Asthma survey items as predictors of respiratory problems in children 2 yrs later: a longitudinal study

The study compared the ability of characteristics defined by an asthma survey (wheeze versus cough and asthma diagnosis versus no diagnosis) to predict later respiratory problems in a cohort of 108 schoolchildren who had reported either recent wheeze or recurrent cough in a 1987 asthma survey. The children recorded daily respiratory symptoms and peak flow from April 1989 until May 1990. The frequency and severity of lower respiratory symptom episodes and peak flow dips were compared in the wheeze and cough groups and in the diagnosed versus nondiagnosed children. The independent effects of initial wheeze, atopy, diagnosis and bronchial hyperresponsiveness (BHR) on the longitudinal outcome measures were assessed using multiple linear regression. Children with initial wheeze had more chronic symptoms and peak flow variability than those with cough alone, but wheeze had only a weak effect on frequency and severity of acute lower respiratory episodes. Children with both wheeze and atopy had more acute symptomatic episodes and more chronic symptoms than did the other children. Children with diagnosed asthma (versus no diagnosis) had significantly more frequent and severe lower respiratory exacerbations, more days symptomatic and greater peak flow variability. The predictive effects of diagnosis were independent of (and stronger than the effects of) wheeze, atopy and BHR, or combinations of these variables. The results suggest that among children who report respiratory symptoms, survey-reported wheeze on its own is a weaker marker of significant respiratory disease than is a doctor's diagnosis of asthma.