Identifying adult asthma phenotypes using a clustering approach

Area of Science:

• Pulmonary Medicine
• Epidemiology
• Biostatistics

Background:

• Asthma characterization requires integrating diverse disease aspects.
• Current classifications may not fully capture asthma heterogeneity.

Purpose of the Study:

• To identify distinct asthma phenotypes using latent class analysis (LCA).
• To apply LCA to large epidemiological cohorts (EGEA2 and ECRHSII) for robust phenotype identification.

Main Methods:

• Latent class analysis (LCA) applied to 19 variables.
• Variables included personal characteristics, symptoms, treatment, onset age, allergy, lung function, and airway hyperresponsiveness.
• Analysis conducted on two large adult asthma cohorts: EGEA2 (n=641) and ECRHSII (n=1,895).

Main Results:

• Four distinct asthma phenotypes were identified in both EGEA2 and ECRHSII cohorts.
• Consistent phenotypes included: active treated allergic childhood-onset asthma and active treated adult-onset asthma.
• Other phenotypes comprised inactive/mild untreated asthma, differing by atopy and onset age.
• Phenotypes effectively discriminated quality of life and blood eosinophil/neutrophil counts.

Conclusions:

• Latent class analysis successfully delineated four distinct asthma phenotypes.
• These identified phenotypes offer more homogeneous patient groups for future research.
• Utilizing these refined phenotypes may improve the identification of asthma risk factors and guide personalized treatment strategies.