Asthma is a common problem worldwide and longitudinal studies of children followed up into adult life enable the assessment of clinical outcomes, examine the pattern of lung function outcomes, and importantly provide insight into aetiology and prognosis for patients with asthma. The aim of this review is to examine the major childhood asthma cohort studies which have continued into adult life, describing the strengths and weaknesses and the lessons that can be learnt regarding pathophysiology and potential future directions for research.
Asthma is one of the commonest conditions affecting the worldwide community. Monitoring the long-term patterns of childhood asthma into adult life has provided important information on clinical outcomes, highlighted lung function trends over time, enabled investigations into the aetiology and pathophysiology of asthma, measured treatment responses, and, importantly, provided prognostic information for families and their children. In the last 2 decades, there have been few prospective community studies which have explored the long-term outcome of childhood asthma into adult life. Some of the studies have followed up children to adolescence [
The lung function outcome in adult life is determined by 2 factors. Firstly lung growth occurs in children to young adult life and suboptimal growth may be potentially impacted by childhood diseases such as asthma and premature gestation. Secondly, following maximal lung growth, there is a decline that occurs in adult life and this decline may be accelerated by insults such as smoking.
The purpose of this review is to highlight the strengths, weaknesses, and lessons which can be learnt from the landmark studies of children with asthma who have been followed up into adult life.
Characteristics of longitudinal cohorts.
Age of recruitment (yrs) | Most recent followup age (yrs) | Number of original subjects | Recent followup (participation rate) | Atopy measurements | Lung function measurements (years) | |
---|---|---|---|---|---|---|
Melbourne cohort | 7 | 50 | 484 | 346 (76%) | Eczema, allergic rhinitis, skin prick tests | 7, 10, 14, 21, 28, 35, 42, 50 years |
Dunedin cohort | 9 | 38 | 1139 | 840 (81%) | Eczema, allergic rhinitis, skin prick tests | 9, 11, 13, 15, 18, 21, 26, 38 |
Tucson cohort | Birth | 22 | 1246 | 123/169 (73%) who had infantile lung function | Eczema, allergic rhinitis, skin prick tests, serum IgE | 1 ( |
Tasmanian cohort | 7 | 45–47 | 8683 | 1389 (selected cohort) | Eczema, allergic rhinitis, skin prick test | 13, 30, 45–47 |
British cohort | 7 | 45 | 14571 | 1266 (selected cohort) | Not collected | 34-35, 45 |
In 1964, the late Williams initiated a community-based, prospective study of a group of children who were born in Melbourne from the 1957 birth cohort with a history of wheezing [
As part of the original study, 401 children were randomly selected from 30,000 Melbourne primary school children and were classified as follows: control group (C), 105 children who had never wheezed; mild wheezy bronchitis group (MWB), 74 children with fewer than 5 episodes of wheezing associated with bronchitis or respiratory tract infection; wheezy bronchitis group (WB), 104 children with 5 or more episodes of wheezing associated with bronchitis or respiratory tract infection; asthma group (A), 113 children with wheezing not associated with respiratory tract infection. When the children were reviewed at age 10 years, it was realized that there were very few with severe asthma, and a further sampling was performed from the same birth cohort to establish a severe asthma group (SA), 83 children with onset of symptoms before 3 years of age, persistent symptoms at 10 years of age, and a barrel-chest deformity or a reduction of FEV0.5/FVC ratio to 50% or less.
These subjects have been reviewed at ages ten, fourteen, twenty-one, twenty-eight, thirty-five, and forty-two years and most recently at 50 years of age making this study the longest, most comprehensive followup study of childhood asthma into adult life [
The overall participation rate ranged from 76% to 90% with good representation of the original cohort across the four decades. At the 50-year followup, 64% of the MWB/WB (intermittent asthma) groups, 47% of the A group and 15% of the SA group had achieved asthma remission. In the wheezy bronchitis groups, a majority of the remissions had occurred prior to age ten (46%). In the asthma group, remission occurred most commonly between the ages of 14 to 21 years. The remission rate for the severe asthma group remained low throughout the study periods. The main childhood predictors of asthma persistence at age 50 were the severe childhood asthma group, female sex, and childhood hay fever [
The reduction in lung function that occurred in the asthma and severe asthma group was established in early childhood with a similar trend throughout adult life. In those with wheezy bronchitis in childhood, lung function was preserved and is not different from controls. Importantly, there did not appear to be an increased rate of decline in lung function in the asthma and severe asthma groups when compared to the control and wheezy bronchitis groups despite the persistence of symptoms. This decline in lung function was not affected when variables such as gender and smoking behavior at the age of 21 or 50 years were taken into account.
