Asthma definitions, relative validity and impact on known risk factors in young Brazilians

BACKGROUND: An asthma score was proposed in the European Community Respiratory Health Survey (ECRHS) framework, as dichotomous definitions could be less appropriate in the study of chronic diseases. The aims of this study were to assess the value of this asthma score in comparison with other definitions of asthma in another population setting, using as criteria bronchial hyperresponsiveness (BHR) to methacholine and diagnosed asthma, and the association of these definitions to known risk factors of asthma. METHODS: We used the ECRHS questionnaire on 2063 Brazilians, aged 23-25 years, and measured their BHR. We assessed the positive and negative likelihood ratios (PLR and NLR) of the asthma score (0-8), a three question score (ECRHS definition) and single asthma symptoms in relation to BHR and diagnosed asthma. RESULTS: The PLR were relatively low for all asthma definitions with odd ratios varying from 1.47 for asthma score to 5.50 for wheeze and waking with breathlessness without a cold. The NLR were near 1. The PLR were lower for assessments using the score than for dichotomous assessments or the ECRHS definition. The PLR increased with asthma scores, but the prevalence with higher scores was too low for useful analysis. The asthma score was slightly better for identifying associations from a set of known risk factors than the other two approaches. CONCLUSION: Our study provided little support for a greater validity of this asthma score over other asthma definitions, and only marginal advantage for identifying risk factors.