in the first year of life had a runny nose an average of 16% of days, 7.9% cough, 4.9% attachment, and 1.5% wheezing. These findings are in line with our results. In comparison to the study from Copenhagen, in our cohort there were slightly more days with runny or blocked nose (17% vs. 16% of days) and a little more cough (10% vs. 7.9% of days). In the younger age strata, we detected only a few days with sore throat and chills. Children at this age cannot yet report sore throat, and the occurrence of chills is certainly also very rare, so we did not expect high frequencies in these age strata in our cohort either.

In 2018, Sarna et al. estimated the average age of the first infection in life at a median of 2.9 months [9] . This is consistent with our estimate of a median of 91 days after birth.

There are many different ARI definitions that have been used in recent research to analyze recorded symptoms [5, 16, 18, 25, 27, 38] , so there is a legitimate question of whether our definition leads to more overestimation or even to underestimation of ARI episodes based on the choice of definition. We adapted the ARI definition proposed by Lambert et al. [25] , which in direct comparison of different definitions provided middle estimates in a past study [24] .

The strengths of this study were the prospective birth cohort study design and the presence of detailed diary data of respiratory symptoms in the first two years of life. Compared with retrospective data, symptom diaries provide more valid data, i.e., higher reporting and incidence rates, thereby mitigating recall bias [39] [40] [41] . However, the data collection is very time-consuming and challenging. Parents in our cohort were required to keep entries for each day, even if the child was asymptomatic.

It is well known that comparison of ARI episodes is very difficult when different recording methods (retrospective or by physician consultations) have been used. In addition, the studies were conducted in different regions with different environmental factors. However, it must also be said that a comparison is also difficult even with the same recording method if different definitions are used for the identification of an ARI episode based on the symptoms. This is shown by Zoch et al. [24] for six definitions for the identification of an ARI in a single dataset.

Completing a daily symptom diary can be a burden for participants and can affect compliance [42, 43] and also leads to tiredness [44] . This is likely the reason why in our study only a subsample of participants submitted complete diaries. We also observed some dropout (11% dropped out of the study in the first two years and even some more stopped recording symptoms). In addition, the symptom diary as a study component likely kept many people from participation.

This study provides up-to-date, detailed data on the incidence of respiratory diseases in the first two years of life of German children and shows the effects of increasing age, seasonality, daycare attendance, breastfeeding, and the presence of siblings. This study provides pediatricians and researchers with information on the range of infection frequency in generally healthy children. It can be considered as a guideline for the normal occurrence of ARIs in the 21st century. These results show a previously undescribed high frequency and high burden of acute respiratory disease in German children in the first two years of life, which consequently can also represent a great burden for parents and should therefore receive more public attention in this phase. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the participants to publish this paper.

Most of the quantitative results are provided in the tables. Distributions and individual level data in anonymized form can be obtained upon request and from the R-package lkstaR [26] .

History of respiratory infections in the first 12 yr among children from a birth cohort

Duration of symptoms of respiratory tract infections in children: Systematic review

Antibiotic use in the first year of life and risk of atopic disease in early childhood

The Economic Burden of Non-Influenza-Related Viral Respiratory Tract Infection in the United States

Occurrence and management of acute respiratory illnesses in early childhood

Respiratory Tract Infections in Children in the Community: Prospective Online Inception Cohort Study

The role of respiratory tract infections and the microbiome in the development of asthma: A narrative review

Genetics and epidemiology: Asthma and infection

Timing of First Respiratory Virus Detections in Infants: A Community-Based Birth Cohort Study

The Underrecognized Burden of Influenza in Young Children

Cohort profile: 'children of 1997 : A Hong Kong Chinese birth cohort

Rationale and methods

Early daycare is associated with an increase in airway symptoms in early childhood but is no protection against asthma or atopy at 8 years

The Tecumseh study of respiratory illness. I. Plan of study and observations on syndromes of acute respiratory disease

Incidence and comparison of retrospective and prospective data on respiratory and gastrointestinal infections in German households

A Prospective Study of Proneness to Acute Respiratory Illness in the First Two Years of Life

Role of respiratory viruses in acute upper and lower respiratory tract illness in the first year of life: A birth cohort study

The burden of community-managed acute respiratory infections in the first 2-years of life

Bruijning-Verhagen, P. Impact of early daycare on healthcare resource use related to upper respiratory tract infections during childhood: Prospective WHISTLER cohort study

Epidemiology and Risk Factors of Infection in Early Childhood

Prospectively assessed incidence, severity, and determinants of respiratory symptoms in the first year of life

Breastfeeding Reduces Risk of Respiratory Illness in Infants

Community epidemiology of human metapneumovirus, human coronavirus NL63, and other respiratory viruses in healthy preschool-aged children using parent-collected specimens

Effect of Disease Definition on Perceived Burden of Acute Respiratory Infections in Children: A Prospective Cohort Study Based on Symptom Diaries

Respiratory illness during winter: A cohort study of urban children from temperate Australia

lkstaR: An R-Package to Analyse the Loewenkids Symptom Diary (v0.0.2-beta)

Acute respiratory symptoms and general illness during the first year of life: A population-based birth cohort study

Acute respiratory infections in early childhood and risk of asthma at age 7 years

Residential fungal growth and incidence of acute respiratory illness during the first two years of life

Bundeszentrale für Politische Bildung (bpb)

Infectious disease in pediatric out-of-home child care

Daycare attendance and respiratory tract infections: A prospective birth cohort study

Frequency and severity of infections in day care: Three-year follow-up

A community study of clinical traits and risk factors for human metapneumovirus and respiratory syncytial virus infection during the first year of life

Breast feeding and lower respiratory tract illness in the first year of life. Group Health Medical Associates

The relationship between breastfeeding and reported respiratory and gastrointestinal infection rates in young children

Predominance of rhinovirus in the nose of symptomatic and asymptomatic infants

Nitrogen dioxide and respiratory illnesses in infants

Health diaries

Health diaries: A method of collecting health information

Prospective daily diary study reporting of any and all symptoms in healthy adults in Pakistan: Prevalence and response

Patient non-compliance with paper diaries

Diary keeping in asthma: Comparison of written and electronic methods

Best practice in symptom assessment: A review

We sincerely thank all participants and their parents for participating in this study. We would like to thank our study assistants of the LoewenKIDS study at MLU in Halle, especially Mareike Kunze and Bianca Gebhardt for their great and tireless efforts in participant and database management. We would also like to thank Beate Zoch-Lesniak for her great support in the study design at the Helmholtz Centre for Infection Research, Braunschweig, Germany.

The authors declare no conflict of interest.