The other major finding at the age of 50 years was that children with severe asthma were at significantly increased risk of developing adult chronic obstructive pulmonary disease but this was not the case for those with mild childhood asthma [
The strength of this study has been the high level of participation which has been maintained throughout the last 4 decades. The comprehensive nature of testing involving consistent questionnaires, repeated lung function testing, and skin prick tests at each followup enabled accurate comparisons and detailed trends across the study periods.
A major limitation of the study is the lack of information regarding pregnancy (e.g., smoking), birth (e.g., prematurity, birthweight), and early life events (e.g., infections, eczema). The other major limitation is the fact that children in the original cohort were managed in an era when appropriate medications such as inhaled corticosteroids were not available. Therefore it would not be possible to generalize the findings to current children with asthma. Data which is collected by questionnaire is clearly subject to memory recall error which may have incorrectly defined patient wheeze phenotype and smoking behaviour. Lastly, there were inconsistencies noted in the smoking data which meant that the true impact of smoking quantity and behaviour on lung function decline had to be interpreted with caution.
This study was initiated as a New Zealand birth cohort study which has followed up children till the age of 38 years [
In this study, wheeze phenotype was defined as follows: persistent wheezing, wheezing reported at every assessment after it is first mentioned; remission, absence of wheezing after wheezing had been reported at two or more successive prior assessments; relapse, wheezing reported at two or more successive assessments and then absent at one or more successive assessments and then reported at all subsequent assessments; intermittent wheezing, presence of symptoms at some assessments but not the others; transient wheezing, wheezing reported at one assessment only.
In the initial assessment at 9 years of age, 815 study members (78.6% of the cohort of 1037) completed respiratory questionnaires. By the age of 26 years, 613 (59.1%) of the study members had provided respiratory data at every assessment. 27.4% of the study members had never reported wheezing at any time periods and 72.6% of the study members had reported wheezing during at least one assessment. 26.9% of the study members were currently wheezing at age 26 years with a majority either in remission or reporting intermittent wheezing. 12.4% had had a remission followed by a relapse by the age of 26 years. In 14.5%, wheezing had persisted from onset.
The risk factors that promote persistence of wheeze included being female, having a higher prevalence of sensitivity to House Dust Mite (HDM) and cat allergen, smoking at 21 years, and airway hyperresponsiveness. Those with asthma persistence and relapses had lower lung function measurements (expressed as the ratio of FEV1/FVC) at each assessment during childhood, adolescence, and adulthood compared to those who have never wheezed. There was also no statistical difference in the slopes of change in the FEV1/FVC ratio over time in any outcome category.
At the most recent followup at 38 years, 840 (81%) subjects were recruited and the cohort were reclassified to the following groups with childhood-onset persistent asthma, late-onset asthma, asthma in remission, and being nonasthmatic. Cumulative smoking history was associated with lower FEV1/FVC ratios among subjects without asthma at age 38 years and those with late-onset or remittent asthma but smoking was not associated with lower FEV1/FVC ratios among those with childhood-onset persistent asthma.
This was a major unselected, population birth cohort study conducted and the sample at 38 years was representative of the original birth cohort. This was supported by the findings that there were no significant differences in sex ratio, family history of asthma and hay fever, symptoms, proportion with atopy, and lung function measurements between the participants and the original cohort of 1037. Identical challenge testing with methacholine was used consistently at each review to the age of 26 years, as a measure of bronchial hyperresponsiveness which enabled consistent comparison between groups at different time intervals. The findings of increased bronchial hyperresponsiveness in the persistent and relapsing groups suggest that active inflammation may be potentially occurring in the airways and that this was a risk factor for asthma persistence beyond childhood years. This study highlighted that children with more severe wheeze phenotypes (persistent wheezing) had reduced lung function when first measured at age 9 years, with no apparent increase in rate of decline when compared to all the other groups. At age 38 years, there was no evidence that cumulative smoking had a synergistic effect in the lung function decline in those with childhood persistent asthma.
The histories of wheezing in early childhood were obtained from the parent when the children were already nine years of age. This raises the possibility of recall error such as underestimated reports of early wheezing. There was no description of pregnancy and early birth events in the studies either. Clearly, by the age of nine years, some children with early wheeze may have already resolved by the time when the respiratory questionnaires were conducted. The definition of remission is also subject to error as some subjects may be wheeze-free due to medication use which was not elaborated in the definition. By the age of 26, complete analysis was carried out on 59.1% of the subjects who had measurements at each study time period which may potentially lead to selection biases, although the participation rate of 81% at age 38 years is a marked improvement.
This study recruited 1246 newborns born between May 1980 and October 1985 and has followed up the children into young adulthood. At the time of enrolment, the parents completed a questionnaire about their history of respiratory illness, smoking habits, and education. Parents completed a questionnaire during their child’s second and sixth years of life. During the first year of life, 169 infants underwent pulmonary function testing involving partial expiratory flow-volume curves obtained by the chest compression technique. This technique measured the
Skin prick testing was performed concomitantly with lung function testing at the time of the six-year survey with extracts of seven common aeroallergens. Total serum IgE levels were measured with the paper radioimmunosorbent test in samples obtained from cord blood, from blood obtained at a median age of 9.3 months, and from blood obtained at the time of the 6 year survey.
Wheeze phenotype at the assessment at 6 years of age was defined as follows: children who never had wheezing, no recorded lower respiratory tract illness with wheezing during the first three years of life and no wheezing at six years of age; transient early wheezing, children with at least one lower respiratory tract illness with wheezing during the first three years of life but who had no wheezing at six years of age; transient early wheezing was a term which was uniquely coined to reflect children with wheezing in early years of life which had ceased by 6 years of life; wheezing of late onset, those who had no wheezing during the first three years of age but who had wheezing at six years of age; persistent wheezing, those who had at least one lower respiratory tract illness with wheezing in the first three years of life and had wheezing at six years of age.
In the 6th year analysis [
This birth cohort has been recently followed up to early adult life at the age of 22 years. Children who had infant lung function in the lowest quartile also had lower prebronchodilator values for FEV1/FVC ratio, FEF25–75 and FEV1 up to age 22 years. Results in this study suggest that poor airway function shortly after birth should be recognized as a risk factor for airflow obstruction in young adults and these subjects may reach the threshold of FEV1 and FEV1/FVC ratio that define chronic obstructive pulmonary disease (COPD) at an earlier age [
This is the longest prospective birth cohort study which has followed up children from birth to young adulthood. The Tucson cohort has also been instrumental in defining three key phenotypes of asthma: transient wheezing of infancy, nonatopic wheezing, and IgE mediated wheezing. The uniqueness of lung function measurements from the 1st year of life to young adulthood has demonstrated that lung function was partly determined in early life, therefore highlighting the possibility of in-utero effects and the risks associated with maternal smoking in pregnancy. The most recent followup in young adulthood also reports the possibility that adult chronic obstructive pulmonary disease (COPD) may have its origins in early life.
The participation rate of the subjects has decreased throughout the study periods which raise the possibility of selection bias. Reports on symptoms are based on questionnaires which are predisposed to memory recall error. There are inherent limitations in the measurement of
In 1968, the parents of all 8683 children born in 1961 and attending school in Tasmania were asked to complete questionnaires on their children’s history of respiratory symptoms and asthma. The survey aimed to study the natural course of asthma and returns from 99% of children were received [
From 2002 to 2005, 7312 (85.2%) of the original participants were sent a detailed respiratory questionnaire which was completed by 5729 (78.4%) of subjects. A sample of respondents enriched for asthma and chronic bronchitis participated in lung function testing (including pre- and postbronchodilator spirometry) and skin prick testing as a measurement for atopy from 2006 to 2008 [
In the 1991–93 followup study, questionnaires were returned by 1494 subjects representing 74.7% of the total sample (75.1% of those with childhood asthma and 75.3% of those without). Interestingly, almost half of the subjects gave responses that contradicted those of their parents in 1968 with regards to onset of asthma by age of 7. Of the subjects with asthma or wheezy bronchitis by the age 7, as reported by their parents 25.6% reported current asthma as an adult. 75% of the children with asthma reported by their parents in 1968 were free of symptoms as adults, consistent with asthma remission. The risk factors measured at the age of 7 that independently predicted current asthma as an adult were being female, having a history of eczema, having a low mid forced expiratory flow, having a mother or father with a history of asthma, having childhood asthma, and having the first attack by the age of 2 or having had more than 10 attacks [
In the 2002–05 followup study, the authors report that childhood infectious diseases (such as rubella, chicken pox, and pertussis) appear to protect against asthma persisting in later life, although pertussis and measles were associated with new-onset asthma after childhood [
There was a high participation rate (75%) at the followup study, with large numbers of subjects, which was representative of the birth cohort and limits response bias. This study has also highlighted the key risk factors of parental asthma, frequent childhood asthma episodes, and rhinitis as determinants of asthma persistence from childhood to adult life. This study has also provided helpful information to suggest that rhinitis in childhood is a risk factor to the development of new-onset adult asthma [
There were significant differences in the parents’ report of symptoms when the children were 7 years of age with the subjects’ report in the followup study indicating the error of recall and unreliable nature of retrospective self-assessment of childhood asthma in adults [
This birth cohort study of Perinatal Mortality started in 1958 with a focus on 18,559 births in a single week [
5,801 (31%) subjects contributed data at ages 7, 11, 16, 23, and 33 years. Persistence of wheezing at each followup was assessed by responses indicating one or more attacks of asthma or wheezy bronchitis in the previous year. The cumulative incidence of wheezing illness was 17% by age 7, 24% by age 16, and 43% by age 33. Incidence during childhood was strongly associated with pneumonia, hay fever, and eczema. Relapse at 33 after prolonged remission of childhood wheezing was more common among current smokers and atopic subjects. A quarter of children with a history of asthma or wheezy bronchitis by age 7 reported wheeze in the previous year at the age of 33.
