Bønnelykke 1

Association between early life virus infections and later asthma

development is independent of virus type

Klaus Bønnelykke MD PhD1, Nadja Hawwa Vissing MD1, Astrid Sevelsted M.Sci1, Sebastian

L. Johnston MD, PhD2, Hans Bisgaard MD Dr Med Sci1

1) Copenhagen Prospective Studies on Asthma in Childhood; Copenhagen University

Hospital, Gentofte, Denmark.

2) Airway Disease Infection Section, MRC & Asthma UK Centre in Allergic Mechanisms of

Asthma, and Centre for Respiratory Infection, National Heart & Lung Institute, Imperial

College London, UK

Correspondence and request for reprints:

Professor Hans Bisgaard

Copenhagen Prospective Studies on Asthma in Childhood

Copenhagen University Hospital; Gentofte

Ledreborg Allé 34

DK-2820 Gentofte

Copenhagen

Denmark

Tel: +45 39 77 7360

Fax: +45 39 77 7129

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

Bønnelykke 2

E-mail: bisgaard@copsac.com

Website: www.copsac.com

Source of funding

COPSAC is funded by private and public research funds listed on www.copsac.com. The

Lundbeck Foundation, the Danish Strategic Research Council, the Pharmacy Foundation of

1991, the Augustinus Foundation, the Danish Medical Research Council, and the Danish

Pediatric Asthma Centre provided the core support for COPSAC research center. SLJ is

supported by ERC FP7 Advanced grant 233015, a Chair from Asthma UK (CH11SJ), MRC

Centre Grant G1000758 and Predicta FP7 Collaborative Project grant 260895.

The funding agencies did not have any role in study design, data collection and analysis,

decision to publish, or preparation of the manuscript.

None of the authors report any conflict of interest relevant to the content of this report.

Article Type: Original article

Word count: 3,239

Figure count: 2

Tables: 2

Data in this manuscript has been published as an abstract at the EAACI congress 2014 but not

in any other format.

This article has an online data supplement.

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

Bønnelykke 3

ABSTRACT

Background: Early respiratory tract infections are associated with later asthma and this

observation has led to a focus on a potential causal role of specific respiratory viruses, such as

rhinoviruses (RV) and respiratory syncytial virus (RSV), in asthma development. However,

many respiratory viruses and bacteria trigger similar respiratory symptoms, and it is possible

that the important risk factors for asthma are the underlying susceptibility to infections and

the exaggerated reaction to such triggers rather than the particular triggering agent.

Objective: To study the association between specific infections in early life and development

of asthma later in childhood.

Methods: 313 children were followed prospectively in the Copenhagen Prospective Studies

of Asthma in Childhood (COPSAC) 2000 high-risk birth cohort. Nine respiratory virus types

(RSV, RV, other picornaviruses, coronaviruses 229E & OC43, parainfluenza viruses 1-3,

influenza viruses, human metapneumovirus, adenoviruses, and bocavirus) and three

pathogenic airway bacteria (Streptococcus pneumoniae, Haemophilus influenzae, Moraxella

catarrhalis) were identified in airway secretions sampled during episodes of troublesome lung

symptoms during the first three years of life. Asthma was determined by age seven.

Results: In unadjusted analyses, all viruses and pathogenic bacteria identified during episodes

of troublesome lung symptoms were associated with increased risk of asthma by age seven

years with similar odds ratios. After adjustment for frequency of respiratory episodes the

particular triggers were no longer associated with asthma.

Conclusion: The number of respiratory episodes in the first years of life, but not the

particular viral trigger, was associated with later asthma development. This suggests that

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

Bønnelykke 4

future research should focus on the susceptibility and exaggerated response to respiratory

infections in general, rather than the specific triggering agent.

63

64

Bønnelykke 5

Key messages

Early respiratory tract infections are associated with later asthma, and this has led to a

focus on a potential causal role of specific viral triggers, particularly rhinoviruses and

respiratory syncytial virus.

In this study, the number of early respiratory episodes was associated with asthma

irrespective of virus type.

This suggests that future research should focus on the susceptibility and response to

respiratory infections in general, rather than the specific triggering agent.

Capsule summary

The number of early respiratory episodes was associated with asthma irrespective of virus

type or presence of pathogenic bacteria suggesting that future research should focus on the

susceptibility and response to respiratory infections in general, rather than specific viral

triggers.

