Rostrum

Is there any role for allergen avoidance in theprimary prevention of childhood asthma?

Onno C. P. van Schayck, PhD,a Tanja Maas, MSc,a Janneke Kaper, PhD,a Andre J. A.

Knottnerus, MD,a and Aziz Sheikh, MD, FRCGPb Maastricht, The Netherlands,

and Edinburgh, United Kingdom

Revie

ws

and

featu

reart

icle

s

1323

In this article we discuss 3 hypotheses to attempt to understand

why preventive measures thus far studied with the aim of

preventing (or delaying) the development of asthma have

shown such disappointing results. The most likely explanation

is that the development of a multifactorial disease, such as

asthma, is extremely difficult, if not impossible, to prevent by

eliminating only one risk factor. In a meta-analysis we

investigated the effect of a multifaceted and monofaceted

intervention in 10 prospective birth cohorts of a total of 3473

children on a diagnosis of asthma. Multifaceted intervention

studies had an odds ratio (OR) of 0.73 (95% CI, 0.55-0.97),

whereas the monointervention studies had an OR of 1.22 (95%

CI, 0.83-1.78) in patients younger than 5 years and an OR of

0.52 (95% CI, 0.32-0.84) versus 0.93 (95% CI, 0.66-1.31) in

patients older than 5 years. We therefore hypothesize that

studies with a multifaceted approach will have a much greater

chance of being successful than studies using a monofaceted

approach, with the latter being unlikely to yield a clinically

relevant reduction of asthma. (J Allergy Clin Immunol

2007;119:1323-8.)

Key words: Asthma, prevention, multifaceted

It is widely accepted that asthma is a polygeneticdisease that can be aggravated by exposure to a range ofenvironmental factors. Genetically predisposed childrenexposed to specific allergens are believed to be atincreased risk of asthma development. On the basis ofthe understanding that the interaction between genotypeand environment plays a crucial role, it is reasonable toassume that reducing exposure to potentially relevantallergens should lead to a reduction in the risk of asthma.This concept is, however, mainly based on observations

From athe Department of General Practice, Research Institute Caphri, Maastricht

University, and bthe Allergy & Respiratory Research Group, Division of

Community Health Sciences: GP Section, University of Edinburgh.

Supported by an education grant of the Netherlands Asthma Foundation.

Disclosure of potential conflict of interest: T. Maas has received grant support

from the Dutch Asthma Foundation. The rest of the authors have declared

that they have no conflict of interest.

Received for publication January 10, 2007; revised February 16, 2007;

accepted for publication February 20, 2007.

Available online April 13, 2007.

Reprint requests: Onno van Schayck, PhD, Department of General Practice,

Research Institute Caphri, Maastricht University, PO Box 616, 6200 MD

Maastricht, The Netherlands. E-mail: onno.vanschayck@hag.unimaas.nl.

0091-6749/$32.00

� 2007 American Academy of Allergy, Asthma & Immunology

doi:10.1016/j.jaci.2007.02.024

made in the early 1980s, which found that dramaticreduction of house dust mite (Der p 1) exposure, by movingchildren with asthma into a hospital or to a mountainsanatorium, could very considerably reduce symptoms ofasthma.1,2 But translating these secondary prevention find-ings into feasible-to-deliver primary prevention strategieshas proved extremely difficult, as demonstrated by the dis-appointing results of large, carefully designed, randomizedclinical trials of house dust mite avoidance strategies inbirth cohorts.3,4 The reasons behind their lack of effective-ness remain unclear. Based on theoretical consideration,there are 3 main hypotheses:

1. It is not possible to reduce allergen exposure to aclinically relevant extent in real-life circumstances.

2. The concept is wrong: exposure to a specific allergenwill not increase but, through induction of tolerance,actually reduce the risk of asthma development.

3. Because asthma is a multifactorial disease, preven-tive measures will only prove effective if most orall relevant environmental factors for a specificchild are simultaneously avoided.

In this article we discuss these 3 hypotheses to attemptto understand why preventive measures thus far studied,which aimed to prevent (or delay) the development ofasthma, have shown such disappointing results. Our hopeis that these deliberations will assist researchers inprogressing beyond the current impasse in relation toinvestigating the role of allergen avoidance as a primaryprevention strategy.

