BACKGROUND There has been no large-scale and comprehensive study of the differences between asthma in elderly asthmatics (EA) and non-elderly (i.e. young) asthmatics (NEA).

METHODS We performed principal component analysis (PCA) using 2067 asthmatics (434 EA and 1633 NEA) from the Korean Cohort for Reality and Evolution of adult Asthma (COREA). EA was defined as asthmatics with the chronological age of 65 or more and eleven clinical variables measured at enrollment were used for PCA; symptom score, symptom duration, number of exacerbation during previous one year, smoking pack year, number of controller medications, body mass index, predicted % of FEV1, predicted % FVC, post-bronchodilator FEV1/FVC ratio, atopy index and number of eosinophils in peripheral blood.

RESULTS PCA of all asthmatics showed that EA and NEA were distinctly separated by the first and second principal component on the plot of individual asthmatics according to their scores. For further analysis, we divided all asthmatics into the EA and the NEA group and performed PCA again in each group. The first four principal components with eigenvalues ≥ 1.0 were identified in both groups and they explained 55.5% of the variance in the EA group and 52.4% in the NEA group respectively. Clinical variables showed distinctly different patterns of loading on the first four principal components between the EA and the NEA group.

CONCLUSION EA and NEA have different compositional patterns underlying their clinical variables. These observations helped in understanding the differences between EA and NEA from the integrated view covering various clinical aspects.