Key words

child, asthma, virus, bacteria, respiratory tract infection

65

66

67

68

69

70

71

72

Bønnelykke 6

Introduction

It is a solid clinical observation that asthmatic children exacerbate in relation to a common

cold, and viruses are more commonly found during asthma exacerbations than in the

asymptomatic state.1 In addition, early life virus infections have been suspected to initiate a

chronic disease trajectory leading to recurrent wheeze and asthma later in childhood, and this

putative causal role in asthma development has stimulated a strong focus on the role and

pathogenic mechanisms of specific viral agents.2–7 With respect to long term consequences on

asthma risk, there has previously been a particular focus on respiratory syncytial virus (RSV)

infection as a cause of instigating abnormal pulmonary function, wheezing, and asthma in

childhood,2–6 while recently there has been a focus on the role of human rhinovirus (RV)

infection in asthma development.7

However, several viral and bacterial agents can elicit similar asthmatic symptoms in young

children8 and trigger asthma exacerbations in school-age.1 Furthermore, it is clear that the

individual susceptibility plays an important role in the response to respiratory infections.

Individuals with asthma have an altered epithelial immune response to rhinoviruses9 and are

more susceptible to develop lower respiratory tract infections in relation to these.10 Also,

hyper-reactive airways are a hall mark of children with asthma and we and others have shown

that children who later develop asthma have increased bronchial responsiveness before any

lower respiratory tract illness.11,12

It is therefore possible that the important risk factors for asthma are the preexisting

susceptibility and inflammatory response to respiratory infections in general, rather than the

specific triggering agent. This is a key question for directing future research: if specific

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

Bønnelykke 7

viruses confer a specific risk, then viral-specific mechanisms should be the focus.

Alternatively, future research should focus on the underlying susceptibility and

hyperreactivity to infections in general.

We analyzed the association between specific viruses and pathogenic bacteria detected during

episodes of troublesome lung symptoms in early life and asthma status at age seven in the

Copenhagen Prospective Studies on Asthma in Childhood (COPSAC) 2000 birth cohort.

Materials and Methods

Study subjects

The Copenhagen Prospective Study on Asthma in Childhood (COPSAC) is a prospective

clinical birth cohort study, conducted in accordance with the Declaration of Helsinki and

approved by the Copenhagen and Frederiksberg Ethics Committee (KF 01-289/96 and KF 11-

107/02) and the Danish Data Protection Agency (2008-41-1754). During 1998-2001 the study

enrolled 411 neonates at one month of age, born to mothers with a history of asthma,

excluding children born before 36 weeks of gestation and anyone suspected of chronic

diseases or lung symptoms prior to inclusion as previously described in detail13,14.

Troublesome Lung Symptoms

Significant troublesome lung symptoms (Trolls) were recorded in daily diaries from 1 month

of age until age 3 years as recently described and analysed in detail.15 Parents were taught to

record their child’s symptoms with emphasis on the lower airways. Trolls were translated to

the parents as any symptom significantly affecting the child’s breathing such as noisy

breathing (wheeze or whistling sounds), shortness of breath or persistent troublesome cough

affecting the sleep or activity of the child. Daily symptoms were recorded as composite

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

Bønnelykke 8

dichotomized scores (yes/no) each day; ie, the parents were taught to make a global

assessment. The complexity of symptoms was detailed in a book that was given to the parents

(www.copsac.com/content/literature-parents). The diary cards were collected and reviewed by

the doctors at the half-yearly clinic visits.

Parents were encouraged to bring the children to the research clinic for an acute visit after

each three-day episode of Trolls recorded in the diary. The research doctors examined the

children at each acute visit in accordance with predefined standard procedures. The children

received a standardised physical examination, including auscultation of the lungs. The doctor

performed nasopharyngeal and hypopharyngeal aspirates under aseptic conditions with two

separate soft suction catheters passed into the upper rhino-pharynx and hypo-pharynx for

analysis of viruses and bacteria, respectively8.

