In the following article, the researcher who coined the phrase "complex sleep apnea" gives an overview of this variant, one that is often poorly understood and has been variously defined. Of special interest, the article includes:
– Si Baker-Goodwin, EdD, SAPCON Patient Engagement Panel member
The original description of complex sleep apnea described a pattern consistent with simultaneous upper-airway obstruction (partial or full collapse of the throat, with effort to breathe) and increased breathing control-related fluctuations ("central" events without associated effort to breathe). It is important to note that accurate distinction of respiratory breathing “events” as “central” vs. “obstructive” is needed to identify this disorder.
The disease was observed to occur mostly in non-rapid eye movement (NREM) sleep (which includes the deep, restorative phase), rather than the REM sleep (dream phase) dominance of typical obstructive sleep apnea. However, it was also seen emerging during CPAP titration or early phases of CPAP treatment.
Several professional entities further narrow its description via their diagnostic criteria. For example: the Medicare criteria for diagnosing complex sleep apnea require:
The International Classification of Sleep Disorders has an entity called “Treatment-Emergent Central Sleep Apnea” which follows the Medicare criteria.
These criteria and diagnostic descriptions capture only a small part of what I call complex apnea.
To understand the cause of the diagnostic confusion and disagreement requires a working knowledge of breathing control and an appreciation of some limitations of the current sleep apnea scoring rules.
To start with breathing control, there is an elaborate system starting with the carotid bodies- small (2-3 mm) structures in the middle of the neck right next to the carotid arteries- which are oxygen (O2) and carbon dioxide (CO2) blood sensors. This information goes to the lower part of the brain, called the brainstem, where there are several connected centers for breathing and blood pressure control. Information streams out of these centers to the heart and the muscles that are used for breathing, which in turn changes the O2 and CO2 in the blood, and the loop continues. Thus, there are natural small fluctuations in O2, CO2, blood pressure, and breathing that are all linked together.
When we fall asleep, breathing volume per breath is reduced, and there is a small increase in CO2 and decrease in O2 in the blood, gently stimulating the carotid bodies. This results in a new balanced condition of breathing during sleep. Abnormalities in this system can lead to either increases or decreases in levels of CO2. If the system is less sensitive, high CO2 and low oxygen levels develop; this is called “hypoventilation”. If the system is too sensitive, a slightly low CO2 develops that can be sensed by the carotid body system as over-breathing, which triggers a change in breathing pattern (small breaths or holding of breaths), thus allowing CO2 to rise until a point is reached when a new breath is triggered. The cycle of fluctuating breathing drive repeats, and breathing volume in turn fluctuates with increased and decreased airflow and breathing effort. When breathing effort completely stops, we have traditionally called this a central apnea.
However, as often the case, what seems like something straightforward, is not so.
It turns out that when the drive to breathe waxes and wanes, the stimulation to the throat muscles and the breathing muscles also goes up and down. When drive to the muscles declines, many types of abnormal patterns can result, including both obstructive and central apneas and hypopneas. The airway can even be closed during central apnea. Central hypopnea is very hard to accurately score by hand. Although one can use the presence of snoring or examination of the pattern of airflow as clues as to whether a hypopnea is central or obstructive, this can be very difficult. Therefore, it is often difficult to accurately score obstructive hypopneas, especially when both obstruction and central events coexist.
It is also a good time to state that there is much research showing how abnormal breathing control causes obstructive features. It is estimated that nearly a third of patients with apparent obstructive apnea have a very strong breathing drive- this overly-active breathing control control system may contribute to their susceptibility to sleep apnea.
The good news is that there are relatively simple ways to identify when the breathing control system is overactive and that complex sleep apnea may be present. For example, seeing that apneas get better in during dream sleep (REM sleep) and are more common in non-REM compared to REM sleep are clues pointing to complex sleep apnea.
In its most simple terms, complex sleep apnea is NREM-dominant obstructive sleep apnea. Once any therapy for obstruction is applied, including oral appliances or jaw surgery, the breathing control component becomes more apparent, and central apneas and hypopneas can emerge. In some individuals, the number of central apneas may reach the threshold for Medicare criteria complex apnea and American Academy of Sleep Medicine criteria for treatment-emergent central sleep apnea. As central hypopneas are frequently missed during routine scoring, only the “tip of the proverbial iceberg” of breathing control abnormalities” is detected by standard criteria.
I coined the term complex sleep apnea in 2004 and a year earlier had described a NREM-dominant obstructive sleep apnea syndrome. However, since early in my sleep medicine training days in the mid 1990's, I was puzzled by the subset of patients who appeared to have obstructive sleep apnea but were difficulty to treat properly in the sleep laboratory. In particular, they persistently stated an inability to use, or intolerance to, positive airway pressure, despite their honest efforts to use it.