HYPOTHESIS 1: IT IS NOT POSSIBLE TOREDUCE ALLERGEN EXPOSURE TOA CLINICALLY RELEVANT EXTENTIN REAL-LIFE CIRCUMSTANCES

The reduction in exposure needs to fulfill 2 essentialcriteria: it must be relevant because the child has to be

Abbreviation usedOR: Odds ratio

J ALLERGY CLIN IMMUNOL

JUNE 2007

1324 van Schayck et al

FIG 1. Numbers of asthmatic children in treatment and control groups are given for different studies at differ-

ent ages. Pooled analysis of the last observation of the different studies are given for a physician’s diagnosis

of asthma (<5 years of age) and for objectively defined asthma (�5 years of age). PREVASC, Prevention of

Asthma in Children; SPACE, Study of Prevention of Allergy in Children in Europe; CAPS, Childhood Asthma

Prevention Study; NACmass, NAC Manchester Asthma and Allergy Study.

Revie

ws

and

featu

rearticle

s

allergic for the specific allergen studied, and the reductionneeds to be of a magnitude sufficient to have a clinicaleffect.

Most studies in this area have focused on secondaryprevention, and the most extensive studied exposure in thisrespect is the house dust mite in those with establishedasthma. In studying the success of reducing allergenexposure measures, direct exposure to the allergen isseldom assessed; rather, surrogate parameters like theconcentration of house dust mite on bare mattresses orfloor covering are typically measured. But even then, notall trials have resulted in clear reductions of house dustmite concentrations. In the latest Cochrane systematicreview and meta-analysis of 49 trials (enrolling a total of2733 patients) investigating the effectiveness of house dustmite avoidance measures for treating asthma,5 31 studiedphysical methods to reduce exposure to mites (eg, barrierbedding), 10 studied chemical methods (acaracides), and8 studied a combination of approaches. Objectively

confirmed mite reduction occurred in only 13 trials andonly to a minor extent in the largest trial,6 and it was unsuc-cessful in 24 trials and not reported in the remaining 12.The final conclusion of the reviewers was that current ap-proaches to reducing exposure to mites are ineffective. Inarriving at this conclusion, the authors assumed that theabsence of an effect was due to an inadequate reductionof mite antigen levels. This interpretation is, however,not necessarily consistent with the evidence becauseeven in a study in which the reduction was clearly suc-cessful (ie, >90% reduction in Der p 1 resulting in a geo-metric mean of less than the accepted absolute thresholdvalue of 2 mg/g), no clinically relevant improvementswere observed in FEV1 or asthma symptoms.7 Only a sig-nificant difference in peak flow variability was observed.

Another point that should be considered is that even inthe controlled circumstances accompanying most trials,it is still possible that children are exposed and becomesensitized to dust mites in other houses.

J ALLERGY CLIN IMMUNOL

VOLUME 119, NUMBER 6

van Schayck et al 1325

FIG 1. Continued.Revie

ws

and

featu

reart

icle

s

Although the house dust mite is the allergen to whichasthmatic subjects are most frequently sensitive (exceptin specific areas in the world, such as Colorado orNorthern Scandinavia), one could argue that it is unlikelythat all subjects studied were allergic to the house dustmite alone. For this reason, we recently performed a trialin which asthmatic subjects were recruited who weresensitive to the house dust mite but not to (or notexposed to) other aeroallergens, such as cat or dog.8

Despite achieving a reduction in house dust mite aller-gens of almost 90% (resulting in a geometric mean<1 mg/g), there was only a slight increase in peak flow

and no change in symptoms. Therefore from these stud-ies, it is reasonable to conclude that in real-life circum-stances it is not possible to reduce the exposure tohouse dust mite in such a way that it has clinically rele-vant consequences for asthmatic patients. Whether thisconclusion for secondary prevention in asthmatic sub-jects can be extrapolated to primary prevention of asthmain susceptible children cannot be completely determined.However, as pointed out already, large studies of primaryprevention by reducing house dust mite allergens in chil-dren with a genetic susceptibility to asthma are not verypromising.3,4

