METHODS

The subjects were participants in the follow up phase of the ECRHS conducted in Melbourne in 1998/9. Full details of the original sampling protocol have been described elsewhere.⁴ Participants completed the detailed ECRHS questionnaire, spirometric tests, and a methacholine challenge. QoL was evaluated by the short form (SF-36) health survey which was completed by participants upon arrival at the laboratory. All questionnaires were checked for missing data by one of the trained interviewers after completion. Only the physical component summary (PCS) score and the mental component summary (MCS) score are reported in this analysis which were calculated using the three step procedure recommended by the developer.⁵ A total of 426 participants completed the methacholine challenge and were included in this analysis.

Wheeze only was defined as a positive response to the question: “Have you had wheezing or whistling in your chest at any time in the last 12 months?”.¹ The ECRHS defined asthma as a positive response to any of the following questions: “Have you had an attack of asthma in the last 12 months?”, “Have you been woken by an attack of shortness of breath at any time in last 12 months?”, and “Are you currently taking any medicine for asthma?”.⁴ BHR was defined as a provocative dose of <2 mg methacholine causing a 20% fall in FEV₁ (PD₂₀). Current asthma was defined as a positive response to the question: “Have you had wheezing or whistling in your chest at any time in the last 12 months?” and measured BHR.¹ Doctor diagnosed asthma was defined as a positive response to the question: “Have you ever had asthma?” and then to the question: “Was this confirmed by a doctor?”. BHR alone and FEV₁ % predicted were used as purely physiological definitions of asthma.

Comparisons in QoL scores were made between subjects meeting the various definitions of asthma and the remaining subjects who did not meet the criteria. The Mann-Whitney U test was used to test if the distribution of the SF-36 scores was the same across the different definitions and to assess whether SF-36 scores were associated with sex, age, and smoking status. A multivariate regression model was used to test for significance between cases and controls after adjusting for age, sex, and current smoking status. Spearman's rank correlation was used to assess the association between FEV₁ % predicted and SF-36 scores. A p value of <0.05 was considered significant. All analyses were conducted using Stata (Stata for Windows, Stata Corporation 1997, Texas, US). Post hoc power calculations indicated that there was a greater than 80% power to detect a three point difference in PCS and MCS scores between groups.

RESULTS

Participants had a mean (SD) age of 39.7 (6.4) years (range 26–50), 50.0% were women, and 81.5% were Australian born. The mean (SD) FEV₁ was 109.3 (14.5)% predicted. Eighty five participants (25%) were current smokers and 214 (50.2%) reported ever smoking. Doctor diagnosed asthma was reported by 120 participants (28.2%) and an attack of asthma in the previous 12 months was reported by 71 (16.7%).

The key asthma symptom of wheeze in the preceding 12 months was reported by 177 participants (41.5%). Nocturnal shortness of breath (SOB) and spontaneous SOB were reported by 43 (10.1%) and 53 (12.4%) participants, respectively. BHR was demonstrated by 106 participants (24.9%). The number of participants who reported that they were currently on medication for their asthma was 145 (34.0%). While 38 (8.9%) reported seeing a GP specifically because of their breathing, only eight (1.9%) reported seeing a specialist physician.

In this population of young adults, sex, age, and current smoking status were not important determinants of QoL scores. SF-36 summary scores for the alternative definitions of asthma are presented in table 1⇓. The PCS and MCS scores were both significantly worse in those with wheeze in the preceding 12 months than in those without wheeze. Similarly, for the ECRHS definition of asthma both the PCS and MCS scores were significantly different between the cases and controls. For doctor diagnosed asthma only the PCS score was significantly worse in adults with self-reported doctor diagnosed asthma than in those without. In contrast to these results, for current asthma neither the PCS nor MCS scores were significantly worse in those with current asthma than in those without. For those with BHR only neither PCS nor MCS scores were significantly different between cases and controls. The rank correlations between FEV₁ % predicted and the PCS and MCS scores were not significant (PCS ρ=0.3, p=0.06; MCS ρ=0.2, p=0.4).

DISCUSSION

This is the first study of which we are aware to investigate specifically the relationships between different epidemiological definitions of asthma and reported QoL using the SF-36 questionnaire in a community based sample of young adults. There is some controversy over which symptom questions are the most valid in diagnosing asthma for epidemiological studies.² We have found that subjective definitions of asthma such as wheeze in the previous 12 months, the ECRHS definition of asthma, and self-reported doctor diagnosed asthma define individuals with significantly worse QoL. On the other hand, the epidemiological definition of current asthma and purely physiologically based definitions of asthma such as BHR alone or FEV₁ were not associated with significantly worse QoL. These results would suggest that merely having respiratory symptoms of any type is sufficient to adversely affect QoL and, in the context of this particular population, the underlying problem was likely to be asthma.

Data from the French centres of the ECRHS found that subjects with asthma had significantly lower PCS and MCS scores than control subjects using a combination of self-reported symptoms and BHR to define cases.⁶ Several other studies have reported strong correlations between poor QoL and subjective measures of asthma severity such as diary obtained symptom scores and β₂ agonist use for the relief of symptoms.^3,7 On the other hand, correlations between objective measures such as FEV₁ have been found to be poor.⁸ Our results are consistent with these studies and show that, even for individuals with mostly mild asthma from a community setting, QoL is more strongly associated with self reported “subjective” than objective measures of asthma.

Clinical observations suggest that patients' concerns with regard to their asthma tend to focus on symptom frequency, activity limitation, and avoidance of irritants.⁹ For subjects with mild asthma who experience few symptoms QoL is equivalent to or better than population norms but when they experience symptoms their QoL is significantly affected,¹⁰ while individuals with more severe asthma and more frequent respiratory symptoms who have accommodated their lifestyle to this chronic condition might perceive less impact on QoL. This would suggest that most people tend to focus on subjective rather than objective measures of their asthma and, given that QoL is a self-assessed measure, it is not surprising that subjective definitions of asthma correlate more closely than do objective measures of asthma with poor QoL.

In conclusion, we have confirmed that symptom based definitions are more closely related to QoL than are physiological definitions of asthma. We suggest that, for epidemiological studies investigating the social, psychological, and economic costs of asthma, it might be more appropriate to use symptom based definitions when defining subjects. While our conclusions are valid for epidemiological studies of asthma in the community, they should not be generalised to clinical trials of patients with more severe disease. However, we do suggest that in clinical trials QoL measures would provide clinicians with valuable further information.