In those days the type of breathing monitoring now available in positive pressure devices did not exist, so we were largely blind to what happened during patients' nightly use. I also remember noting that bilevel positive airway pressure (BILEVEL) was often resulted in worse responses than continuous positive airway pressure (CPAP, including auto CPAP). In fact, in patients with NREM dominant sleep apnea, auto CPAPs seemed to “chase” changing breathing patterns with pressures that went up and down during sleep, resulting in even poorer results than use of fixed CPAP.
One particularly fine morning, I realized that the sleep of these “complaining patients” showed strong similarities to sleep apnea at high altitude, which occurs as a result of physiological changes as the body works to breathe in a lower-oxygen environment. (CO2 levels in the body control breathing, especially during sleep.). This realization explained why these patients had only a partial response to CPAP, and why their sleep was more severely impaired in NREM sleep – when the brain’s response to changing O2 and CO2 levels influences breathing most strongly.
The next part of the story involves meeting a patient with complex apnea, with whom I collaborated in designing a device to use CO2 to stabilize breathing rhythms, and a method to use mask-venting modifications to trap part of exhaled CO2 when using PAP, called Enhanced Expiratory Rebreathing Space (EERS). EERS is used routinely (not FDA-approved and unlikely as it is a simple reconfiguration of common components; its use is “off label”) at the Beth Israel Deaconess Medical Center’s (BIDMC) and some other sleep centers to treat complex apnea patients, while the CO2 device remains investigational.
We found we could stabilize this abnormal breathing rhythm with this method. To date, we have treated over 1,000 patients successfully. The original patient has built and used a CO2-regulating device of his own for over 10 years now with continued benefit.
Along with EERS and CPAP, more recently I have used low doses of acetazolamide, which is commonly used to prevent and treat high-altitude sickness and sleep apnea.
Also, I have found that some patients benefit from use of sedative drugs to reduce awakenings from sleep which tend to especially worsen complex apnea. Brief awakenings from sleep result in large breaths which lower CO2 and thereby prompt central apneas. Sedative use tends to reduce these awakenings. Some newer sedatives have been shown to be relatively safe in patients with sleep apnea, but need to be used very cautiously, by highly-experienced sleep physicians who closely monitor patient responses. In particular, sedatives can make sleep apnea more severe in REM sleep. It seems that to recover from an apnea in REM sleep, arousal is needed, while in NREM sleep, arousals increase the risk of apnea by lowering CO2. It is crucial that patients struggling with CPAP do not use over-the-counter sedatives without discussion first with their physicians.
The reported prevalence varies widely depending upon the particular criteria used for lab test scoring and diagnosis. It has been reported that 5 to 15% of patients with sleep apnea have this subtype. At BIDMC, we estimated that at least 10% of all apnea patients have complex apnea. Research from several different centers has shown that about 25 to 30% of apnea patients have a very strong breathing rhythm that causes apnea. If those who try hard but fail to get benefit from CPAP, and those who in frustration abandon its use altogether are also considered, it is probably far more frequent.
Atrial fibrillation, family history of complex/central sleep apnea, heart failure, male sex, renal failure, and stroke are factors that are associated with complex sleep apnea. The majority of complex apnea patients probably have a genetic risk, just as the amount of central apnea seen in individuals going to high altitude varies substantially even with the same degree of exposure. The genes controlling sensing of O2 and CO2 could alter the control pathways at many sites as the breathing control system is very elaborate.
The sleep study report may indicate features of abnormal breathing rhythm. The terms used are often central apnea, mixed apnea, periodic breathing, and ataxic breathing. The patients may feel they are fighting the air pressure, or rip the mask off frequently, or feel no benefit, or even feel worse, when using PAP. Despite AHI numbers that suggest “normal” sleep on CPAP, the patient may still feel they slept poorly, and function poorly during the day.
A high AHI while wearing CPAP, as recorded by the PAP device, can suggest complex apnea. This is especially true if no large leak is present. While modern PAP devices now track use and breathing, they are not accurate enough to determine if breathing is normal, and different vendors’ measure respiration in different ways.
Also, while physicians can look directly at the breathing patterns using software and Internet-based systems, patients can now do so as well, using free open-source software called SleepyHead. Written by a programmer in Australia, it is downloadable at sleepyhead.jedimark.net, and it provides a wealth of information patients can use to track and improve their sleep.
However, to use this software properly requires some understanding of normal and abnormal breathing patterns, and how to use all the information available through this method. MyApnea.Org will soon have a “How to Use SleepyHead” resource. You can also ask your sleep medicine physician how to use it.
