INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: asthma; atopy; cohort study; bronchial hyperresponsiveness

Asthma is one of the most common chronic disorders among children and young adults. Cross-sectional studies have shown that the prevalence of asthma and allergic disorders has increased over recent decades (1, 2) and is highest in western developed countries (3, 4). Although such cross-sectional studies are able to measure prevalence at a particular point in time and place, longitudinal cohort studies measuring incidence are required to study both the natural history and outcome of wheezy disorders. Cohort studies also have the crucial advantage over retrospective studies of avoiding recall bias.

Longitudinal cohort studies of childhood asthma are few. Previous studies differ in the population from which the cohort has been drawn, age of subjects at entry to the study, length of follow-up, and whether the study is based on questionnaire data alone or includes objective measurements of bronchial hyperresponsiveness and atopy. Clinic-based cohort studies, which suffer selection bias, have shown that 43% of 10-yr-old children with asthma have symptoms 16 yr later (5). Similarly a follow-up of school aged children in Denmark showed 86% were asthmatic 10-12 yr later (6). In general practice a 20-yr follow-up of children with asthma showed only 28% of adults to be symptom free (7).

Longitudinal population-based studies avoid the bias inherent in clinic-based studies. Two of the longest and largest studies did not include children younger than 7 yr old (8, 9), whereas a large British birth cohort study had no prospective information on children between birth and 7 yr (10). A large longitudinal study from Tuscon has provided more valuable information on the natural history of childhood asthma but has not yet reported on adult outcome (11). The natural history of allergic sensitization in childhood has been previously reported from this cohort (12) and by others (13), in particular by the German Multicentre Allergy Study (MAS) (16). Few studies have followed subjects from birth to adulthood.

This study follows prospectively a cohort of subjects at risk of atopic disorders from birth to 22 yr of age. Earlier results from the same cohort have been reported (17). The aim of this latest follow-up is to describe the course of wheeze, bronchial hyperresponsiveness, and atopy over time and to determine the outcome of childhood asthma in early adult life.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

One hundred subjects were enrolled before birth over a 12- mo period in 1976-1977. All babies were born in Poole Hospital, and were selected on the basis that one or both parents had a history of asthma and/or hayfever. Informed consent was obtained, and the local ethical committee approved the study.

The cohort was studied every 3 mo for the first year, annually thereafter until the age of 5 yr, and again at 11 yr. This study reports on the latest follow-up of subjects at 22 yr. There remained 92, 73, 67, and 63 subjects in the cohort at 1, 5, 11, and 22 yr, respectively.

Methods

During the first 5 yr, parents reported all respiratory illness and this was confirmed by clinical examination wherever possible. At the age of 11 yr children completed a questionnaire with the help of their parents. Questions covered asthma and allergic symptoms in the previous 12 mo and over the period since their last follow-up. A similar questionnaire was administered at the age of 22. Subjects were regarded as having current wheeze if they answered yes to the question "Have you had wheezing (loud breathing with a whistling sound coming from the chest) at any time in the last 12 months?"

The atopic status of subjects was determined at each follow-up by skin prick testing using a standard technique to a panel of six allergensDermatophagoides pteronyssinus, mixed grass pollen, cat fur, dog dander, hen egg, and cows' milk (Bencard UK Ltd 1976-1989, ALK-Abelló, Denmark 1999) with positive histamine and negative saline control. The reaction was regarded as positive if the mean diameter of the skin weal at 15 min (after subtracting the diameter of the weal produced by the saline control) was 2 mm or greater up to the age of 2 yr, and 3 mm or greater in subjects older than 2 yr.

Bronchial responsiveness to inhaled histamine was assessed at ages 11 and 22 using the method of Yan and coworkers (21). The provocation dose of histamine that caused a reduction in forced expiratory volume in 1 s (FEV₁) by 20% (PD₂₀) was calculated (21, 22). Subjects were considered to show bronchial hyperresponsiveness (BHR) if this occurred at or before the maximum cumulative dose of 7.8 µmol.