Three consecutive days of Trolls with available virus or bacteria aspirates was the predefined

base-unit of the symptom burden (episodes of Trolls) and the number of such episodes was

summarized as an indicator to reflect respiratory symptom burden.15

Viruses were detected by PCR analysis of nasopharyngeal aspirate samples for respiratory

syncytial virus (RSV), rhinoviruses (RV), other picornaviruses, coronaviruses 229E and

OC43, parainfluenza viruses 1-3, influenza viruses AH1, AH3 and B, human

metapneumovirus (HMPV), adenoviruses and bocavirus, as previously described.8

Pathogenic airway bacteria (Streptococcus pneumoniae, Haemophilus influenzae, and

Moraxella catarrhalis) were identified according to standard procedures as previously

described8 and analysed as the dichotomized measure of at least one bacteria detected.

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

Bønnelykke 9

Sensitization

Sensitization was determined by skin prick test and/or determined by specific IgE

measurement to any allergens at six years of age. Specific IgE levels were measured by

ImmunoCAP (Phadia AB, Uppsala, Sweden)16 against cat, dog, horse, birch, timothy grass,

Dermatophagoides pteronyssinus, Dermatophagoides farinae, mugwort, molds (Paspalum

notatum, Cladosporium herbarium, Aspergillus fumigatus, and Alternaria alternata), egg,

milk, peanut, cod, wheat, and soya bean allergens. Skin prick test was performed with cat,

dog, horse, birch, timothy grass, D. pteronyssinus, D. farinae, mugwort, A. alternata, C.

herbarium, egg, milk, peanut, cod, wheat, rye, beef, pork, and soya bean allergen extracts

(ALK-Abelló, Copenhagen, Denmark), as well as raw egg and milk.17 Sensitization was

defined as having any positive skin prick test (>3mm) or having any specific IgE >3.5 kU/L.

Infant lung function

Infant lung function lung function was measured at 1 month of age by the rapid raised volume

thoraco-abdominal compression technique as previously described.18 The forced expiratory

flow at 50% of the functional vital capacity (FEF50) was used in this study since this was the

measure previously shown to be most closely associated with later asthma.18

Asthma at age seven

Asthma at the age of seven was diagnosed by the doctor at the research unit in accordance

with international guidelines. The burden of recurrent symptoms was quantified from an

algorithm of five episodes of Trolls as defined above within six months and in need of short-

acting ß2-agonists as previously described in details.13,15 The quality of symptoms was judged

by the study clinician to be typical of asthma exercise-induced symptoms, prolonged

nocturnal cough, persistent cough outside common cold, and symptoms causing wakening at

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

Bønnelykke 10

night. Furthermore, the diagnosis required symptom improvement during a three-month trial

of inhaled corticosteroids and relapse when this medication was stopped.

Confounders

Possible confounding effect were evaluated for selected risk factors with suspected relevance

to risk of childhood asthma: 17q21 locus (ORMDL3) 19 and Filaggrin (FLG) genotypes,20

mothers level of education, nicotine in hair by 12 months, and father’s asthma.

Statistical analyses

Distribution of risk factors in the study population and in children excluded from the analysis

was compared by chi-square for categorical variables and t-test for continuous variables.

Nicotine in hair was log-transformed to obtain normal distribution.

Children with a positive identification of a particular virus or bacteria during an infection

were compared to children who never had a positive identification of this specific pathogen,

either because their aspirates were never positive for that particular pathogen or because they

were never ill enough to have nasopharyngeal suction performed in the first 3 years of life.

Associations between each specific virus or bacteria and asthma at age 7 were investigated

with logistic regression. Subsequently we adjusted for the child’s total number of episodes of

Trolls with an available sample for virus or bacteria to obtain the specific trigger effect

adjusted for general infection susceptibility.

Possible interaction with allergic sensitization and infant lung function was investigated by

including the interaction term in the regression analysis.

A significance level of 0.05 was used in all types of analyses. All analyses were conducted in

SAS version 9.3 for Windows (SAS Institute Inc, Cary, NC, USA).

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

Bønnelykke 11

Results

Study Base

The COPSAC birth cohort included 411 new-borns. This study analysed those 313 children

with complete follow-up in the first three years of life (66 children missing) and known

asthma status at seven years of age (32 children missing) (Figure 1).

Children in the study group and excluded children did not differ with respect to 17q21 locus

and FLG genotypes; mothers level of education, or nicotine in hair by 12 months, but father’s

asthma was more prevalent in the study group than the excluded group (20 % versus 7 %,

p=0.0022).