The only FDA-approved treatment is adaptive servo-ventilation (ASV). ResMed and Philips-Respironics have devices approved world-wide, and Weinmann has a device in Europe. These devices track breathing rhythm and create air pressure rhythms equal and opposite to the patient’s. They are at times very useful and effective, but at times the rhythm does not properly match and patients cannot tolerate the pressure fluctuations. The ResMed device was recently shown to worsen outcomes in patients with a certain type of severe heart failure. Interested readers can search the Internet with the keywords “SERVE-HF” and “ResMed” to find more information. One useful link is http://www.resmed.com/us/en/serve-hf.html
If you use one of these devices, it is important that you discuss the benefits and risks with your physician. These are powerful devices that need to be used with care. All other treatment options (EERS, acetazolamide, oxygen, and sedatives) are off-label (not yet FDA approved) and must be used only by physicians after a careful risk/benefit discussion. In my experience, oxygen is frequently “tried,” but rarely provides a high level of benefit. I have had some success combining an oral appliance with acetazolamide.
The most common problem is an inadequate response to, and difficulty tolerating, CPAP. There is thus a risk for complex apnea patients to remain untreated, with all the resultant consequences. If any patient with apparent obstructive apnea gets minimal benefit from substantial use of CPAP, complex apnea is a likely reason. If a patient feels worse with CPAP, one should strongly consider complex sleep apnea as a possible explanation.
I also have observed patients with complex sleep apnea to be at increased risk for developing new atrial fibrillation, which is supported by the literature showing a strong link between central sleep apnea and atrial fibrillation (see last reference article).
There currently is very little research directly addressing complex apnea that is currently ongoing, perhaps partly due to some controversy in the research community about its perceived importance. Ongoing research at BIDMC in collaboration with the Manton Center for Orphan Diseases at Children’s Hospital Boston is looking at genetic abnormalities in patients with complex apnea. The American Sleep Medicine Foundation is funding a project that will help identify factors for persistently abnormal respiration during long-term CPAP use (Grant Title: Polysomnographic markers of clinical sleep apnea outcomes). In this grant, my group will apply both conventional and novel analysis techniques to identify sleep study features that could predict persistent respiratory abnormality in CPAP users, including patients with complex sleep apnea.
Since complex sleep apnea seems to be less common that typical obstructive sleep apnea, collaborative research across centers may be needed to address these important issues. Perhaps one group to begin to focus on is the set of patients who try to use CPAP but do not get good responses to treatment.
Stanchina M, Robinson K, Corrao W, Donat W, Sands S, Malhotra A. Clinical Use of Loop Gain Measures to Determine Continuous Positive Airway Pressure Efficacy in Patients with Complex Sleep Apnea. A Pilot Study. Ann Am Thorac Soc 2015;12:1351-7.
Neu D, Balkissou AD, Mairesse O, Pefura-Yone EW, Noseda A. Complex sleep apnea at auto-titrating CPAP initiation: prevalence, significance and predictive factors. Clin Respir J. 2015 [Epub ahead of print] PMID: 26072986.
Bazurto Zapata MA, Martinez-Guzman W, Vargas-Ramirez L, Herrera K, Gonzalez-Garcia M. Prevalence of central sleep apnea during continuous positive airway pressure (CPAP) titration in subjects with obstructive sleep apnea syndrome at an altitude of 2640 m. Sleep Med. 2015;16:343-6.
Gilmartin G, McGeehan B, Vigneault K, Daly RW, Manento M, Weiss JW, Thomas RJ. Treatment of positive airway pressure treatment-associated respiratory instability with enhanced expiratory rebreathing space (EERS). J Clin Sleep Med 2010;6:529-38.
Thomas RJ, Daly RW, Weiss JW. Low-concentration carbon dioxide is an effective adjunct to positive airway pressure in the treatment of refractory mixed central and obstructive sleep-disordered breathing. Sleep 2005;28:69-77.
Gilmartin GS, Daly RW, Thomas RJ. Recognition and management of complex sleep-disordered breathing. Curr Opin Pulm Med 2005;11:485-93.
May AM, Blackwell T, Stone PH, Stone KL, Cawthon PM, Varosy PD, Redline S, Mehra R. Central sleep apnea predicts atrial fibrillation in older men. Am J Resp Crit Care Med. 2015: Nov 23
This article was written by Robert Thomas, M.D., Associate Professor of Neurology, Harvard Medical School and Beth Israel Deaconess Medical Center, in response to one of the top rated research questions on MyApnea.Org, according to our users.