J ALLERGY CLIN IMMUNOL

JUNE 2007

1326 van Schayck et al

HYPOTHESIS 2: EXPOSURE TO SPECIFICALLERGENS IN CHILDREN WILL, THROUGHINDUCTION OF TOLERANCE, DECREASE THERISK OF ASTHMA DEVELOPMENT

What actual evidence do we have that exposure tospecific allergens in children will decrease the risk ofasthma? Early-life exposure to aeroallergens from cats anddogs has been identified as possibly protecting againstallergy, with the proposed mechanism being immunetolerance9,10 or concomitant exposure to endotoxin.11

However, it is important to note that findings suggestinga role for possible tolerance are based on observational(cross-sectional and cohort) studies and not on observationsfrom randomized controlled trials. A fundamental limita-tion with nonexperimental designs is that it is extremely dif-ficult, if not impossible, to prove a causal relationship.Also, in cohort studies there are validity problems becausethe exposure is not determined by the play of chance andcould be influenced by behavior. Behavior might thereforebe a confounder because it can be associated with both ex-posure and outcome.12 Could, for example, allergic parentsand parents of children with respiratory symptoms be morelikely to decide not to have a cat or dog in their householdsthan other parents? Allergic parents might have experi-enced what effects these pets have on their own health,and parents of susceptible children can take the precautionof avoiding pets. Indeed, behavior of parents is likely toplay an important role in the exposure of high-risk childrento aeroallergens. In the Prevention and Incidence of Asthmaand Mite Allergy study, a randomized clinical trial investi-gating the efficacy of reducing allergen exposure, the aller-gen load before the intervention measures was so low thataccording to the authors this was likely to be caused by in-creased public awareness of the potential adverse effect ofallergen exposure, particularly among atopic families.13

Moreover, in a large study of more than 14,000 families,it was concluded that pet keeping seemed to be protectivefor the development of allergy, but this was mainly dueto the fact that parents avoided exposing their child topets because they believed this was a risk factor for the de-velopment of allergy.14 Although in natural cohort studiesattempts are often made to control for confounding behav-ior through multivariate analyses, one has to consider thatcorrection for confounding cannot always exclude residualconfounding because, for instance, unknown selection phe-nomena and poorly measured and unknown confoundingfactors cannot be adjusted for post hoc.

We would like to emphasize that the mentioned obser-vations do not exclude immune tolerance because the resultsof the studies of Ownby et al11 and Platt-Mills et al10 seem toindicate that high levels of allergen exposure to cats (andmaybe to dogs as well) were protective for some childrenand a risk factor for others, possibly genetically determined.

We conclude that on the basis of the present data, thereis insufficient evidence to assume that exposure to partic-ular allergens will decrease the risk of asthma develop-ment in all children.

Revie

ws

and

featu

rearticle

s

HYPOTHESIS 3: A MULTIFACETEDAPPROACH TO ALLERGEN AVOIDANCEWILL PROVE SUPERIOR TO A SINGLEINTERVENTION

Because asthmatic subjects are usually not only sensi-tive to one allergen, the successful elimination of that oneallergen is only likely to be of limited clinical benefit.It is therefore logical, although experimentally somewhatcomplex, to test the efficacy and effectiveness of multi-faceted approaches to allergen avoidance. The mainpotential disadvantage of a multifaceted approach is thatit is difficult to disentangle the effects of one interventionfrom those of another.