Symptom burden

The majority of children had between one and five episodes of Trolls; 10% never reported

Trolls in their diary cards, and 30% had Trolls in the diary card but were not seen at the clinic,

hence had no virus aspirate taken. These children were assumed not to have had any clinically

significant episode of Trolls.

In total, 228 (%) of the 313 children had at least one episode of Trolls with a virus or bacteria

aspirate taken.

Prevalence of viruses and bacteria

A total of 650 virus aspirates were collected. Viruses were identified in 423 of 650 samples

(65%). RVs were most frequent, found in 147 samples (23%), followed by RSV in 124

samples (19%) and coronaviruses in 79 samples (12%). 184 children had at least one

identifiable virus; 106 had several viruses, some on several occasions.

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

Bønnelykke 12

A total of 614 bacteria aspirates were collected. Bacteria were identified in 533 (87%) of

these. S. pneumonia was found in 295 samples (48%), H. influenzae,was found in 275

samples (45%) and M. catarrhalis was found in 304 samples (50%). 206 children had at least

one identifiable pathogenic bacteria; 132 had several.

Asthma Prevalence

The asthma prevalence by age seven was 15% (46 of 313). 30% of the children with asthma

and 17% of children without asthma were sensitized to one or more allergens at six years of

age (p=0.052).

Risk of asthma in relation to respiratory infections

Simple logistic regression analyses showed that children with RV were at significantly greater

risk of asthma at age seven compared to subjects never having RV (OR 2.58 [1.36-4.89];

p=0.0037). The risk of asthma was similarly increased in children with RSV (OR 2.69 [1.43-

5.09]; p=0.0023); coronavirus (OR 3.14 [1.59-6.22]; p=0.0010), other viruses (OR 3.88 [2.01-

7.49]; p<0.0001), or pathogenic bacteria (table 1). Having an episode of Trolls without any

virus detected was also a significant risk factor for asthma at age 7 (OR 3.62 [1.82-7.20];

p=0.0002) (not in table). Also the simple number of episodes of Trolls per child was

significantly associated with developing asthma (OR 1.43 [1.26-1.63]; p<0.0001), table 1.

There was no significant risk from any of the viruses or bacteria when the simple association

analyses were adjusted for the child’s total number of episodes of Trolls, (Table 1 and Figure

2) while the association between the number of episodes of Trolls and asthma remained

significant after adjustment for any positive virus or bacteria aspirate (OR 1.39 [1.21-1.59];

p<0.0001), table 1.

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

Bønnelykke 13

Analysing number of episodes with specific virus types and bacteria gave similar results

(Table 2). The effect estimate was similar for the different triggers in the unadjusted analyses,

and in the multivariable model including both numbers of the specific infection and the total

number of Trolls, only the number of Trolls remained significantly associated with asthma.

Limiting the analysis to the events in the first year of life (Online Table E-1) or to children

having at least episode of Trolls in the first 3 years of life (Online Table E-2) did not change

the conclusions.

The associations between specific infections and asthma were largely unchanged after

adjustment for sensitization at age six or infant lung function (data not shown). We further

tested if sensitization or infant lung function modified the association between specific

infections and later asthma. There were some trends of stronger association between specific

viral infections and asthma in some strata, including a stronger association for RSV in

children with lower infant lung function and an opposite tendency for RSV and corona virus.

However, overall there was no statistical evidence of effect modification (all interaction p-

values were >.12) (Online Table E-3).

DISCUSSION

Simple association analyses showed association between common respiratory viruses and

later development of asthma but with no differences in risk between the most common virus

types (RV, RSV and coronavirus). There was no significant risk of school age asthma from

any virus or pathogenic bacteria when adjusting for the number of respiratory episodes. That

is, for a given burden of acute respiratory episodes it did not matter if any of these, or how

many of these, were related to a particular trigger. This suggests that the most important risk

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

Bønnelykke 14

factor for later development of asthma is the number of acute respiratory episodes rather that

the specific viral trigger.

Strengths of the Study

The major strength of the COPSAC study is the meticulous prospective clinical monitoring,

diagnosing, and treating lung symptoms based on standard operating procedures. The birth

cohort was followed prospectively with diary cards and six-monthly routine visits (three-

monthly for children diagnosed with asthma). In addition, the child attended the clinic for

episodes of Trolls, which was part of the nested randomized controlled trial of inhaled

corticosteroids, showing no clinical effect compared to placebo21. At each visit the child was

seen by the research doctor at the dedicated research clinic where full clinical work-up was

performed, including objective assessment of the respiratory status of the child and aspirates

for microbiology.

