Dr Barry Krakow’s findings on the relationship between insomnia and sleep-disordered breathing have far-reaching implications for diagnosis and treatment. The Medical Director of the Maimonides Sleep Arts & Sciences Ltd in Albuquerque, NM, speaks with Sleep Review about why the comorbidity is frequently overlooked, what treatments work best, and how to file insurance reimbursements.

Interview by Sree Roy

Photos by Steven St. John

Sleep Review: A lot of research looks at coaching insomniacs on how to go back to sleep after they wake up, but you looked at why they woke up in the first place. What did you find?

Barry Krakow, MD: We asked insomnia patients to tell us why they wake up1 and found that the causes they attribute to their awakenings are very different from what we see in the sleep lab. Fifty percent say it’s mental and 50% say it’s physical. They’ll also point to things after the fact, such as “I woke up because my mind is racing” and then realize that the racing thoughts really emerged after the awakening. In the lab, however, breathing events were the most common cause of their awakenings.

SR: In your 2012 article in Sleep, you purposely selected insomnia patients without signs or symptoms of sleep-disordered breathing (SDB), yet they still suffered many SDB events.1 Did these results surprise you?

Krakow: We first realized in 1994 that many patients were presenting to our sleep center with a possible link between insomnia and SDB. In this newer research, we expected some patients to show SDB events, but we were surprised that 17 of the 20 patients were diagnosed with either OSA or UARS.

SR: What percent of patients suffering from insomnia do you think suffer comorbid SDB?

Krakow: In my opinion, the number is astronomical, especially among treatment-seeking chronic insomnia patients who complain their sleep aids don’t work. For them, the number is over 90%. Among insomniacs with middle of the night awakenings, the number approaches 100%. For sleep-onset insomnia, the proportion with SDB is difficult to estimate, but I doubt the number is low.

In terms of primary care, there is scant data. We conducted one study among patients sitting in waiting rooms of primary care clinics and surveyed their sleep breathing symptoms and insomnia complaints,2 but we excluded anyone who reported past sleep disorder diagnoses or treatments. At least one-third still seemed likely to suffer comorbid insomnia and SDB or what we like to call “complex insomnia.”

SR: Is there any chance the insomnia came first and the SDB came second? Could sleep fragmentation destabilize the airway?

krakow CoverSwiggles SIZEDRespiratory event related arousals (RERAs) appear as plateaus on a PSG. In Krakow’s lab, insomnia patients with comorbid sleep-disordered breathing are titrated to round their inspiration and expiration curves, resulting in significant insomnia improvement.

 Krakow: That’s the working theory of several researchers. We first came upon it because of our work with post-traumatic stress disorder (PTSD) patients 20 years ago. We noticed patients seeking treatment for nightmares and insomnia also suffered high rates of SDB. Many of these individuals described their sleep as perfectly normal prior to their traumatic exposure.

Later, we discovered Series et al’s3 work, which showed that noise-induced fragmentation increased upper airway collapsibility in sleepers who previously had been breathing normally. Since then, we have wondered whether insomnia, PTSD, depression, or other mood and anxiety disorders cause initial sleep fragmentation that over a period of months or years leads to SDB but presents on the surface as insomnia.

SR: What are the clinical ramifications of the link between insomnia and sleep-disordered breathing?

Krakow: The biggest problems are delay in care and inappropriate care. Some examples are highlighted by work we published on several hundred hypnotic-dependent insomniacs.4,5 They had presented having used sedating agents, on average, for at least 4 years and suffered from insomnia for more than a decade. Upon further evaluation, we discovered that between 80% and 90% had SDB. This delay in their SDB diagnosis certainly affects their health and quality of life, including the consequences of untreated SDB, such as high blood pressure, low energy, and mood symptoms, and in our view acts as an aggravating influence on their insomnia, which then seems to trigger more prescriptions for sleep aids.

SR: What are the specific SDB diagnoses common to patients with insomnia?

Krakow: A mix of OSA and UARS, either of which causes sleep fragmentation. In the study we conducted showing breathing events had caused awakenings,1 the dominant trigger wasn’t apneas or hypopneas; it was respiratory effort-related arousals (RERAs), the scoring term for UARS. Yet, unfortunately, some sleep professionals, insurance carriers, and government agencies like Medicare pay scant attention to UARS, which sends a message that this type of breathing event does not need treatment.