A number of randomized clinical trials have beenconducted in newborn children to investigate whether thedevelopment of asthma can be prevented or delayed byreducing exposure to inhalant allergens, feeding aller-gens, or both. Although the majority of these studiesinvestigated only one measure, others have used amultifaceted approach. Because there is no publishedsystematic review on possible differences between these2 approaches, we performed a search of the CochraneCentral Register of Controlled trials (Central, issue 1,2006), Medline (from January 1966 through December2006), and EMBASE (from January 1989 throughDecember 2006) using an appropriate search strategy(details available from first author). All identified refer-ences were screened by the second author. When therewas any doubt about whether to select a study, the firstauthor was consulted. Both reviewers assessed the full-article version of the selected studies. Data from theincluded studies were independently extracted, and dis-agreements were resolved through consultation with thethird author. Familiarity with many of the studies madeus aware of the many different outcome parameters anddifferent observation periods. Therefore to make maxi-mum use of available data, our primary outcome measureof interest was agreed a priori as being a physician’s di-agnosis of asthma in patients younger than 5 years and anobjectively defined diagnosis of asthma in patients olderthan 5 years. Because of the nature of the question wewere addressing in this review, it was likely that therewould be some variation in the length of follow-up car-ried out in each of the studies. We therefore aimed to an-alyze data on the diagnosis of asthma in 2 ways. First, weanalyzed the last observation available in each of thestudies, irrespective of time point, because the diagnosisof asthma is more reliable when a child grows older.We also reanalyzed data from multiple time points (aspooled subgroups). A formal test for statistical heteroge-neity, the natural approximate x2 test, assessed whetherthe observed variability in effect sizes is greater thanwould be expected to occur by chance. For dichotomousend points, an odds ratio (OR) and 95% CI were pre-sented by using a random effects model. We used theCochrane collaboration’s RevMan 4.2 program for theanalysis.

J ALLERGY CLIN IMMUNOL

VOLUME 119, NUMBER 6

van Schayck et al 1327

In total, 10 high-quality studies enrolling 3473 patientswere included, of which 1124 had a multifaceted interven-tion design15-17 and 2349 had a monointervention de-sign.4,18-23 Several studies had repeated measurements ondifferent ages of the children. In Fig 1 we displayed allpublished observations at different ages from the studies.The Isle of Wight study was published at ages 2, 4, and 8years.15,24,25 The Canadian Asthma Prevention study waspublished at ages 2 and 7 years.16,26 The Zeiger studywas also published at ages 2 and 7 years.19,27 The AustralianChildhood Asthma Prevention Study was published at ages3 and 5 years.21,28 Fig 1 shows that the multifaceted inter-vention studies had an OR of 0.73 (95% CI, 0.55-0.97),whereas the monointervention studies had an OR of 1.22(95% CI, 0.83-1.78) when patients were younger than 5years and an OR of 0.52 (95% CI, 0.32-0.84) versus anOR of 0.93 (95% CI, 0.66-1.31) when patients were olderthan 5 years.

These results suggest that a multifaceted approach iseffective in delaying or preventing asthma, whereas amonointervention does not seem to have such a protectiveeffect. The multi-faceted Prevention of Asthma in Childrentrial comparing the combination of mite allergen reduction,breast-feeding (thereby reducing cow’s milk protein expo-sure), and smoking cessation versus usual care showed anindependent and significant effect of mite allergen reduc-tion, breast-feeding, and maternal smoking on currentwheezing at 2 years of age.17 Epidemiologic observationsconfirmed the independent effects of different allergens(as well as smoking) on the incidence of asthma and onearly indications of asthma in healthy subjects.29,30 Also,a (secondary prevention) multifaceted intervention in asth-matic children was clearly shown to be effective whencompared with monointervention studies.31,32 All theseobservations support the concept that a multifaceted ap-proach to allergen avoidance is superior in diminishingthe development of asthma.

HYPOTHESIS THAT NOW NEEDSTO BE TESTED

It is clear that the enthusiasm induced by early obser-vations that reduction of a single allergen would diminishthe development of asthma has not been confirmed inexperimental studies. On the basis of the above describeddata, no definite conclusions can yet be drawn to explainthe lack of effectiveness of primary preventive measureson the development of asthma found in most studies.However, the most likely explanation is that the develop-ment of a multifactorial disease, such as asthma, isextremely difficult, if not impossible, to prevent byeliminating only one risk factor.28,33 We therefore hypoth-esize that studies with a multifaceted approach will have amuch greater chance of being successful than studies witha monofaceted approach, with the latter being unlikely toyield a clinically relevant reduction in asthma. The onlyway to draw definite conclusions is to study the effective-ness of separate and combined intervention measures in

one and the same study and to compare these effects onthe development of asthma. House dust mite allergenavoidance, cow’s milk protein avoidance, and a combina-tion of these approaches could, for example, be investi-gated in a randomized controlled trial, with the separateand combined measures investigated in separate arms.To the best of our knowledge, such a trial has not yetbeen performed, probably because of the large numbersneeded for such a trial. Mounting such a necessarily largetrial ideally involving at least 6 years of follow-up wouldbe challenging and expensive. If conducted, however,it would, we believe, shed important new light on therole of allergen avoidance measures in preventing thedevelopment of asthma in susceptible children.

REFERENCES

1. Boner AL, Niero E, Antolini I, Valletta EA, Gaburro D. Pulmonary func-

tion and bronchial hyper-reactivity in asthmatic children with house dust

mite allergy during prolonged stay in the Italian Alps (Misurina, 1756

m). Ann Allergy 1985;54:42-5.

2. Platt-Mills TA, Tovey ER, Mitchell EB, Moszoro H, Nock P, Wilkins

SR. Reduction of bronchial hyperreactivity during prolonged allergen

avoidance. Lancet 1982;2:675-8.

3. Koopman LP, van Strien RT, Kerkhof M, Wijga A, Smit HA, de Jongste

JC, et al. Placebo-controlled trial of house dust mite-impermeable mat-

tress covers: effect on symptoms in early childhood. Am J Respir Crit

Care Med 2002;166:307-13.

4. Woodcock A, Lowe LA, Murray CS, Simpson BM, Pipis SD, Kissen P,

et al. Early life environmental control: effect on symptoms, sensitization,

and lung function at age 3 years. Am J Respir Crit Care Med 2004;170:

433-9.

5. Gotzsche PC, Johansen HK, Schmidt LM, Burr ML. House dust mite

control measures for asthma. Cochrane Database Syst Rev 2004;18:

CD001187.

6. Woodcock A, Forster L, Matthews E, Martin J, Letley L, Vickers M,

et al. Control of exposure to mite allergen and allergen-impermeable

bed covers for adults with asthma. N Engl J Med 2003;349:225-36.

7. Cloosterman SG, Schermer TR, Bijl-Hofland ID, Van Der Heide S, Bru-

nekreef B, Van Den Elshout FJ, et al. Effects of house dust mite avoid-

ance measures on Der p 1 concentrations and clinical condition of mild

adult house dust mite-allergic asthmatic patients, using no inhaled

steroids. Clin Exp Allergy 1999;29:1336-46.

8. van den Bemt L, van Knapen L, de Vries MP, Jansen M, Cloosterman S,

van Schayck CP. Clinical effectiveness of a mite allergen-impermeable

bed-covering system in asthmatic mite-sensitive patients. J Allergy

Clin Immunol 2004;114:858-62.

9. Hesselmar B, Aberg N, Aberg B, Eriksson B, Bjorksten B. Does early

exposure to cat or dog protect against later allergy development? Clin

Exp Allergy 1999;29:611-7.

10. Platts-Mills T, Vaughan J, Squillace S, Woodfolk J, Sporik R. Sensitisa-

tion, asthma, and a modified Th2 response in children exposed to cat al-

lergen: a population-based cross-sectional study. Lancet 2001;357:752-6.

11. Ownby DR, Johnson CC, Peterson L. Exposure to dogs and cats in the

first year of life and risk of allergic sensitization at 6 to 7 years of age.

JAMA 2002;288:963-72.

12. van Schayck CP, Knottnerus JA. Can the hygiene hypothesis be ex-

plained by confounding by behaviour and health? J Clin Epidemiol

2004;57:435-7.

13. van Strien RT, Verhoeff AP, van Wijnen JH, Doekes G, de Meer GE,

Brunekreef B. Der p I concentrations in mattress surface and floor dust

collected from infants’ bedrooms. Clin Exp Allergy 1999;25:1184-9.

14. Bornehag CG, Sundell J, Hagerhed L, Janson S, DBH Study Group.

Pet-keeping in early childhood and airway, nose and skin symptoms

later in life. Allergy 2003;58:939-44.

15. Hide DW, Matthews S, Matthews L, Stevens M, Ridout S, Twiselton R,

et al. Effect of allergen avoidance in infancy on allergic manifestations at

age two years. J Allergy Clin Immunol 1994;93:842-6.

