Early eczema and the risk of childhood asthma: a prospective, population-based study
© Saunes et al.; licensee BioMed Central Ltd. 2012
Received: 31 January 2012
Accepted: 15 October 2012
Published: 24 October 2012
Skip to main content
© Saunes et al.; licensee BioMed Central Ltd. 2012
Received: 31 January 2012
Accepted: 15 October 2012
Published: 24 October 2012
Severe eczema in young children is associated with an increased risk of developing asthma and rhino-conjunctivitis. In the general population, however, most cases of eczema are mild to moderate. In an unselected cohort, we studied the risk of current asthma and the co-existence of allergy-related diseases at 6 years of age among children with and without eczema at 2 years of age.
Questionnaires assessing various environmental exposures and health variables were administered at 2 years of age. An identical health questionnaire was completed at 6 years of age. The clinical investigation of a random subsample ascertained eczema diagnoses, and missing data were handled by multiple imputation analyses.
The estimate for the association between eczema at 2 years and current asthma at 6 years was OR=1.80 (95% CI 1.10-2.96). Four of ten children with eczema at 6 years had the onset of eczema after the age of 2 years, but the co-existence of different allergy-related diseases at 6 years was higher among those with the onset of eczema before 2 years of age.
Although most cases of eczema in the general population were mild to moderate, early eczema was associated with an increased risk of developing childhood asthma. These findings support the hypothesis of an atopic march in the general population.
The Prevention of Allergy among Children in Trondheim study has been identified as ISRCTN28090297 in the international Current Controlled Trials database
The atopic march is a term used to describe the relationship between allergy-related diseases, starting with food allergy and eczema in early childhood, and the subsequent development of asthma and rhino-conjunctivitis. An association between eczema and asthma in childhood has been documented in several studies, and severe eczema is associated with an increased tendency to produce immunoglobulin E (IgE) and developing asthma [1–7]. The nature of the relationship between eczema and asthma has, however, been debated. Although these diseases share some genetic and environmental risk factors, it has been argued that eczema and asthma are unrelated and that the progression of eczema to asthma is due to the co-manifestation of eczema and wheezing early in life [3, 6]. On the other hand, plausible biological pathways have been described, with eczema as the first step in a progressive atopic march leading to asthma and/or rhino-conjunctivitis [8, 9]. It has also been argued that the atopic march is restricted to males only . A defective skin barrier due to mutations in the gene encoding profilaggrin/filaggrin (FLG) is associated with increased trans-epidermal water-loss, dry skin, itching and eczema . Studies on FLG mutations have not, however, found any association with FLG mutations and asthma/rhino-conjunctivitis in children who did not have eczema.
Because previous studies have been based primarily on cross-sectional data, high-risk or hospitalised patients, population-based studies with prospective designs are warranted [3, 5, 6, 14, 15]. The aim of the present study was to prospectively investigate the association between a history of eczema at 2 years of age and current asthma at 6 years of age in a general population. We also aimed to determine the prevalence of allergy-related diseases at 6 years of age according to eczema status at 2 years of age.
The Prevention of Allergy among Children in Trondheim (PACT) study is a comprehensive, controlled, primary intervention study of allergy-related diseases in Trondheim, central Norway. The study is described in detail elsewhere . Briefly, the control cohort included pregnant women and parents with their young children at ages 6 weeks, 1 year, 2 years and 6 years attending ordinary scheduled appointments with GPs, midwives or health-workers. The inclusion of 2-year-olds started on September 1st of 2000 and closed on March 31st of 2005; the inclusion of 6-year-olds started on September 1st of 2000 and closed on December 31st of 2008. In the present study, children from the PACT control cohort with cross-sectional data at 2 years were followed-up at the age of 6 years. In addition, we used information on sensitisation from a random subsample of children evaluated clinically at 2 years of age.
Parents of children at age 2 (baseline) were given questionnaires concerning environmental exposures and the family history of different allergy-related diseases in the parents and siblings. A detailed questionnaire concerning the child’s health, with an emphasis on allergy-related diseases, was also administered. An identical health questionnaire was administered when the child was 6 years of age. The questions were adapted from the International Study of Asthma and Allergies in Childhood (ISAAC) study to fit both age groups.
Both questions “Has your child ever had eczema?” and “Has your child ever had an itchy rash which was coming and going for at least 6 months?” had to be answered positively to obtain the designation of “history of eczema” at 2 years of age. To identify children who had current eczema at 6 years of age, the two former questions were combined with “Has your child during the last 12 months used any kind of medication, ointment, cream, tablets or herbal medicines against eczema?” A history of asthma was denoted by a positive answer to the question “Has your child ever been diagnosed as having asthma by a doctor?” To obtain current asthma status, the former question was combined with a positive answer to “Has the child during the last 12 months used tablets, inhalation medications or other treatments for wheezing, tightness in the chest or asthma?”
