An Early Indicator of Complex Sleep Apnea

Patients with undiagnosed or improperly treated complex sleep apnea frequently experience central events while receiving continuous positive airway pressure or bilevel therapy. They often become frustrated with persistent symptoms like fatigue, headaches, and irritability, which may lead them to view therapy as ineffective and stop it completely.

The standard treatment approach is to titrate CPAP to maximum tolerable pressures during the in-lab polysomnography. However, this often fails. Clinicians then usually switch the patient to a bilevel ventilation device, with limited or no success. Often, the patient spends a second night in the sleep lab undergoing a trial of adaptive servo-ventilation (ASV).

Used as an adjunct to standard scoring criteria, Cardiopulmonary Coupling (CPC) analysis is an automated operator-independent method. It identifies the presence of sleep-disordered breathing and evaluates treatment efficacy.

A sleep spectrogram of the initial split-night evaluation (top) shows the lack of HFC, consistent with the lack of efficacy of continuous and bilevel pressure treatments. Using ASV on the second night (bottom), the AHI decreased to 16.6 events/hour from a baseline supine AHI of 161 events/hour. HFC increased slightly (circle bottom) relative to the first polysomnogram. The bottom panel shows its persistence on ASV. The coupling frequency is faster (20 mHz) at approximately 5 a.m., which is reminiscent of the short respiratory even cycles of altitude-induced periodic breathing.

Embla Systems Inc. developed this technology into a marketable product in collaboration with a team of researchers at Boston’s Beth Israel Deaconess Medical Center and Harvard Medical Center sleep laboratories, headed by Robert Thomas, MD.

The electrocardiogram-derived sleep spectrogram estimates the coupling of heart rate variability and R-wave amplitude variations induced by respiration, yielding the CPC sleep spectrogram.1

High-frequency coupling (HFC) band – 0.1 to 0.4 Hz – is associated with physiological respiratory sinus arrhythmia and stable sleep and respiration, while low-frequency coupling (LFC) – 0.1 to 0.01 Hz – is associated with unstable sleep or periodic sleep behaviors, including sleep apnea.2

The resultant spectrograph provides a “picture” of the dominant oscillatory frequency patterns of autonomic drive coupled with respiration.

Early recognition of complex sleep apnea using CPC to phenotype patients with SDB could improve the level of care and significantly reduce laboratory and health care costs.

Case study

A 47-year old male presented with complaints of loud snoring, apneic episodes, waking up gasping, and hypersomnolence. At 5-foot, 7-inches and 240 pounds, he had a body mass index of 37.6.

A split-night study showed a pre-CPAP supine apnea hypopnea index (AHI) of 161.0 and a non-supine AHI of 99.4 per hour of sleep during the initial 155.5 minutes of total recording time.

Using a medium nasal mask (Mirage Swift, ResMed Corp.), clinicians first evaluated the efficacy of CPAP 4 to 15 cm H2O then bilevel ventilation 14/10 to 19/15 cm H2O in eliminating respiratory events. The AHI remained high, with the best AHI of 21.5 per hour achieved on bilevel at 12/10 cm H2O.

CPC analysis showed the HFC decreased at 6.6 percent compared to an expected adult value of > 50 percent. The LFC increased to 79.8 percent compared to an expected adult value of < 30 percent, and the elevated LFC was 2.5 percent compared to an expected value of 0 percent.

On the second night, the patient underwent a trial of ASV using a full face mask (Mirage Quattro, ResMed Corp.). Sleep professionals initiated ventilation therapy at an EEP of 5 cm H2O and titrated to 10 cm H2O. The AHI decreased to 16.6 events per hour of sleep at 9 cm H2O. The HFC increased to 12.9 percent, but the LFC remained high at 82 percent. The elevated LFC narrow band coupling was persistent at 5.5 percent.

Providers prescribed the patient an ASV unit with EEP at 9 cm H2O. Overall, he has found his sleep more restful. By using CPC to detect SDB, providers could deliver more appropriate medical care on the first night to patients with a phenotype like this case study subject. In addition, patient compliance would likely improve with proper identification and adequate treatment that addresses both underlying causes of the complex sleep apnea.


1. Schramm PJ, Madison S. An electrocardiogram-based technique to identify complex sleep apnea in patients with diabetes and sleep disordered breathing. (In preparation)

2. Thomas RJ, Mietus JE, Peng CK, et al. Differentiating obstructive from central and complex sleep apnea using an automated electrocardiogram-based method. Sleep. 2007;30(12):1756-69.

Preetam Schramm, PhD, RPSGT, is clinical specialist at Embla Systems Inc. in Broomfield, Colo.

Case Study Sponsored By: