| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Dr. Weissman is Senior Medical Officer, National Institute for Occupational Safety and Health, Health Effects Laboratory Division.
Correspondence to: David N. Weissman, MD, FCCP, Senior Medical Officer, National Institute for Occupational Safety and Health, Health Effects Laboratory Division, Mailstop L-4218, 1095 Willowdale Rd, Morgantown, WV 26505; e-mail: Dweissman{at}cdc.gov
How should asthma be defined in population studies? The question is deceptively simple, and its answer remains elusive. Since questionnaires are the most practical tools to use in screening populations for asthma, much attention has focused on developing survey definitions of asthma based on questionnaires. In general, the approach to validating such definitions has been to assess the ability of individual questions and combinations of questions to predict which individuals in a population have either clinical diagnoses of asthma or nonspecific bronchial hyperreactivity (BHR) to agents such as histamine or methacholine.1 Unfortunately, physicians’ diagnoses of asthma and BHR are not particularly good "gold standards" for identification of asthma. It is likely that a physician’s diagnosis of asthma underdetects subclinical mild asthma. Thus, using it as a "gold standard" will tend to underestimate the specificity of a questionnaire. In contrast, BHR is present in many people without asthma.1 2 3 Therefore, use of BHR as a "gold standard" will underestimate sensitivity.
Recognizing these limitations, many studies1 4 5 6 have assessed the ability of questionnaires to predict a physician’s diagnosis of asthma and/or BHR. In general, questions about ever having asthma, ever having asthma diagnosed by a physician, and having wheezing during the previous 12 months have been the questions with best sensitivity and specificity for prediction of the flawed "gold standards." Thus, responses to these questions are often used in survey definitions of asthma. In this issue of CHEST (see page 135), Ponsonby et al report a study suggesting that evaluation of severity can be used to classify into subsets individuals identified by questionnaire to have symptoms of asthma or the disease itself. The study is a cross-sectional survey for asthma conducted in 1999, evaluating children aged 8 to 10 years from randomly selected schools in the Australian Capital Territory. Children of the same age presenting in 1999 to the only three hospitals in the Australian Capital Territory able to manage acute pediatric asthma were also evaluated. For all of these children, asthma was identified using a questionnaire and atopy by a panel of allergy skin tests. Among those reporting wheezing in the previous 12 months, a stronger relationship was noted with atopy for those reporting > 12 episodes of wheezing in the last 12 months than for those reporting 1 to 3 episodes in the last 12 months (odds ratios [ORs], 8.70 vs 3.27, respectively). Atopy was also found to be more strongly associated with 1999 hospital attendance for asthma than with ever having had asthma (ORs, 16.95 vs 2.09, respectively). The proportion of "asthma-ever" attributable to atopy was 33%, while for hospital attendance in 1999, this proportion was 89%. Based on these findings, the authors suggest that atopy contributes more to frequent or severe asthma than to infrequent or mild asthma.
These findings are consistent with those of other studies. The important association of atopy with childhood asthma is well accepted.7 A review8 of studies relating atopy to asthma notes that in cross-sectional studies conducted exclusively or predominantly in children, the proportion of cases attributable to atopy varied from 25 to 63%, with a weighted mean of about 38%. Previous studies8 have also suggested a relationship between atopy and asthma severity. Atopy is also related to degree of BHR.9 10 Conversely, in patients having wheeze in the previous 12 months, BHR is related to both atopy and measures of disease severity such as peak flow variability.11
Thus, it has become increasingly apparent that populations identified by survey definitions of asthma based on self-report of asthma or asthma symptoms are a heterogeneous population. This population can be further subdivided into more homogenous subsets. Those with mild or inactive disease are less likely to be atopic or exhibit BHR. In contrast, those with more severe disease are more likely to be atopic and exhibit BHR. It has already been proposed that measurement of BHR can be used in combination with questionnaire responses to define subpopulations of asthmatics.3 11 Perhaps it will also prove useful to define subpopulations based on severity of disease using questions such as those in the wheezing module of the International Study of Asthma and Allergies in Childhood questionnaire.12 This approach has already been applied to evaluation of asthma prevalence, documenting that increases in the prevalence of asthma diagnosis and symptoms in Sheffield, UK, between 1991 and 1999 were confined to mild symptoms.13 Identification of more homogeneous asthmatic subpopulations should also facilitate population studies addressing issues such as asthma pathogenesis and effectiveness of preventive interventions such as allergen avoidance.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