From 1992 to 1993, lung function with postbronchodilator responses was performed in 1060 subjects with a history of asthma, wheezy bronchitis, or wheezing and 275 controls. The adults were aged between 34 and 35 years. Among 551 cases reporting no wheeze in the year before examination, ventilatory function did not differ significantly from the control, except for FEV1 in 192 subjects with wheezing before the age of 7. Among 509 cases reporting wheeze in the past year, FEV1 and FEV1/FVC ratio were reduced to a greater extent in those with an earlier age of onset of wheeze highlighting the importance of childhood asthma as a determinant of adult lung function [
The main strengths of the study include the largest prospective followup study of children from a birth cohort and a thorough assessment of new incident cases of asthma in adult life describing smoking and atopy as major risk factors. The British cohort also described smoking to be associated with an increase in the rate of decline in lung function, which was one of the few longitudinal studies to describe this additional decline in lung function.
The response rate at the 33-year-old study had dropped to 31% reflecting significant loss over time primarily by relocation. The definitions of asthma and wheezy bronchitis were based on the term “attacks” which may have underestimated subjects with “wheeze” alone. Lung function was firstly measured at age 33 years with no prospective measurements made in childhood. There were also no objective markers of atopy throughout the study limiting its assessment.
The clinical and lung function outcomes of adults with asthma appear to be partly established in childhood years. Children with mild asthma are more likely to achieve remission and those in the severe phenotype are more likely to have ongoing symptoms into adult life. Some of the main risk factors that predispose children to persistence of asthma symptoms into adult life include female sex, parental history of asthma, and atopy in particular allergic rhinitis.
There have been studies that describe increasing prevalence of asthma in females from adolescence into adulthood [
It is interesting to speculate on the “one airway, one disease” hypothesis as a basis to explain the association between allergic rhinitis and asthma [
Lung function data from the Melbourne, Dunedin, and Tucson cohorts indicates that children with more severe asthma achieve suboptimal lung growth. The comprehensive lung function measurements in the Melbourne and Dunedin cohorts demonstrate a parallel decline in lung function across the severe asthma groups which was not different when compared to the control and mild asthma groups. This finding is contrary to publications from adult based asthma cohorts. The adult Busselton study was a multiple cross-sectional study where asthma was acquired at any time during the study period. The reduction in lung function that occurs with the onset of asthma during a longitudinal study would influence the rate of decline over the length of the study period [
The results of the Melbourne cohort suggests that children with severe asthma or those with deficits in lung function from childhood are clearly at increased risk of fulfilling the diagnosis of adult chronic obstructive pulmonary disease [
Given the findings of the childhood asthma cohorts, it would support the focus of studies in the preschool years which appear to be the period when changes of airway remodeling might be occurring. Future research needs to continue into understanding the mechanisms that precede airway remodeling. Ideally, airway epithelial biopsies would provide the most information but given the invasive nature of such a procedure, further noninvasive tools need to be developed. High resolution CT (HRCT) scanning in adults has shown good correlation between spirometric indices of airflow obstruction, reticular basement membrane thickness, and HRCT measurements of airway wall thickness [
Lastly, other noninvasive techniques to measure airway inflammation and biomarkers should be explored in preschool children. Exhaled breath condensate (EBC) is a simple tool which may be of use in preschool children abut at this stage; no useful biomarkers for remodeling have been identified. Salivary proteins may also be a useful tool in determining candidate proteins that identify children who may have an atopic tendency and predisposition to asthma.
The author declares that there are no competing interests regarding the publication of this paper.