SR: RERAs still seem controversial in the field of sleep medicine. Can you comment?

Krakow: A paradox has arisen because the American Academy of Sleep Medicine (AASM) made it optional to score RERAs on diagnostic studies, yet it mandates treatment of RERAs on titration studies. This is confusing to other sleep facilities, so we wrote a commentary that was accepted for publication in the Journal of Clinical Sleep Medicine on the rationale for scoring RERAs on all sleep studies.6 Clinically, we see our best results when we aggressively treat RERAs in insomnia patients by using advanced PAP therapy devices; whereas, if we don’t treat the RERAs, we find most of these patients can neither use PAP therapy consistently nor achieve the greatest improvements in their insomnia.

I have never understood the rationale for optional scoring of RERAs. An analogous situation occurs with cardiac arrhythmias; no one thinks the only arrhythmia is asystole (no heart beat). It’s widely acknowledged various arrhythmias must be treated. In sleep, there are various airflow patterns that cause problems. UARS events have been described in the scientific literature for more than two decades and linked to daytime sleepiness, trips to the bathroom at night, high blood pressure, and possibly depression. In my opinion, failing to treat UARS borders on medical malpractice.

SR: Some sleep doctors argue that a fairly low apnea/hypopnea index (AHI) unaccompanied by symptoms, such as daytime sleepiness, does not need treatment. Would that hold for insomnia patients with mild SDB?

krakow Patient SIZEDSleep techs play the critical role in treating complex insomnia patients in the lab, because they must learn to manually titrate RERAs in tandem with the auto-adjusting feature of the PAP device.Krakow: I hope that physicians with that perspective are in the minority, because the majority of OSA patients have so many other symptoms beyond classic daytime sleepiness. For example, as much as 50% of OSA patients have insomnia, and many of these patients also report fatigue, nonrestorative sleep, difficulty in concentrating, or waking up to use the bathroom. Moreover, such a patient may have a low AHI but a high respiratory disturbance index (RDI) when the RERA index is properly accounted for. In other words, many so-called mild cases are really more severe cases, but failing to measure the full RDI leads to false impressions.

In addition, it is important to view this problem in a larger framework. The current measurement techniques for SDB are extremely primitive. For example, we often cannot determine what a low AHI or low RDI means until the patient has been treated. I’ve treated patients with so-called low AHI or RDI or both who report severe insomnia, sleepiness, or fatigue. They reported marked improvements in their symptoms, and some patients were so impressed with the results from their PAP devices, they paid out of pocket when insurance failed to cover them. So, we make the assumption that many treatment-seeking patients suffer significant sleep fragmentation that does not always correlate with the number of breathing events, yet still responds to PAP.

SR: How do you see complex insomnia influencing policies on the appropriate use of sleep testing, whether via an in-lab polysomnogram (PSG) or an at-home sleep test?

Krakow: Home sleep tests are ineffective tools for insomnia sufferers at this point because most do not measure the full RDI. A few offer an option to score RERAs manually, and the WatchPAT devices appear to score a full or true RDI automatically. Regardless, an insomnia patient’s shortened sleep duration may yield an overestimate of total sleep time on a home test, which would underestimate the breathing event index. Insomnia patients more naturally see their problem as psychological, therefore they need a more hands-on experience, so we bring them into the sleep lab to engage them more actively by reviewing the breathing events on the computer and explaining how the breathing disorder may be an integral component of their overall sleep health and specifically their insomnia.

I have been very pleased in discussing this comorbidity problem with medical directors of several insurance companies who agree that home testing is neither appropriate nor valid for insomniacs with suspected SDB. When pieces to this puzzle are properly explained, we experience rare difficulties in obtaining prior authorizations for the in-lab PSG.

SR: So you regularly test insomniacs in the sleep lab?