Asthma was defined as wheeze within the previous 12 mo and bronchial hyperresponsiveness.

Statistical Methods

Age at first wheeze was summarized using a Kaplan-Meier survival curve. Comparison between survival curves was made using the log rank test.

The possible influence of sampling error on the relationships between pairs of categorical variables was assessed with Fisher's exact test and between normally distributed variables using the independent samples t test. The relationship between pairs of binary variables is expressed as an odds ratio where appropriate. Comparisons of prevalence rates at different ages in this cohort were made using Cochran's Q test.

Data were analyzed using SPSS for Windows Version 8. A 5% significance level and 95% confidence intervals (preceded by the abbreviation CI) are used.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sixty adults were recalled for study at 22 yr, and three further subjects returned questionnaires only. Of the remaining 37 subjects 19 dropped out under the age of 2 yr. Thus, early life data were available in 81 subjects, but 18 of these could not be recalled at 22. Of the 37 subjects without data at 22 yr 1 had died (sudden infant death syndrome), 5 are known to have emigrated, and the remainder either refused to participate or could not be traced. Thirty-seven of the 63 22-yr-old subjects were female. Not all of these 63 subjects attended on all 11 occasions (birth, 3, 6, and 9 mo, 1-5, 11, and 22 yr). Complete longitudinal data on wheeze were available in 58 subjects, skin prick tests in 52, and bronchial challenge in 55 subjects. Early life data were available in 81 subjects up to the age of 2 yr, 18 of whom could not be recalled at 22. The sex distribution, social class, mean maternal age, occurrence of early wheeze, and eczema in those lost to follow-up were not significantly different from the 63 adult subjects who remained (Table 1). There was a significant difference in history of maternal asthma between those followed up and those dropping out, but it was in the opposite direction to that expected (those dropping out of the study were more likely to have a mother who had asthma). There is no obvious explanation for this and it is unlikely to lead to bias in the results. Twenty-three of 60 adults (38%) admitted to current smoking (at least one cigarette per day). There was no significant difference in the prevalence of asthma in smokers and nonsmokers.

Wheeze

The first report of wheeze could occur at any age (Figure 1) with variable patterns of remission (Table 2). The annual prevalence of wheeze showed a rise from birth to 22 yr (p < 0.001, using Cochran's Q test) (Figure 2). The increase at 2 yr was not significant. The lifetime prevalence of wheeze in the whole cohort (81 subjects) was 69% (CI 58, 80%) (Figure 1).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Age of onset of first reported wheeze (n = 81).

View larger version (40K): [in this window] [in a new window]   Figure 2.   Twelve-month prevalence of reported wheeze at each follow-up (n = 58).

The outcome of preschool wheezing was assessed by dividing the 58 subjects for whom data on wheeze were available at each follow-up into three groups according to their wheezing history during the first 5 yr: (1) never wheezed between 0 and 5 yr (55%, n = 32), (2) wheezing starting prior to 2 yr follow-up (28%, n = 16), and (3) wheezing starting between the second and fifth year follow-up (17%, n = 10).

In these groups the proportion of subjects wheezing at 11 or 22 was (1) 38%, (2) 38%, and (3) 80% (p = 0.065). Children younger than 2 yr who wheezed were therefore no more likely to wheeze in adolescence than those who had never wheezed in the preschool years, whereas those starting to wheeze between 2 and 5 yr were more likely to have later wheezing. Five subjects reported wheezing for the first time between the ages of 11 and 22.

The frequency of early wheeze was related to the persistence of symptoms. Of 32 children who never wheezed when younger than 5 yr, 12 (38%) wheezed at 11 or 22. Of 15 children who had one or two episodes of wheeze when younger than 5 yr, 4 (27%) had further wheezing beyond age 5. By contrast, of 11 children who had three or more wheezy episodes when younger than 5 yr, 10 (91%) had later wheezing (p = 0.003).