The base-unit of symptom burden was defined as three-day episodes of Trolls (troublesome

lung symptoms significantly affecting the wellbeing of the child, excluding croup) and having

information on virus. We avoided the use of specific terms such as wheeze, which we have

demonstrated has a very low sensitivity to asthma.15,22,23 This base-unit was defined

objectively from diary cards, which we have previously found to be the more sensitive

measure of asthma propensity with a closer association to known risk factors of asthma such

as 17q21 locus genotypes than the traditional temporal categories.15,24

When investigating any specific virus we chose to include all other clinically assessed

episodes of Trolls in the comparison group, i.e. the non-exposed group include children

exacerbating with other triggers than the case trigger. This was a conservative approach. If we

had used the group of children with no respiratory infections ever, this would only have

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

Bønnelykke 15

boosted the unadjusted odds ratios, but would not affect the result of the adjustment for

recurrent episodes of Trolls.

Asthma diagnosis followed a standardized algorithm based on the recorded symptom severity

and including request for improvement on treatment and relapse off treatment, as well as a

clinical history compatible with asthma as judged by the study physician.

Limitations of the study

We presume that children active in the cohort, but without clinic visits for episodes of Trolls

in the first three life years did not have clinically significant episodes of Trolls. This

introduces a possible bias against the null hypothesis, i.e. if this control group indeed had

similar symptomatic episodes this would reduce the contrast between two groups and reduce

the risk estimates of asthma. Furthermore, it did not change the conclusions if we excluded

children not attending the clinic for episodes of Trolls from the comparator group, i.e. we

found similar conclusions when we limited the analyses to the 228 children presenting to the

clinic with episodes.

Due to the size of the study, we had limited statistical power to detect effect modification by

allergic sensitization.

Similar to previous studies on this subject, this is an observational study and we therefore

cannot draw conclusions about any potential causal effects of virus infections on asthma risk.

The external validity of this study is limited by the high-risk nature of the cohort. Still, this is

unlikely to affect the distribution of virus types or the risk of asthma between them.

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

Bønnelykke 16

Interpretation

Our study suggests that the specific microbial trigger is not important for the risk of later

development of asthma. This directs future research towards studies of individual

susceptibility and inflammatory consequences of respiratory infections in general.

Our findings contrast previous reports from another birth cohort study finding that acute

respiratory episodes with detection of RV infection were more strongly associated with later

asthma than episodes with detection of RSV.7 However, those data are from a relatively small

study and the very high risk estimates observed for RV are associated with considerable

statistical uncertainty as evidenced by the wide confidence intervals. Another birth cohort

study reported, in line with our findings, that respiratory infections with RSV and RV in the

first year of life were associated with wheeze and asthma by age 5 years with no statistically

significant difference between the 2 viruses.25 We have previously shown that pathogenic

bacteria are associated with wheezing episodes in young children8 but, similar to our finding

for viruses, it did not seem to determine the long term course of asthma if the episodes were

triggered by bacteria. None of the previous studies of viral infections analyzed the presence of

pathogenic bacteria.

Our study relates to the ongoing debate of early viral infections being either causative for

asthma or merely a marker of the underlying asthma constitution, or perhaps both. There is a

significant body of work published on the association between particularly RSV2–6 and more

recently RV7 and later development of asthma. On the other hand, predisposition to asthma

was associated with increased risk of lower respiratory tract infection and RSV bronchiolitis,

the intermediary phenotype associated with asthma.26 Furthermore, early wheezy symptoms

were found to be a strong risk factor for subsequent RSV-hospitalization.27 We recently

