Twenty-three adults with cystic fibrosis and diabetes started a GLP-1 drug, and over the next six months their lung function rose by a median of 6 percentage points. The part that stands out: the improvement held even after the researchers accounted for the weight the patients lost.

The finding comes from a retrospective cohort study ↗ published September 2 in Respiratory Medicine by a UT Southwestern Medical Center team led by endocrinologist Sasan Mirfakhraee. The authors span the Division of Endocrinology and the Division of Pulmonary and Critical Care Medicine. It is small and uncontrolled, and the authors say plainly it is a starting point, not an answer. But it lands on a question that barely existed a few years ago: what happens when you give a weight-loss drug to people with a disease that used to be defined by the struggle to keep weight on.

A disease that changed

Cystic fibrosis is an inherited disease that thickens the mucus in the lungs and gut. For most of its history it was a wasting condition. Patients burned more calories than they could absorb, and staying heavy enough was itself a survival strategy, because low body weight tracked with worse lungs and a shorter life.

That math flipped with CFTR modulators, the drugs led by Vertex's elexacaftor-tezacaftor-ivacaftor (sold as Trikafta) that fix the underlying protein defect. People who once fought to gain weight now sometimes gain too much, and obesity and its metabolic problems, including cystic-fibrosis-related diabetes, are showing up in a group that never used to face them. GLP-1 drugs, the semaglutide ↗ and tirzepatide ↗ class that mimics a gut hormone at the GLP-1 receptor ↗ to curb appetite and lower blood sugar, are the obvious tool. Almost nothing is known about using them here.

What the chart showed

The UT Southwestern team pulled records for 23 adults with cystic-fibrosis-related diabetes who were put on a GLP-1 drug. Their FEV1, the standard lung-function measure (how much air you can force out in one second), rose a median of 6.0 percentage points, with most patients landing between 3.5 and 10 points, over six months.

Two things tracked with a bigger gain: losing more weight, and actually reaching a full therapeutic dose of the drug rather than stopping at a starter dose. The result the authors flag is that the dose-and-lung link survived statistical adjustment for weight change. If less weight were the whole story, accounting for it should have erased the connection. It did not, which hints the drugs might reach the lung through something other than the mechanical relief of carrying less mass, perhaps inflammation or an effect on the airway itself.

What it does not show

Everything about the size and shape of this study argues for caution. Twenty-three patients, no control group, one center, six months. FEV1 in cystic fibrosis is a restless number that moves with modulator adjustments, with infections cleared or caught, and with the season. With no comparison arm there is no clean way to separate the drug from the ongoing benefit of the CFTR modulators these patients were also taking, or from simple regression to the mean. "Associated with" is doing real work in every sentence of the results. Nobody has shown the drug caused the gain.

What the study earns is a trial. GLP-1 use in cystic fibrosis is new, off-label, and spreading ahead of the evidence, and a signal that it might help the lungs and not just the waistline is exactly the kind of thing a controlled study should confirm or kill before the practice hardens. One of the authors reports funding from Vertex, the company whose modulators created the metabolic problem these drugs are now being aimed at.

The pattern is familiar from the rest of GLP-1's expansion. When the drugs were tested in obstructive sleep apnea ↗, they moved the machine-measured breathing number convincingly while the symptom patients actually feel barely budged. A GLP-1 story that opens with a striking surrogate is a reason to look closer, not a reason to conclude. peptidemodel also hosts a card for dulaglutide ↗, another member of the class.