Eosinophil-associated protein networks differ between COPD and asthma in UK Biobank

A study published in Allergy examined associations between blood eosinophils and incident COPD, asthma, and asthma–COPD overlap in 302,370 UK Biobank participants. Researchers used multivariable-adjusted Cox models, bidirectional two-sample Mendelian randomization, and baseline plasma proteomic data from 32,121 participants to investigate disease risk and potentially distinct inflammatory pathways.

Over approximately 13.5 years, higher eosinophil counts were associated with increased risks of all three conditions. Unlike asthma, pre-existing COPD did not predict subsequent eosinophil elevation. Mendelian randomization supported an eosinophil-to-COPD effect and bidirectional effects for asthma. Proteomic analyses identified a broader interaction network for COPD, comprising 378 proteins, compared with 70 for asthma. COPD-associated proteins mapped to a wider inflammatory network, whereas the asthma network was mainly linked to Type 2 inflammation.

Exploratory mediation analyses estimated that proteomic signatures statistically accounted for 87.8% of the eosinophil–COPD association (p = 0.076) and 59.1% of the eosinophil–asthma association (p < 0.001). These analyses assumed an eosinophil-to-protein direction, and the COPD mediation result did not meet conventional statistical significance. The findings suggest that eosinophilic COPD should not automatically be equated with asthma-like Type 2 inflammation, but do not establish treatment implications. Assessment is limited to the abstract; effect sizes for disease-risk associations were not provided.