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

Bønnelykke 17

demonstrated that children developing acute severe virus bronchiolitis were born with

increased bronchial responsiveness compared to children not developing bronchiolitis,28 and

allergic sensitization was a risk factor for RV-induced wheezing.29 A differential risk from

different virus types during acute episodes could be interpreted in support of a causal effect

from certain viruses. However, our study did not support this observation. Similar to other

observational studies, we cannot draw any conclusions on causality of early respiratory

infections on later asthma from this study. It is plausible that early respiratory infections

interacts with genetic susceptibility to asthma and initiates a trajectory towards asthma. It has

been reported that timing of birth in relation to the yearly peak in respiratory viral infections

was associated with later asthma in support of a causal role of early infections.30 A causal role

can only be proven in interventional studies showing that removing the virus or preventing

infection changes the risk of subsequent asthma. A recent randomized trial of monoclonal

antibodies against RSV demonstrated a protective effect against wheeze in the first year of

life, also after the end of treatment, and further follow-up will provide information on the

causal role of early RSV infection on asthma.31 Our results showing a non-differential

response to different viral infections would suggest that a potential causal effect on asthma

would not be restricted to RSV infections.

Genetic studies may help understanding the effects of environmental exposures. An

interaction between early respiratory infections and the asthma locus at chromosome 17q21

was recently reported from the COPSAC and COAST birth cohorts.32 This locus seemed

primarily to be associated with increased susceptibility to respiratory episodes triggered by

RV infection. Future studies must show if genetics can improve our understanding on the

relationship between early infections and asthma.

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

Bønnelykke 18

Conclusion

Our study suggests that the number of early respiratory episodes, but not the specific viral

trigger, is associated with development of school age asthma. Particularly, we saw no

difference in asthma risk associated with RV or RSV. This suggests that the susceptibility and

response to early respiratory infections in general, rather than the specific triggering agent

should be the target of future research.

Acknowledgements

We thank the children and families of the COPSAC2000 cohort study for all their support and

commitment. We also acknowledge and appreciate the unique efforts of the COPSAC

research team and T. Kebadze and J. Aniscenko for the virologic analyses.

The guarantor of the study is HB who has been responsible for the integrity of the work as a

whole, from conception and design to acquisition of data, analysis and interpretation of data

and writing of the manuscript. HB had full access to all the data in the study and had final

responsibility for the decision to submit for publication. KB and NHV contributed to analysis

and interpretation of data and writing of the manuscript. AS performed statistical analyses and

contributed to acquisition of data, data interpretation and writing of the manuscript. SLJ

contributed the virologic analyses and interpretation of data and writing of the manuscript.

All authors made important intellectual contributions and critical final revision of the

manuscript.

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

Bønnelykke 19

Conflict of interest statement

The authors have no conflicts of interest relevant to this topic.

The corresponding author had full access to the data and had final responsibility for the

decision to submit for publication.

References

1. Johnston SL, Pattemore PK, Sanderson G, et al. Community study of role of viral

infections in exacerbations of asthma in 9-11 year old children. BMJ

1995;310(6989):1225–9.

2. Stein RT, Sherrill D, Morgan WJ, et al. Respiratory syncytial virus in early life and risk

of wheeze and allergy by age 13 years. The Lancet 1999;354(9178):541–5.

3. Sigurs N, Gustafsson PM, Bjarnason R, et al. Severe Respiratory Syncytial Virus

Bronchiolitis in Infancy and Asthma and Allergy at Age 13. Am J Respir Crit Care Med

2005;171(2):137–41.

4. Kuehni CE, Spycher BD, Silverman M. Causal links between RSV infection and asthma:

no clear answers to an old question. Am J Respir Crit Care Med 2009;179(12):1079–80.

5. Pullan CR, Hey EN. Wheezing, asthma, and pulmonary dysfunction 10 years after

infection with respiratory syncytial virus in infancy. Br Med J (Clin Res Ed)

1982;284(6330):1665–9.

6. McBride JT. Pulmonary function changes in children after respiratory syncytial virus

infection in infancy. J Pediatr 1999;135(2 Pt 2):28–32.

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

Bønnelykke 20

7. Jackson DJ, Gangnon RE, Evans MD, et al. Wheezing Rhinovirus Illnesses in Early Life

Predict Asthma Development in High-Risk Children. Am J Respir Crit Care Med

2008;178(7):667–72.

8. Bisgaard H, Hermansen MN, Bønnelykke K, et al. Association of bacteria and viruses

with wheezy episodes in young children: prospective birth cohort study. BMJ

2010;341:c4978.

9. Edwards MR, Regamey N, Vareille M, et al. Impaired innate interferon induction in

severe therapy resistant atopic asthmatic children. Mucosal Immunol 2013;6(4):797–

806.

