Cluster analysis of phenotypes, job exposure, and inflammatory patterns in elderly and nonelderly asthma patients

Abstract

Background: Asthma has been identified as different phenotypes due to various risk factors. Age differences may have potential effects on asthma phenotypes. Our study aimed to identify potential asthma phenotypes among adults divided by age as either younger or older than 65 years. We also compared differences in blood granulocyte patterns, occupational asthmagens, and asthma control-related outcomes among patient phenotype clusters. Methods: We recruited nonelderly (<65 years old) (n = 726) and elderly adults (≥65 years old) (n = 201) with mild-to-severe asthma. We conducted a factor analysis to select 17 variables. A two-step cluster analysis was used to classify subjects with asthma phenotypes, and a discriminant analysis was used to verify the classification of cluster results. Results: There were three clusters with different characteristics identified in both the nonelderly and elderly asthmatic adults. In the nonelderly patient group, cluster 2 (obese, neutrophilic phenotypes) had a 1.85-fold significantly increased risk of asthma exacerbations. Cluster 3 (early-onset, atopy, and smoker with an eosinophil-predominant pattern) had a 2.37-fold risk of asthma exacerbations and higher oral corticosteroid (OCS) use than cluster 1 (late-onset and LMW exposure with paucigranulocytic blood pattern). Among elderly patients, cluster 2 had poor lung function and more ex-smokers. Cluster 3 (early-onset, long asthma duration) had the lowest paucigranulocytic blood pattern percentages in the elderly group. Conclusions: The novelty of the clusters was found in age-dependent clusters. We identified three distinct phenotypes with heterogeneous characteristics, asthma exacerbations and medicine use in nonelderly and elderly asthmatic patients, respectively. Classification of age-stratified asthma phenotypes may lead to precise identification of patients, which provides personalized disease management.

Keywords

• Asthma
• Cluster analysis
• Elderly
• Inflammatory patterns
• Job exposure