The lungs were not the target

Research presented at the European Respiratory Society congress in September 2026 found that patients with chronic airway disease who were prescribed semaglutide had up to 40 per cent fewer asthma attacks and about 20 per cent fewer COPD flare-ups than comparable patients prescribed sulfonylureas. The work was led by Chloe Bloom, clinical associate professor in respiratory epidemiology at the National Heart & Lung Institute, Imperial College London, and presented by Bohee Lee.

The design was four parallel analyses of UK electronic medical records, each covering between 20,000 and 22,000 people starting either a GLP-1 receptor agonist or a sulfonylurea. Using another diabetes drug as the comparator is a deliberate choice: it compares treated patients with treated patients, which removes some of the confounding that dogs comparisons against no treatment.

The within-class result is the interesting one

Liraglutide, dulaglutide and exenatide — all GLP-1 agonists — showed no association with reduced exacerbations, across doses and respiratory phenotypes. Only semaglutide did.

That heterogeneity is what raises the finding above the usual noise. If the benefit were simply a consequence of weight loss, or of being the sort of patient who gets prescribed a modern drug, one would expect the whole class to move together. A difference confined to one molecule points either to something specific about it — potency, receptor kinetics, a distinct anti-inflammatory effect — or to something specific about who receives it, which is the very confounding an active comparator is meant to suppress and does not fully.

What would settle it

A randomised trial with exacerbations as a pre-specified endpoint, stratified by weight change so the mediation question can be answered rather than argued. Until then this is a well-constructed hypothesis with a large denominator, which is a good thing to have and not the same thing as a result.