Forty (49%) of the 81 subjects had at least one parent with asthma. There was no significant difference in age of onset of wheeze in subjects with and without a parental history of asthma (p = 0.52 using log rank test).

Atopy

At 22 yr, 43 (72%) (CI 59, 83%) of the 60 subjects were atopic. Fifty-two subjects underwent skin prick tests at every time point during the study. The age of first sensitization was in infancy to ingestants (only one occurred when older than 2 yr) and throughout childhood for aeroallergens (Figure 3) (p < 0.001 using Cochran's Q test). The number of children with aeroallergen sensitization increased throughout childhood. The positive skin reactions to egg and milk tended to be transient, whereas those for airborne allergens tended to be permanent. For example, of 13 participants with a positive reaction to egg or milk at 2 yr or younger none had positive tests at every subsequent follow-up compared with 6 of 16 for airborne allergens (p = 0.02). Those children who showed sensitivity to ingestants (egg or milk) when younger than 2 yr were more likely to develop aeroallergen sensitivity, and this occurred at an earlier age than those who had not shown ingestant sensitivity (Table 3).

View larger version (23K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   Figure 3.   (A) Frequency of sensitization to food allergens (n = 52). (B) Frequency of sensitization to airborn allergens (n = 52).  Sensitization first detected in previous follow-up; sensitization first detected in this follow-up; sensitization at any follow-up.

The lifetime prevalence of sensitization to the allergens tested was house dust mite 54% (CI 39, 69%), grass pollen 54% (CI 39, 68%), cat 50% (CI 36, 64%), dog 33% (CI 20, 47%), egg 25% (CI 14, 39%), and milk 8% (CI 2, 19%).

Bronchial Hyperresponsiveness

The prevalence of BHR was 29% (CI 18, 43%) at 11 and 40% (CI 29, 54%) at 22 yr. Of 14 subjects who changed their BHR status, 10 gained hyperresponsiveness at 22 and 4 lost it (p = 0.18 using McNemar's test). Six of the 10 who became hyperresponsive when older than 11 yr also developed wheeze at this time. The 4 nonwheezy subjects showed mild BHR (PD₂₀ > 3 µmol). There was an association between wheeze at 3-5 yr and BHR at 11 yr, 50% of 18 subjects with wheeze compared with 21% of 48 subjects without wheeze showed BHR at 11 yr (p = 0.03).

Asthma

Fifteen subjects (25%) had asthma at the age of 22 yr. The majority of those shown to have asthma had mild disease. Six subjects used ₂-agonist medication alone, 2 used inhaled steroids, and 7 were not on any medication or used only very occasionally ₂-agonists. Of the 12 children with asthma at 11 yr, 7 (58%) had asthma at 22 yr compared with 8 of 33 (24%) children who did not have asthma at 11 yr (p = 0.01). Ten of the twelve 11-yr-old children with asthma (83%) were still wheezing at 22 yr; the 2 who denied wheeze still showed BHR. Four subjects with asthma developed wheeze for the first time between 11 and 22 yr. Three of these subjects with late-onset asthma were nonsmokers, all had symptoms less frequently than one per month, and only one used asthma medication (inhaled ₂-agonist alone).

Relationship among Symptoms, Atopy, and Bronchial Hyperresponsiveness

By longitudinal follow-up it was possible to divide the 58 subjects for whom data were available on each occasion into three subgroups, according to the pattern of wheezing they had shown throughout the study period (Table 2).

