Sleep apnea subtype predicts who benefits most from tirzepatide

Clinical Takeaway: Sleep apnea features beyond severity itself may help predict which patients will see the strongest tirzepatide benefit. Endotyping could eventually shift counseling away from population-average expectations toward more individualized treatment predictions.

Tirzepatide reduces obstructive sleep apnea severity in patients with comorbid obesity, but average treatment effects mask substantial individual variation. Clinicians have had little to offer beyond population-level expectations when patients ask how much benefit they should expect. This analysis aimed to identify a responder phenotype using physiological features of sleep apnea itself.

Patients in the strong-response subgroup experienced roughly twice the reduction in sleep apnea severity compared with other treated patients. The defining traits were a combination of younger age, less severe obesity, more pronounced upper-airway collapsibility, greater breathing-control instability (high loop gain), and a lower threshold for arousal during airflow obstruction.

Loop gain as a predictor was initially counterintuitive, since unstable breathing control is not an obvious target of GLP-1 therapy. On further analysis, the researchers found tirzepatide appeared to improve both breathing instability and airway collapsibility, suggesting the drug may act on sleep apnea through pathways beyond weight loss alone.

The analysis used data from SURMOUNT-OSA, the earlier randomized trial of tirzepatide in patients with obstructive sleep apnea and obesity. Researchers applied sleep apnea endotyping methods to identify physiological subgroups, then compared treatment response across those subgroups. As a secondary analysis of a single trial presented as a conference abstract, the findings are exploratory and have not yet been peer reviewed. Prospective validation in independent cohorts would be needed before endotype-based selection could enter clinical practice.

If validated, the approach could shift sleep medicine toward a more personalized model where patients are told not just that tirzepatide cuts sleep apnea severity roughly in half on average, but where they likely fall on the response curve. It could also help identify patients likely to need adjunctive therapy from the start rather than after a disappointing trial of GLP-1 monotherapy.

"We are excited about these findings because they offer the chance to now be able to share with patients: 'You have the characteristics that place you in a group of patients who respond remarkably to tirzepatide weight loss therapy.' Or, by contrast, 'You have characteristics that suggest additional treatments might still be needed,'" said first author Scott Sands, PhD, associate professor of medicine at Brigham and Women's Hospital, Harvard Medical School.

Source: Sands S. ATS 2026. May 19, 2026. Tirzepatide outcome prediction using sleep apnea endotyping in SURMOUNT-OSA