Patients with type 2 diabetes taking glucagon-like peptide-1 receptor agonists face increased risk of developing chronic cough compared with those using alternative diabetes medications, according to a large cohort study published in JAMA Otolaryngology–Head & Neck Surgery.
The retrospective analysis drew from the TriNetX research network, examining electronic medical records from 70 US health care organizations between April 28, 2005, and April 15, 2025. The study, led by Tyler J. Gallagher, MD, of the Caruso Department of Otolaryngology–Head & Neck Surgery at Keck School of Medicine of the University of Southern California, and colleagues identified 427,555 adults (mean age, 56 years; 59% female ) with type 2 diabetes prescribed a glucagon-like peptide-1 receptor agonists (GLP-1 RA) and 1,614,495 patients (mean age, 64 years; 44% female) with type 2 diabetes prescribed other second-line diabetes medications.
After propensity score matching, patients prescribed a GLP-1 RA had significantly increased risk of new chronic cough compared with patients prescribed any non–GLP-1 RA second-line medication, dipeptidyl peptidase-4 inhibitor, or sulfonylurea. The risk difference was negligible compared with sodium-glucose cotransporter 2 (SGLT-2) inhibitors. When patients with any history of gastroesophageal reflux disease were excluded, the association strengthened across all comparison groups, including SGLT-2 inhibitors.
The study also examined whether medication formulation influenced outcomes. Long-acting GLP-1 RAs were associated with increased risk of chronic cough compared with any non–GLP-1 RA. Short-acting GLP-1 RAs showed no significant association with chronic cough, although power was limited.
Prior to matching, 2% of patients prescribed GLP-1 RAs developed chronic cough, followed by 1% of those receiving SGLT-2 inhibitors , 1% receiving any non–GLP-1 RA second-line agent, 1% receiving dipeptidyl peptidase-4 inhibitors, and 1% receiving sulfonylureas.
Cohorts were matched for current age, age at index event, sex, race, ethnicity, and multiple comorbidities including chronic obstructive pulmonary disease, asthma, diabetic gastroparesis, chronic bronchitis, gastroesophageal reflux disease, chronic rhinitis, postnasal drip, heart failure, tobacco use, overweight status, obesity status, metformin use, insulin use, and angiotensin-converting enzyme inhibitor use.
The findings suggest multiple potential mechanisms. "One possible reason for this is that chronic cough may be a presenting symptom of GERD; previous literature has shown that as many as 75% of individuals with GERD present without gastrointestinal symptoms yet manifest with other symptoms, such as chronic cough," the researchers wrote. Patients without gastrointestinal symptoms may not receive a diagnosis until cough prompts an evaluation.
"One direct etiology could be activation of the vagal nerve. Previous studies have shown that GLP-1RAs stimulate vagal afferent signaling, which results in various different afferent signals," noted researchers. Studies have described the presence of GLP-1 receptors in the respiratory epithelial cells and vasculature of the lungs.
The researchers also noted that otolaryngologists and general practitioners should be aware of the possibility of an association between GLP-1 RA use and chronic cough, particularly when other etiologies are not consistent with a patient's clinical picture.
"Management of chronic cough in these patients should continue to focus on the most likely etiologies of cough," they wrote. "When history and examination findings are suggestive of laryngopharyngeal reflux, a treatment approach targeted at addressing this mechanism as suggested by current guidelines should be used."
The study authors suggested that more aggressive management of reflux may be needed for symptom control given the pharmacological mechanism of GLP-1 RAs, which delay gastric emptying. For refractory cases, they noted that a patient-centered discussion regarding a discontinuation trial of the medication based on the patient's unique risk-benefit profile may be necessary.
The investigation acknowledged several limitations. Data on medication compliance, dosage, and duration of use were unavailable. The analysis relied on International Classification of Diseases, Tenth Revision codes for chronic cough diagnosis rather than validated cough severity measures. Body mass index data were unavailable, though researchers controlled for obesity through diagnostic codes.
The research received no external funding, and researchers reported no conflicts of interest.
