For decades, the apnea-hypopnea index (AHI) has provided clinicians with classifications to describe obstructive sleep apnea (OSA) severity based on the number of breathing pause events per hour of sleep. But a single number can only tell part of the story. AHI tells us how often respiratory events occur. It does not necessarily explain why they occur, why two patients with similar AHI values can experience OSA very differently, or why they may respond differently to the same treatment. That is where sleep apnea endotyping is beginning to change the conversation.
As Professor Danny J. Eckert of Flinders University’s College of Medicine and Public Health notes, understanding a patient’s underlying endotypes can help explain why that individual develops OSA — and may ultimately help match the right patient to the right treatment.
At Sleep Europe 2026, Nox Medical will bring together experts for the symposium “From Signals to Clinical Strategy: Democratizing Sleep Apnea Endotyping for Precision Care.” The session will explore how advances in airflow measurement, signal analytics, and AI may make physiological insights that were once largely confined to research settings more accessible in clinical care.
Looking beneath the AHI
OSA is a heterogeneous disorder. Different physiological mechanisms can contribute to airway obstruction and instability during sleep, including differences in upper-airway collapsibility, ventilatory control, arousal threshold, and muscle responsiveness.
Understanding those mechanisms may provide information that conventional severity measures cannot.
“AHI is useful for describing the frequency of respiratory events, but it doesn’t tell us why a particular patient has obstructive sleep apnea,” says Heidi Riney, MD, Chief Medical Officer at Nox Health. “Endotyping gives us a window into the underlying physiology.”
That distinction matters clinically. Two patients of similar age and BMI may have almost identical AHIs yet report very different symptoms, experience different physiological disturbances during sleep, and respond differently to the same therapy. Eckert extends that point to treatment selection: endotype information may identify patients — including some with severe OSA — whose physiology is associated with response to non-CPAP therapies, a possibility that may be obscured by the headline severity number alone.
Rather than simply asking how severe is this patient’s OSA?, endotyping introduces another question: what is driving it?
From trial and error toward more informed treatment selection
The timing is increasingly relevant as treatment options for OSA expand. Positive airway pressure remains a cornerstone of treatment, but clinicians now have a growing range of alternatives. Selecting among them is still frequently based on broad clinical characteristics followed by a degree of therapeutic trial and error.
According to Jón Skírnir Ágústsson, PhD, VP of Artificial Intelligence and Data Research at Nox Medical, understanding a patient’s endotype could eventually help clinicians identify patients who may struggle with a particular therapy and consider strategies to improve treatment success earlier. One example is arousal threshold. Ágústsson notes that research has found that patients with a low arousal threshold may tolerate PAP differently and may experience different treatment effects on sleep parameters than patients with a higher threshold. Insights like these could add important context to conventional metrics.
Eckert sees potential across existing and emerging therapies. He points to oral appliance therapy as one of the clearest examples today, noting evidence that endotype-informed patient selection could improve treatment success from roughly 50% to around 80%. If validated and translated into routine pathways, this type of stratification could help shorten the treatment journey and get appropriate patients to an effective therapy sooner.
Riney sees a similar opportunity: “If endotyping can reliably identify the physiology driving an individual patient’s disease, it could help us better match patients to therapies they are more likely to respond to and tolerate, and potentially get them to effective treatment sooner.”
The goal is not to replace established measures such as AHI. It is to add physiological context that could support a more individualized treatment strategy.
Making complex physiology clinically accessible
Endotyping itself is not new. What is changing is the possibility of deriving increasingly sophisticated physiological information from signals already collected during sleep testing.
The Sleep Europe symposium will examine how advances in airflow measurement, signal analytics, and AI may help turn routine sleep signals into clinically meaningful physiological traits — without necessarily adding substantial testing burden for the patient or care team.
That accessibility may be critical to moving endotyping beyond specialized research laboratories.
“For endotyping to become part of routine care, it has to translate complex physiology into information that is simple, reliable, and clinically actionable,” Riney says. “Ideally, those insights should be derived from the sleep testing clinicians are already doing.”
Eckert argues that this translation now needs to become tangible in everyday workflows. Echoing the recent American Thoracic Society statement on translating OSA endotyping to the clinic, he highlights a practical next step: making endotype information readily available in sleep study reports, supported by educational resources that help clinicians interpret and apply it.
Important questions remain. Prospective evidence is still needed to demonstrate that using endotype information to direct therapy meaningfully improves patient outcomes. Clinical guidelines, reimbursement structures, and practical treatment pathways will also need to evolve.
As sleep medicine moves toward more personalized care, understanding the mechanisms beneath respiratory events may help clinicians see patients not simply as an AHI category, but as individuals with distinct physiology, treatment needs, and opportunities for better outcomes.
From Signals to Clinical Strategy: Democratizing Sleep Apnea Endotyping for Precision Care takes place at Sleep Europe 2026 on Thursday, October 22, from 12:45–13:45 in Session Room 0.11, featuring Heidi Riney, Jón Skírnir Ágústsson, and Danny Eckert.
Have symposium questions for our experts? Submit your questions here.
Topic: Events