The presence of wheezing was determined by a positive answer to both questions “Has your child ever had whistling in the chest?” and “Has your child ever had episodes of wheezing or tightness in the chest?”
Positive answers to the question “Has your child ever had hay fever, sneezing or itchy-watery eyes” defined a history of rhino-conjunctivitis.
A history of an allergy test was defined as an affirmative response to the question “Has your child ever been allergy tested by skin prick test or blood test?” A positive allergy test was present if the child’s parent(s) reported a positive reaction to at least one allergen.
“Atopy in the family” was registered as present if the child’s mother, father or any siblings had answered “yes” to one of several questions regarding asthma, eczema and allergic rhino-conjunctivitis.
The term “Homeowner” was used as a proxy for socioeconomic status and was based on the question “Do you/your family own your apartment/house”.
From March 2001 to September 2002, a randomly sampled group of 720 pregnant women from the same control cohort described above were invited to attend a sub-study of PACT  . Their children were invited to a clinical examination at 2 years of age, including an assessment for eczema according to the United Kingdom Working Party (UKWP) diagnostic criteria. Disease severity was scored by the Scoring of Atopic Dermatitis (SCORAD) index and sensitisation was evaluated by standard prick test (SPT) and specific IgE (sIgE) measurements .
For SPT testing, the following standardised extracts from Soluprick® allergens (ALK Albello, Denmark) were used: mite (Dermatophagoides pteronyssinus), mould (Cladosporium herbarum), cat and dog dander, birch, timothy (grass) and mugwort pollens, hen’s egg white, codfish, hazelnut and peanut. For cow’s milk fresh skimmed milk was used. In addition, two positive histamine controls (Histamine 10mg/ml) and one diluents-negative control (NaCl) were applied on the volar surface of the child’s forearms. A mean wheal diameter of at least 3 mm larger than the negative control was considered a positive test.
Sera from venous blood samples were analysed for sIgE using assays testing for the same allergens as the SPTs. Venous sampling was attempted only once. The sIgE analyses were carried out in the immunology laboratory at St Olavs University Hospital, Trondheim, using the Immulite® 2000 Allergen-specific IgE system (Siemens Medical Solutions Diagnostics, Deerfield, IL, U.S.A.). A sIgE ≥0.35 kUL-1 was considered positive.
Sensitisation was defined as positive if the child either had a positive SPT or a positive sIgE.
Baseline characteristics as well as allergy-related diseases are described as prevalences with 95% confidence intervals (CI) for dichotomous variables.
Multivariable, logistic regression models were used to estimate the adjusted associations between a history of eczema at 2 years and current asthma at 6 years. Eczema and the age of eczema onset were used as explanatory variables. The age of eczema onset was divided in tertiles (0–3 months, 4–12 months, 13 months or older), and all other explanatory variables in the model were dichotomised. The possible confounding factors were identified by a priori knowledge, and in the final models, adjustments were made for sex, atopy in the family, wheezing, homeowner status and a smoking mother. Interaction on the multiplicative level was tested between eczema and sex and between eczema and wheezing using the likelihood ratio test. Children with a history of doctor-diagnosed asthma at age 2 years were excluded from the follow-up analyses. In addition, follow-up analyses excluding both children with doctor-diagnosed asthma and children who had ever experienced wheezing at age 2 years were performed.
Due to the large amount of missing data, we used multiple imputations (MI) to assess the potential impact of missing data in the regression analyses. We assumed that data were missing at random given the results from the observed data, and we used chained equations (regression switching) with 50 sets of imputations to impute missing values as implemented in the STATA’s ICE command. The following predictor variables reported at 2 years were included in the imputation model: eczema, age in tertiles, siblings, atopy in the family, a cat in the household, symptoms of wheezing/whistling in the chest, symptoms of hay fever, history of hospitalisation for any allergy-related diseases, homeowner status, smoking mother/father and sensitisation. In addition, the outcome variable, current asthma at 6 years, was included. Because reported sensitisation was not considered missing at random, sensitisation-data from the subsample, where all of the randomly selected participants included were tested regardless of the presence of disease symptoms, were used as predictors in the ICE command. A multiplicative interaction term between eczema and sensitisation was also included in the predictive ICE command and tested in the final models. For each outcome variable, a separate MI dataset was created, and resulting estimates were combined by the MIM command in STATA. To study the association between a history of eczema at 2 years and current asthma at 6 years, logistic regression analyses were performed on the MI dataset.
