Prognostic utility of childhood characteristics for predicting adult asthma

A range of clinical tests are available to assist clinicians to predict the likelihood that an individual will either develop or continue to have asthma. However, the presence or absence of a characteristic does not provide the same predictive ability. The aim of this analysis was to quantify the prognostic utility of childhood characteristics for predicting asthma in adulthood. Methods: In 1982 we enrolled a random sample of 718 schoolchildren aged 8-10 years and restudied a representative sample of 575 during 1997-1999. We measured airway hyperresponsiveness (AHR) by histamine inhalation test, FEV| and FVC by Vitalograph spirometry, atopy by skin prick test to eight common aero-allergens and respiratory symptoms, allergic illnesses and family history by questionnaire. The presence of asthma symptoms at age 23-25 was identified, in 1997-1999, on the basis of self-reported wheeze or sleep disturbance due to asthma or inhaled steroid use in the previous 12 months. Likelihood ratios are reported and this statistic represents the extent to which the odds of having asthma in adult life are increased, or decreased, by the presence or absence of a childhood characteristic. Results: Likelihood ratios for the presence and absence of childhood characteristics predicting asthma symptoms are shown in the table. Childhood characteristic Likelihood Likelihood Ratio Ratio (absence) (presence) Atopic (any wheal > 3mm) 2.4 0.8 AHR(PD20FEV, <3.9umol) 4.3 0.8 Wheeze within the previous 12 months 3.5 0.9 FEVI/FVC ratio < 80% 3.1 1.0 Hayfever 2.1 0.8 Conclusions: Although the presence of some childhood characteristics confer a significant risk of asthma symptoms in adulthood, the absence of any characteristic is not useful for predicting that asthma will not develop.