How Sleep Apnea Endophenotyping Could Shape the Future of Sleep Medicine


Imagine if clinicians could easily and quickly pinpoint the exact pathophysiology of how sleep apnea emerges in each person’s body.

This level of understanding of the disease pathways in each patient would pave the way for sleep specialists to take a hyper-targeted approach when practicing sleep medicine.

Instead of automatically prescribing continuous positive airway pressure (CPAP), long considered the gold standard sleep apnea therapy, clinicians could more quickly turn to alternatives if a patient’s unique pathophysiology calls for another treatment strategy that might work better for them.

In the future, endophenotyping, the process of identifying specific traits to form subgroups of a disease, could make this advanced level of precision medicine a reality.

How Nox’s Endophenotyping Research Could Shape the Future

We at Nox Medical are at the forefront of endophenotyping research and have already developed an endophenotyping software tool called PUPpy that can identify specific traits and mechanisms that fall under the umbrella of obstructive sleep apnea.*

A recent research collaboration between researchers at the China Medical University Hospital, Harvard Medical School and Nox Research, used the PUPpy tool to analyze and describe three key pathological endotype clusters among patients with moderate to severe obstructive sleep apnea (OSA).

Led by China Medical University Hospital’s Wan-Ju Cheng, MD, PhD, and Liang-Wen Hang, MD, PhD, the research found that within these three clusters, each exhibited distinct polysomnographic characteristics and clinical symptom profiles that could one day be used in precision medicine for sleep apnea patients.(1)

The researchers reviewed PSG data, collected from a clinical cohort of 509 patients in Taiwan with moderate to severe OSA, to identify three endotype clusters. Each cluster showed clear polysomnographic characteristics and was linked with specific symptoms.

“Until now, there have not been many studies done on the endotype values in East Asian populations,” says Jón Skírnir Ágústsson, VP of artificial intelligence and data science at Nox Medical.

The researchers determined that linking endotype clusters with symptom profiles indicated that the pathological mechanisms of OSA may involve distinct symptom profiles, treatments, and comorbidities.

Sleep apnea is not a monolithic disease. One person’s airway could collapse during the night due to muscle tone loss, while another patient may stop breathing due to a neurological issue. Regardless of the cause, the gold standard treatment is CPAP therapy, but CPAP doesn’t work for everyone. A recent paper looking at one hundred eighty-seven patients with OSA who started PAP treatment between July 2018 to March 2020, published in the Journal of Clinical Sleep Medicine, found that PAP compliance at 3 months and 12 months was 79.1% and 51.3%, respectively.(2)

By forming a classification system composed of sleep apnea subgroups, we hope to help clinicians better identify patients who might be ideal candidates for alternative treatments. That way, clinicians can quickly turn to the most ideal treatment for each individual patient’s needs based on the patient’s endotype.

Endophenotypes for Precision Sleep Medicine

So far, scientists have identified three key endophenotype clusters for sleep apnea:

High Collapsibility combined with high loop gain and high arousal threshold (N=161) — Associated with obesity and disturbed sleep in PSG studies, Collapsibility is characterized by a tendency of the pharyngeal tissues to obstruct the airway during sleep, a primary cause of obstructive sleep apnea. Participants exhibited severe daytime sleepiness, snoring and moderate insomnia.

Low Arousal Threshold combined with low collapsibility and high compensation (N=163) — High hypopnea fraction in PSG studies. Participants experienced prominent insomnia symptoms, restlessness, moderate sleepiness and snoring.

Low Upper Airway Compensation (N=185) — Participants showed symptom profile between the first two clusters.

By using these sleep apnea subgroups, physicians and others could then evaluate each patient’s sleep disordered breathing with the patient’s endophenotype in mind to choose an ideal treatment modality. For instance, patients with less collapsibility and lower loop gain seem to respond favorably to oral appliance therapy. In other cases, those patients with lower loop gain, good compensation, and higher arousal threshold might be candidates for hypoglossal nerve stimulation.

While additional research is needed, endophenotyping offers a promising way to highlight subsets of patients who are more likely to benefit from different therapies based on the underlying mechanisms of their disease and Nox Medical diagnostic tests are in the right position to capture the data needed for such complex analysis.

*Endophenotyping software analysis by Nox Medical is currently being used for research purposes only.


Cheng WJ, Finnsson E, Arnardóttir E, Ágústsson JS, Sands SA, Hang LW. Relationship Between Symptom Profiles and Endotypes Among Patients with OSA: A Latent Class Analysis. Ann Am Thorac Soc. 2023 Jun 15. doi: 10.1513/AnnalsATS.202212-1054OC. Epub ahead of print. PMID: 37321164.

Park SI, Kim BK, Lee KE, Hong SD, Jung YG, Kim HY. Predictors for short-term and long-term automatic PAP compliance. J Clin Sleep Med. 2023 Jan 1;19(1):17-26. doi: 10.5664/jcsm.10236.

Topic: Industry News