Krakow: Yes, we recommend in-lab PSGs on 100% of treatment-seeking chronic insomnia patients. We receive few rejections from insurance carriers. The key is not to ask for a sleep study to evaluate their insomnia. We are asking to evaluate the physiological problem that is comorbid with the insomnia. Ironically, most insomniacs who come to a sleep center actually report some breathing symptoms. Indeed, 80% of insomniacs have awareness of sleep breathing issues,7 but almost none connect it to their insomnia. Many will even say, “I’ve got sleep breathing problems too, but I don’t care about that. I want something for my insomnia.” Because the patient is focused on the insomnia, he or she steers the doctor away from investigating the need for a PSG.

SR: What about doctors who are still getting insurance company rejections for PSGs?

Krakow: There will always be some rejections; I just got one today. But I think if you read the AASM guidelines on sleep testing for insomnia patients, you see there is tremendous flexibility built into those practice parameters.8 If a patient’s insomnia is not resolving or sleep aids are not working, these factors justify the need for objective evaluation. My sense is that sleep centers that receive denials on prior authorizations may be describing their patients in such a way the insomnia complaint sounds more prominent than the potential for comorbid SDB.

Although the AASM guidelines are very helpful, in my opinion, the AASM should consider a stronger stance on this issue, because the comorbidity of insomnia and SDB is one of the most common presentations to a sleep center. I would like to see the AASM consider revising or updating its policies to be more assertive and push back against insurance companies by explaining why home sleep testing for insomniacs with suspected SDB is less effective and why an in-lab study is needed. If the AASM takes a stronger stance, it would help individual doctors deal more expediently with insurance carriers.

krakow Team SIZEDKrakow and his team’s research on complex insomnia—that is, insomnia with comorbid sleep-disordered breathing—spurs more effective treatments for insomnia patients.

SR: With regard to treatment, you’ve had success using ASV (adapt-servo ventilation) versus standard PAP devices with insomnia patients. Why do you prescribe more advanced devices?

Krakow: Yes, I’ve had great success, personally, using ASV for nearly 4 years. It eliminates 95% of awakenings at night, which are noticeably fewer disruptions than what I experienced with other PAP devices. In general, insomnia patients do better with advanced devices. The driving influence is the anxiety factor so common to insomnia patients. During expiration, the discomfort from standard pressurized airflow seems to trigger an anxiety response that the insomnia patient cannot overcome; whereas with the use of auto-bilevel devices, including ASV, the patient’s comfort level is noticeably higher.

In these cases, the subjective and objective finding is expiratory pressure intolerance (EPI). Conventional wisdom alleges EPI dissipates with time. We dispute this point. We have seen nearly a thousand, second-opinion patients who were using CPAP or APAP for several years. But, when we looked at the expiratory limb of their airflow signal during the initial titration protocol at our sleep center, we found they were still fighting with the machine. You can also spot this problem during the desensitization period before the sleep study formally begins.

SR: How do you get insurance to pay for ASV?

Krakow: You can’t start with ASV for the majority of patients. In our setting, the patient can objectively and subjectively fail CPAP or BPAP at either the desensitization period or the early phases of the titration. Also, we can quickly spot that the device is not eliminating RERAs; instead the device is producing expiratory pressure intolerance. Other objective markers of an inadequate titration that show up include: failure to generate REM sleep, excess sleep stage transitions, and other signs of sleep fragmentation. After these adverse findings are observed, the sleep tech switches the patient to the advanced technology for the rest of the night. The sleep tech then must constantly override the auto mode, because only through a manual titration combined with the device set for auto mode do we gain the best chance to treat RERAs. With this method, there’s a potential for normalizing the air flow curve, rounding both the inspiratory and expiratory limbs.

About half to three-quarters of patients do well on standard auto-bilevel devices initially, but some patients have residual central apneas. When outcomes are not improving, patients typically are covered by insurance to return for another titration, which we think of as a “treatment procedure.” During this procedure, we see the residual central apneas, after which the patient can be titrated on ASV, assuming they qualify by insurance standards for this switch. Moreover, it is essential to recognize that when a patient uses one device and experiences no benefit, it is within the physician’s prerogative to talk to the insurance company about trying another device for that individual patient.

SR: Many patients are diagnosed with SDB, but never report insomnia. What separates insomnia sufferers from the rest of the population, and what prevents them from falling back asleep?

Krakow: This is a very important question because most likely we’re dealing with an anxiety response again. Anxiety patients may have an altered, if not exaggerated, fear response. When insomnia patients wake up at night, they feel wide awake and their minds are racing. This response contrasts with classic sleep apnea patients, who may wake up with no anxiety and no fear response, so they roll over and return to sleep. Even though both types of patients may be experiencing the “same” physiologic, sympathetic activation that follows a breathing event, the psychology of the insomnia patient yields a more intense response, which in turn feeds the sleeplessness.

krakow01 copySIZEDKrakow and his team have worked diligently to understand why the comorbidity is frequently overlooked, what treatments work best, and how to file insurance reimbursements.

SR: Our discussion all ties in with your new relationship with Classic Sleepcare. Please comment.

Krakow: I’m very excited about my work with Classic Sleepcare because the company has a strong desire to address the comorbidity of insomnia and SDB and is coming at the issue from the unique position of a durable medical equipment (DME) company.

Classic Sleepcare is exploring a number of different approaches to tackle this problem, and, most importantly, they recognize that efforts to manage the insomnia component in their sleep apnea patients will increase PAP therapy compliance while at the same time improve their patients’ insomnia. Classic Sleepcare is highly motivated to integrate appropriate treatment steps within this domain of care, because they have become very aware of how much insomnia is interfering with their patients’ ability to use PAP devices. I was very impressed by their decision to move in this direction, and I know of no other DME companies that are doing so.

SR: How did your relationship with Classic Sleepcare start?

Krakow: We started about 1 year ago, because we discovered we shared a mutual perspective on the need for very intensive and extensive follow-up in managing sleep apnea patients.

SR: Finally, does your research have clinical ramifications on whether we should be thinking of insomnia as a symptom or a disorder?

Krakow: It is both a symptom and a disorder. However, for the purposes of advancing our efforts to help patients with complex insomnia and to create a paradigm that draws more attention to this comorbidity, it is advisable to label insomnia as a disorder. The current “symptom” paradigm may create confusion if a patient’s complaints of sleeplessness steer the doctor away from any discussion of SDB. In my opinion, when you think of insomnia as a disorder, you are more likely to view the condition more seriously and therefore complete a more thorough evaluation, hopefully one that includes polysomnography. 

Sree Roy is editor of Sleep Review. She can be reached at [email protected].

Read about Krakow’s work with PTSD sufferers in “Vision in the Desert. You can also ask Krakow questions about sleep therapy via SR’s Expert Insight page.







  1. Krakow B, Romero E, Ulibarri V. Prospective assessment of nocturnal awakenings in a case series of treatment-seeking chronic insomnia patients: a pilot study of subjective and objective causes. Sleep. 2012;35(12):1685-1692.
  2. Krakow B, Ulibarri V, Romero E, McIver N. A two-year prospective study on the frequency and co-occurrence of insomnia and sleep-disordered breathing symptoms in a primary care population. Sleep Med. 2013;14:814-823.
  3. Sériès F, Roy N, Marc I. Effects of sleep deprivation and sleep fragmentation on upper airway collapsibility in normal subjects. Am J Respir Crit Care Med. 1994;150(2):481-5.4.
  4. Krakow B, Ulibarri V, Romero E. Persistent insomnia in chronic hypnotic users presenting to a sleep medical center. J Nerv Ment Dis. 2010;198(10):1-8.
  5. Krakow B, Ulibarri V, Romero E. Patients with treatment-resistant insomnia taking nightly prescription medications for sleep: a retrospective assessment of diagnostic and treatment variables. Prim Care Companion J Clin Psychiatry. 2010;12(4):PCC.09m00873.
  6. Krakow B, Krakow J, Ulibarri V, McIver N. Commentary on frequency and accuracy of “RERA” and “RDI” terms in the Journal of Clinical Sleep Medicine from 2006 to 2012. Journal of Clinical Sleep Medicine. In press.
  7. Krakow B, Ulibarri VA. Prevalence of sleep breathing complaints reported by treatment-seeking insomnia patients on presentation to a sleep medical center: a preliminary report. Sleep Breath. 2013;17(1):317-22.
  8. Littner M, Hirshkowitz M, Kramer M, et al; American Academy of Sleep Medicine, Standards of Practice Committee. Practice parameters for using polysomnography to evaluate insomnia: an update. Sleep. 2003;26(6):754-60.