Revie

ws

and

featu

reart

icle

s

J ALLERGY CLIN IMMUNOL

JUNE 2007

1328 van Schayck et al

16. Becker A, Watson W, Ferguson A, Dimich-Ward H, Chan-Yeung M. The

Canadian asthma primary prevention study: outcomes at 2 years of age.

J Allergy Clin Immunol 2004;113:650-6.

17. Schonberger HJ, Dompeling E, Knottnerus JA, Maas T, Muris JW, van

Weel C, et al. The PREVASC study: the clinical effect of a multi-facet

educational intervention to prevent childhood asthma. Eur Respir J

2005;25:1-11.

18. Mallet E, Henocq A. Long-term prevention of allergic diseases by using

protein hydrolysate formula in at-risk infants. J Pediatr 1992;121(suppl):

S95-100.

19. Zeiger RS, Heller S. Genetic and environmental factors affecting the de-

velopment of atopy through age 4 in children of atopic parents: a pro-

spective randomized study of food allergen avoidance. Pediatr Allergy

Immunol 1992;3:110-27.

20. Horak F Jr, Matthews S, Ihorst G, Arshad SH, Frischer T, Kuehr J, et al.

Effect of mite-impermeable mattress encasings and an educational

package on the development of allergies in a multinational randomized,

controlled birth-cohort study—24 months results of the Study of Pre-

vention of Allergy in Children in Europe. Clin Exp Allergy 2004;34:

1220-5.

21. Peat JK, Mihrshahi S, Kemp AS, Marks GB, Tovey ER, Webb K, et al.

Three-year outcomes of dietary fatty acid modification and house dust

mite reduction in the Childhood Asthma Prevention Study. J Allergy

Clin Immunol 2004;114:807-13.

22. Johnstone DE, Dutton AM. Dietary prophylaxis of allergic disease in

children. N Engl J Med 1966;247:715-9.

23. Kjellman NIM, Johansson SGO. Soy versus cow’s milk in infants with a

biparental history of atopic disease: development of atopic disease and

immunoglobulins from birth to 4 years of age. Clin Allergy 1979;9:

347-58.

Revie

ws

and

featu

rearticle

s

24. Hide DW, Matthews S, Tariq S, Arshad SH. Allergen avoidance in

infancy and allergy at 4 years of age. Allergy 1996;51:89-93.

25. Arshad SH, Bateman B, Matthews SM. Primary prevention of asthma

and atopy during childhood by allergen avoidance in infancy: a rando-

mised controlled study. Thorax 2003;58:657-93.

26. Chan-Yeung M, Ferguson F, Watson W, Dimich-Ward H, Rousseau R,

Lilley M, et al. The Canadian Childhood Asthma Prevention Study:

outcomes at 7 years of age. J Allergy Clin Immunol 2005;116:49-55.

27. Zeiger RS, Heller S. The development and prediction of atopy in high-

risk children: Follow-up at age seven years in a prospective randomized

study of combined maternal and infant food allergen avoidance. J Allergy

Clin Immunol 1995;95:1179-90.

28. Marks BG, Mihrshahi S, Kemp AS, Tovey ER, Webb K, Almqvist C,

et al. Prevention of asthma during the first 5 years of life: a randomized

controlled trial. J Allergy Clin Immunol 2006;118:53-61.

29. Sunyer J, Sdoriano J, Anto JM, Burgos F, Pereira A, Payo F, et al. Sen-

sitization to individual allergens as risk factors for lower FEV1 in young

adults. European Community Respiratory Health Survey. Int J Epidemiol

2000;29:125-30.

30. Chen Y, Dales R, Tang M, Krewski D. Sex-related interactive effects of

smoking and household pets on asthma incidence. Eur Respir J 2002;20:

1162-6.

31. Morgan WJ, Crain EF, Gruchalla RS, O’Connor GT, Kattan M, Evans R

3rd, et al. Results of a home-based environmental intervention among

urban children with asthma. N Engl J Med 2004;351:1068-80.

32. Sheffer AL. Allergen avoidance to reduce asthma-related morbidity.

N Engl J Med 2004;351:1134-6.

33. Custovic A, van Wijk RG. The effectiveness of measures to change the

indoor environment in the treatment of allergic rhinitis and asthma;

ARIA update. Allergy 2005;60:1112-5.