10. Corne JM, Marshall C, Smith S, et al. Frequency, severity, and duration of rhinovirus

infections in asthmatic and non-asthmatic individuals: a longitudinal cohort study.

Lancet 2002;359(9309):831–4.

11. Bisgaard H, Jensen SM, Bønnelykke K. Interaction between asthma and lung function

growth in early life. Am J Respir Crit Care Med 2012;185(11):1183–9.

12. Martinez FD, Morgan WJ, Wright AL, Holberg CJ, Taussig LM. Diminished lung

function as a predisposing factor for wheezing respiratory illness in infants. N Engl J

Med 1988;319(17):1112–7.

13. Bisgaard H, Hermansen MN, Buchvald F, et al. Childhood asthma after bacterial

colonization of the airway in neonates. N Engl J Med 2007;357(15):1487–95.

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

Bønnelykke 21

14. Bisgaard H. The Copenhagen Prospective Study on Asthma in Childhood (COPSAC):

design, rationale, and baseline data from a longitudinal birth cohort study. Ann Allergy

Asthma Immunol 2004;93(4):381–9.

15. Bisgaard H, Pipper CB, Bønnelykke K. Endotyping early childhood asthma by

quantitative symptom assessment. J Allergy Clin Immunol 2011;127(5):1155–1164.e2.

16. Paganelli R, Ansotegui IJ, Sastre J, et al. Specific IgE antibodies in the diagnosis of

atopic disease. Clinical evaluation of a new in vitro test system, UniCAP, in six

European allergy clinics. Allergy 1998;53(8):763–8.

17. Bisgaard H, Li N, Bonnelykke K, et al. Reduced diversity of the intestinal microbiota

during infancy is associated with increased risk of allergic disease at school age. J

Allergy Clin Immunol 2011;128(3):646–652.e5.

18. Bisgaard H, Jensen SM, Bønnelykke K. Interaction between asthma and lung function

growth in early life. Am J Respir Crit Care Med 2012;185(11):1183–9.

19. Moffatt MF, Kabesch M, Liang L, et al. Genetic variants regulating ORMDL3

expression contribute to the risk of childhood asthma. Nature 2007;448(7152):470–3.

20. Bønnelykke K, Pipper CB, Palmer CN, Bisgaard H, et al.Filaggrin gene variants and

atopic diseases in early childhood assessed longitudinally from birth. Pediatr Allergy

Immunol 2010; 21: 954-61.

21. Bisgaard H, Hermansen MN, Loland L, Halkjaer LB, Buchvald F. Intermittent inhaled

corticosteroids in infants with episodic wheezing. N Engl J Med 2006;354(19):1998–

2005.

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

Bønnelykke 22

22. Skytt N, Bønnelykke K, Bisgaard H. “To wheeze or not to wheeze”: That is not the

question. J Allergy Clin Immunol 2012;130(2):403–407.e5.

23. Bisgaard H, Swern AS, Knorr B. “To wheeze or not to wheeze”: that is not the

question--the sequel. J Allergy Clin Immunol 2012;130(2):531–2.

24. Bisgaard H, Bønnelykke K, Sleiman PMA, et al. Chromosome 17q21 gene variants are

associated with asthma and exacerbations but not atopy in early childhood. Am J Respir

Crit Care Med 2009;179(3):179–85.

25. Kusel MMH, de Klerk NH, Kebadze T, et al. Early-life respiratory viral infections,

atopic sensitization, and risk of subsequent development of persistent asthma. Journal of

Allergy and Clinical Immunology 2007;119(5):1105–10.

26. Goetghebuer T, Kwiatkowski D, Thomson A, Hull J. Familial susceptibility to severe

respiratory infection in early life. Pediatric Pulmonology 2004;38(4):321–8.

27. Stensballe LG, Kristensen K, Simoes EAF, et al. Atopic Disposition, Wheezing, and

Subsequent Respiratory Syncytial Virus Hospitalization in Danish Children Younger

Than 18 Months: A Nested Case-Control Study. Pediatrics 2006;118(5):e1360 –e1368.

28. Chawes BLK, Poorisrisak P, Johnston SL, Bisgaard H. Neonatal bronchial

hyperresponsiveness precedes acute severe viral bronchiolitis in infants. J Allergy Clin

Immunol 2012;130(2):354–361.e3.