In addition to the multiple imputation approach, we also performed analyses based on individuals with complete data.
All data were analysed using STATA for Windows (version 11, College Station, Texas, USA).
All parents signed a written consent form to participate in the PACT study. The Regional Committee for Medical Research Ethics and the Norwegian Data Inspectorate Board approved the study (Ref 120–2000) (Ref 2003/953-3 KBE/-).
The PACT study has been registered in Current Controlled Trials database (ISRCTN28090297).
Baseline characteristics and prevalence of allergy-related diseases at 2 years of age
Children with follow-up data at 6 years (N=2192)a
Children without follow-up data at 6 years (N=2588)a
Atopy in the family
Breastfed ≥ 3 months
Dog at home
Cat at home
Allergy related diseases b
Allergy test performed
Prevalence of allergy-related diseases and sensitisation in children at 2 years of age
Multiple imputation (N=4780)
Prevalence of allergy-related diseases in children at 6 years of age
Children with follow-up data
Prevalence of allergy-related diseases reported at 2 and 6 years by eczema-status and sex
Reported at 2 years
Reported at 6 years
Reported eczema 2 yrs
Reported no eczema 2 yrs
Girls (n= 920)
No substantial sex differences were found in the prevalence of eczema at 2 years. Overall, however, boys reported significantly more asthma than girls both at 2 years (5.4% vs. 8.7%, p<0.001) and at 6 years (6.4% vs. 11.2%, p<0.001).
Adjusted associations between reported eczema at age 2 years and current asthma at 6 years
Atopy in family
Adjusted associations between the eczema age of onset at 2 years and current asthma at 6 years
Age of eczema onset in tertilesd
Atopy in family
The analysis using multiple imputations provided nearly the same associations as the complete case analysis. When information on sensitisation was included in the model, no substantial change in the association between eczema and asthma was found (Table 5). We repeated the analysis after removing all children who reported doctor diagnosed asthma (ever) at 2 years and children with wheezing (ever) at 2 years. The association between eczema at 2 years and asthma at 6 years was still significant with OR=2.06 (95% CI 1.09-3.90) (data not shown).
In the model with age of eczema onset categorised in tertiles, the association between age of eczema onset during the first tertile (before 4 months of age) and current asthma at 6 years did not indicate a clear trend or pattern for the age onset eczema and risk of asthma (Table 6). We found no evidence for a multiplicative interaction between eczema and sensitisation (p=0.70).
Boys had a slightly higher risk than girls of developing asthma, but the differences were not significant.
At age 2 years, 441 of 720 randomly selected children from the control cohort included since pregnancy were eligible for follow-up. Of these, 390 children (88%) participated in a clinical examination. The prevalence of a history of eczema, doctor-diagnosed asthma and wheezing reported in questionnaires at 2 years was 21.0%, 8.8% and 23.5%, respectively. The prevalence of clinically investigated eczema according to the UKWP diagnostic criteria was 15.9%.
An allergy test, either a SPT or a sIgE, was performed in 91% of the children. Of those with UKWP-Eczema, some 22.4% (95% CI 12.5-35.3.4) had a positive test, while some 14.5% (95% CI 10.7-19.0) of those without UKWP-Eczema tested positive (p=0.7).
In this large, prospective study from the general population, we found a strong association between eczema at age 2 years and asthma at age 6 years. Asthma was reported more often among boys than girls, regardless of eczema status. Four out of ten children with eczema at age 6 had the onset of their eczema after 2 years of age, but the co-existence of allergy-related diseases reported at 6 years was higher among children with an onset of eczema before 2 years of age. More than one-half of the children with eczema during the first two years of life no longer had eczema when they turned 6 years.
This comprehensive study was conducted in the primary health care setting. However, as only close to half of the included children had follow-up data at 6 years, there is a possibility of selection bias . The design of the PACT study is that of multiple, yearly, cross-sectional cohorts of children consecutively included from the year 2000. During the inclusion period, we have observed a decline in the number of participants over time. However, a non-participant study carried out among 391 parents who consecutively attended maternal postnatal care revealed no substantial differences in age, education, familial allergy-related diseases and smoking behaviour when comparing participants and non-participants . This trend is also supported by the baseline data and clinical characteristics among those children with follow-up data at 6 years compared with children with no follow-up data at 6 years. Apart from socioeconomic differences, indicated by fewer homeowners and more smoking mothers among children without available data at 6 years, only minor differences were found. This socioeconomic difference between participants and non-participants is a well-known challenge in population-based studies, and has also been reported by others [7, 18]. The low inclusion rate is most likely due to low inclusion activity among GPs, health visitors and midwives and is not likely a consequence of self-selection among parents and children.
