Early-life prediction of school-age asthma among recurrent Wheezers in preschoolers: The WHEEP study.

Abstract

Background: Early identification of childhood asthma is critical to prevent over- and undertreatment. Previous asthma prediction models often rely on subjective outcomes, limiting generalizability. This study aimed to develop an asthma prediction model using objective measures.

Methods: This study was embedded in a multicenter prospective cohort of children aged 1-4 years with recurrent wheezing presenting at pediatric secondary care clinics who were followed up for 12 months and re-assessed between ages 6 and 12 years. Candidate predictors included demographic, environmental, and asthma-related factors. Asthma at school age was defined as at least two of the three criteria: asthma medication use, spirometry with bronchodilator reversibility, and elevated fractional exhaled nitric oxide levels. LASSO regression was applied to identify the most predictive variables in an optimal and parsimonious model, and model performance was evaluated.

Results: Of 144 children with recurrent wheezing at age 1-4 years, 25% were classified as having asthma at school age. The optimal model for predicting asthma at school age demonstrated good predictive performance (area under the curve (AUC) (95% CI): 0.83 (0.73-0.93), sensitivity: 0.69, specificity: 0.86, positive predictive value (PPV): 0.66, and negative predictive value (NPV): 0.88). The parsimonious model showed slightly lower performance (AUC (95% CI): 0.82 (0.71-0.94), sensitivity: 0.67, specificity: 0.86, PPV: 0.65, NPV: 0.87).

Conclusion: Our models highlight the importance of using objectively defined asthma in early prediction of asthma at school age, as demonstrated by their high discriminative performance. Future studies should validate these models and incorporate clearly defined objective outcome measures when developing new models for pediatric secondary care clinics.