29. Jackson DJ, Evans MD, Gangnon RE, et al. Evidence for a causal relationship between

allergic sensitization and rhinovirus wheezing in early life. Am J Respir Crit Care Med

2012;185(3):281–5.

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

Bønnelykke 23

30. Wu P, Dupont WD, Griffin MR, et al. Evidence of a causal role of winter virus infection

during infancy in early childhood asthma. Am J Respir Crit Care Med

2008;178(11):1123–9.

31. Blanken MO, Rovers MM, Molenaar JM, et al. Respiratory syncytial virus and recurrent

wheeze in healthy preterm infants. N Engl J Med 2013;368(19):1791–9.

32. Calışkan M, Bochkov YA, Kreiner-Møller E, et al. Rhinovirus wheezing illness and

genetic risk of childhood-onset asthma. N Engl J Med 2013;368(15):1398–407.

439

440

441

442

443

444

445

446

Bønnelykke 24

Table 1. Risk of asthma by age seven in relation to respiratory infections within the first three

years of life.

No. of children with virus/

No. of children with no virus

Unadjusted

OR [95% CI]; p-value

Adjusted for episodes of Trolls

OR [95% CI]; p-value

RV 92/221 2.58 [1.36 - 4.89]; p=0.004

0.63 [0.27 - 1.51]; p=0.30

RSV 101/212 2.69 [1.42 - 5.09]; p=0.002

1.01 [0.47 - 2.16]; p=0.99

Corona 59/254 3.14 [1.59 - 6.22]; p=0.001

1.30 [0.58 - 2.91]; p=0.52

Other virus 117/196 3.88 [2.01 - 7.49]; p<0.001

1.33 [0.58 - 3.05]; p=0.49

Any virus 184/129 3.94 [1.77 - 8.76]; p<0.001

1.09 [0.40 - 2.99]; p=0.86

Any pathogenic bacteria

206/1075.07 [1.94 - 13.25];

p<0.0011.06 [0.12 - 9.17];

p=0.96

No. of children with and

without any episodes

(clinic visit)

UnadjustedOR [95% CI]; p-

value

Adjusted for any virus/ bacteria

infection.OR [95% CI]; p-

valueEpisodes of Trolls (ANY) 228/85 1.43 [1.26-1.61];

p<0.0011.39 [1.21-1.59];

p<0.001

447

448

449450

Bønnelykke 25

Table 2. Risk of asthma by age seven in relation to number of respiratory infections within the first three years of life.

Effect of number of episodes with a specific agentUnadjusted

OR [95% CI]; p-value

Effect of number of episodes with a specific agent Adjusted for episodes of Trolls

OR [95% CI]; p-value

Effect of TrollsAdjusted for number of episodes with each agent

OR [95% CI]; p-value

RV 1.94 [1.43 - 2.62]; p<0.001

0.89 [0.56 - 1.4]; p=0.60

1.47 [1.24 - 1.74]; p<0.001

RSV 1.98 [1.31 - 3.01]; p=0.002

0.94 [0.55 - 1.59]; p=0.81

1.44 [1.25 - 1.65]; p<0.001

Corona 2.14 [1.40 - 3.27]; p<0.001

1.19 [0.73 - 1.94]; p=0.49

1.40 [1.22 - 1.60]; p<0.001

Other virus 1.81 [1.35 - 2.43]; p<0.001

0.97 [0.65 - 1.42]; p=0.86

1.44 [1.23 - 1.68]; p<0.001

Any virus 1.63 [1.36 - 1.95]; p<0.001

1.09 [0.75 - 1.58]; p=0.65

1.36 [1.06 - 1.74]; p=0.016

Any pathogenic bacteria

1.57 [1.34 - 1.84]; p<0.001

1.15 [0.83 - 1.61]; p=0.39

1.30 [1.01 - 1.66]; p=0.040

451452453

454

Bønnelykke 26

Figure legends

Figure 1. Flow chart for the study population.

Figure 2. Risk of asthma at age seven in relation to viral or bacterial infection in the first 3

years of life. Grey squares: crude Odds Ratio with 95% Confidence Intervals. Black squares:

adjusted Odds Ratio (adjusted for the child’s total number of episodes of TROLLS where an

aspirate was taken).

455

456

457

458

459

460

461