Another possible limitation to the study is the use of questionnaires and parental reporting without a clinical verification for some of the diagnoses. The questionnaires are from the ISAAC study, and for the 2-year-olds, we adapted them to fit with this particular age-group. We have validated the eczema diagnoses among the 2-year-olds against the UKWP diagnostic criteria in another study  and the results indicate that the prevalence of eczema in this age-group may be overestimated when self-reported. The reliability of some of the questions regarding health has, however, been tested, and reported doctor-diagnosed asthma and the information obtained from medical records showed excellent agreement . In addition, mother-reported use of anti-asthmatic medications during the previous year among 7-year-old children has been shown to be highly valid . The questions on rhino-conjunctivitis have, to the best of our knowledge, not been validated for 2-year-olds, and due to the high rate of infections with symptoms resembling rhino-conjunctivitis in this age group, the prevalence of rhino-conjunctivitis reported at 2 years must be interpreted with caution.
In addition to the early manifestation of eczema, it is commonly believed that the severity of eczema, male sex, early wheezing, heredity and sensitisation are possible risk factors for the development of childhood asthma. We used information from the clinical investigative subsample, in which 91% of the children were tested either by SPT or blood sample, regardless of the presence of symptoms, to conduct multiple imputations on missing sensitisation data. In doing so, the sensitisation estimate is less precise. Another limitation regarding sensitisation is the fact that we have only tested the children once, at the age of 2 years. We might therefore have missed the children with transient sensitisation before this age. In addition, there is also a possibility that children are sensitised for allergens not tested for, such as wheat and soybean.
The strength of this study is the large number of participants, the unselected population, the prospective design, the random clinical subsample at baseline and the validation of several of the questions.
The prevalence of allergy-related diseases among 6-year-olds has been widely documented through the ISAAC study . The 12-month prevalence for eczema in phase III of the study in Sweden, Germany and the UK was 19.5%, 7.9% and 16.0%, respectively. Our finding of 13.6% is in between the prevalence reported in Germany and the U.K. but, surprisingly, is much lower than in Sweden. The estimated asthma prevalence of 5.5% is lower than in all three countries (10.2%, 12.8% and 20.9%, respectively), whereas the prevalence of rhinitis in our study is somewhat higher. The prevalence of eczema, asthma and rhino-conjunctivitis in our cohort of 6-year-olds are, however, well in line with the prevalence found in a Danish birth cohort among children of the same age .
Whether eczema is a true risk factor for asthma and rhino-conjunctivitis has been debated, and the relationship between the different allergy-related disorders is unclear. A population-based, longitudinal birth cohort study from the U.K. found that impairment in the skin barrier due to mutations in the FLG gene was associated with an increased risk of eczema, both atopic and non-atopic . In addition, patients with these mutations had more persistent eczema. FLG mutations were also associated with an increased risk of asthma, but only when co-existing with eczema [9, 13]. Suggestions were made by the authors that the sub-stratification of eczema based on FLG status could identify those children who might benefit from early therapeutic intervention. We did not, however, have the option of FLG testing for the children participating in this study.
We found that only twenty percent of children with eczema at 2 years were sensitised, whereas others have found sensitisation among children in this age group to be approximately one-third . It has been hypothesised that the course of the disease is different for those with sensitisation as opposed to those without sensitisation, and that rather than a progressive atopic march, there are different phenotypes of eczema leading to asthma [3, 24]. Others have found that FLG mutations with co-existing eczema increase the risk for asthma and hay-fever without an obligate presence of concurrent sensitisation . Our results show an increased risk of asthma at age 6 years among children with eczema at 2 years compared to children without eczema at 2 years, regardless of sensitisation.
Although most cases of eczema in primary health care are mild to moderate, the findings from this study support the hypothesis of an atopic march in the general population.
International Study of Asthma and Allergies in Childhood
Prevention of Allergy among Children in Trondheim
Scoring Atopic Dermatitis
Specific Immunoglobulin E
Skin Prick Test
United Kingdom Working Party.
The authors would like to thank all parents, children, general practitioners, health visitors and midwives for participating in the study. Special thanks to bioengineer Guri Helmersen and nurse Else Bartnes for their invaluable contribution in the PACT-study.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.