Group A-Subjects who never wheezed over 22 yr: 20 subjects (34%)

Group B-Subjects who wheezed only when younger than 5 yr of age: 12 subjects (21%)

Group C-Subjects who wheezed when older than 5 yr of age: 26 subjects (45%)

The allergic sensitization in early life and BHR in later life were examined in these three clinical subgroups (Table 4). Children who stopped wheezing when younger than 5 yr showed the same characteristics as those who never wheezed over the 22 yr. In contrast, there was a strong association between wheezing persisting or beginning when older than 5 yr, early sensitization, and BHR.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we have followed a cohort of subjects at risk of atopic disorders from birth to adult life. Since many wheezing disorders begin in infancy we collected information annually for the first 5 yr when susceptibility is probably determined. It is now believed that wheezing disorders in childhood are heterogeneous with different risk factors and prognoses (11, 23). How to describe such phenotypes and predict outcomes remains uncertain.

We believe that the strength of our data lies in its prospective, longitudinal nature. In any longitudinal study the cohort size suffers attrition over the years. In this study from birth 37% of the cohort were lost over the 22 yr largely due to geographic factors, but only 18% between 2 and 22 yr. We have presented data comparing those who were lost to follow-up with those who were included to demonstrate that those followed up were representative of the whole cohort. Given the sample size it is not possible to be definite that the two groups are identical, but of the six characteristics compared, five were not significantly different. The one significant difference (maternal asthma) could not easily be explained and its impact is unknown. With 63 subjects remaining in the study there is the possibility of making a type II error when statistical tests are not significant and of prevalence and associations being estimated imprecisely. Confidence intervals have been used liberally to help assess the impact of sample size on results.

In the present study there was a steady increase in the annual prevalence of wheeze in subjects older than 3 yr. These prevalence rates cannot be extrapolated to the general population, although the German Multicentre Allergy Study (MAS) (24) found that 50% of children with asthma derive from atopic families, and it is known that the prevalence of atopy in the adult population in the United Kingdom is approximately 40% (25). In a previous report we showed that most toddlers with wheeze became symptom free by 11 yr (17). We have now extended these observations to show that subjects younger than 2 yr with wheeze were no more likely to become adults with asthma than those without wheeze. These observations are consistent with the findings of premorbid respiratory function abnormalities in toddlers with wheeze (26), and the role played by viruses in precipitating the coryzal and wheezy episodes in this age group (27). In contrast, those children who started wheezing between 2 and 5 yr and who had many wheezy episodes were highly likely to wheeze in adulthood compared with children with other patterns of wheezing. Previous longitudinal studies have suggested that those young children with severe disease tend to persist, whereas the two-thirds of young children with mild disease will remit (28). Our study also suggests that the chance of lasting remission diminishes with increasing age. Those with asthma at 11 yr were highly likely to wheeze as adults; indeed, the two subjects with asthma at 11 yr but claimed no wheeze at 22 yr did demonstrate persistent BHR. Our finding that established asthma in mid-childhood may persist into adult life is in keeping with other longitudinal cohort studies (5).

We have shown how skin sensitization changes with age. There was a steady increase in the annual prevalence of atopy over the 22 yr reaching 70% for the whole cohort by adult life. This finding suggests that a single battery of skin prick tests cannot safely predict the life-long atopic status of an individual in childhood. In our cohort skin sensitization to food presented in infancy and subsequently waned, whereas aeroallergen sensitization began later and increased steadily throughout childhood. In Poole, the house dust mite was the commonest aeroallergen to cause sensitization and was demonstrable in 80% of the adults with asthma. This is not the case elsewhere in the world (29), but we know that our subjects were exposed in early life from our measurement of dust mite allergens in dust samples collected from our subjects' homes (18). Those with positive skin prick tests to ingestants in infancy showed sensitization to aeroallergens at an earlier age when compared with those who had not shown egg/milk sensitivity. Similar results are being reported from the MAS study (30).

Cross-sectional studies in older children and adults have shown an association between symptoms of asthma, atopy, and BHR (31). This longitudinal study of a population "at risk" of atopy confirms that wheeze, BHR, and allergic sensitization are strongly associated at 22 yr of age. The study also showed that in children younger than 2 yr patterns of wheeze were very variable but wheeze in children older than 2 yr of age was likely to persist. Allergic sensitization and BHR at 11 yr were unlikely to remit before adulthood.