Sleep Onset Insomnia and Advanced PAP Therapy

In prior posts, we discussed how treatment of sleep onset insomnia with PAP therapy showed mixed results and delved into our research on the potential for greater improvements using advanced PAP devices (ABPAP and ASV) in these patients. With the very recent publication of our research paper, “A Novel Therapy for Chronic Sleep-Onset Insomnia: A Retrospective, Nonrandomized Controlled Study of Auto-Adjusting, Dual-Level, Positive Airway Pressure Technology,” we can now dig much deeper into our study methods and the precise details of the results to understand the clinical relevance of this hypothesis-generating study. This term for the study refers to its design not being at the highest level of evidence, which would have been a randomized controlled trial comparing ASV or ABPAP to a CPAP device as one example. Our nonrandomized design means the claims we offer are not proof ASV or ABPAP effectively treats sleep onset or early insomnia; instead, our findings indicate a need for more research on these ideas, because a hypothesis or theory was developed in our work to explain why advanced PAP devices might treat this problem when other traditional modes of PAP (e.g. CPAP) demonstrated mixed results.

In developing our research protocol, we started with the premise to isolate a group of patients with severe sleep onset insomnia (SOI) as opposed to a group with a mixture of equal degrees of early, middle and late insomnia. Thus, the chief insomnia complaint of eligible patients needed to be severe SOI, rated in two different ways. First, they needed to self-report sleep onset latency at bedtime greater than 60 minutes, and second on the Insomnia Severity Index rating system for question number one, they had to rate their difficulty in falling asleep as severe or very severe. By focusing only on these types of patients, we believed that whatever results we obtained would have more credibility because there would be no confusion about the type of insomnia we were attempting to treat with advanced PAP therapy.

Moreover, to solidify the perspective that these sleep onset insomnia patient were truly suffering from sleep onset problems, we asked them a series of questions that directly related to the most common symptoms seen in these types of patients. Here is a sample of a few of the behaviors indicative of sleep onset insomnia, and the percentages of patients who reported this issue:

  • - Racing thoughts keep me from sleeping (82%)
  • - Mind won’t turn off enough to fall asleep (72%)
  • - Clockwatching causes me frustration (67%)
  • - I clockwatch when trying to fall asleep (48%)

Of a possible 18 different indicators of insomnia in general and sleep onset insomnia in particular, each patient endorsed on average 9 of these maladaptive behaviors, and each patient averaged two or more psychiatric symptoms, conditions or disorders. All in all, our chart review collected 74 individuals with severe to very sleep onset insomnia whose profiles clearly established patterns consistent with the problem of early insomnia or SOI.

Other indications of severe global insomnia as well as severe sleep onset insomnia were demonstrated by their self-reported sleep patterns. They averaged less than 5 hours of total sleep per night despite remaining in bed on average for more than 8 hours, and they required more than 2 hours on average to fall asleep. They suffered from chronic insomnia for an average of more than 8 years, and 68% were using prescription sleeping pills, and 51% were using over-the-counter sleep aids. The sample included a mix of adult men and women, mostly White or Hispanic, average age 49 years-old, and mildly obese with an average BMI of 31 (roughly 25 to 30 pounds overweight).

Next, while all the patients were diagnosed with co-occurring OSA or UARS, we wanted to clarify why they were unable to tolerate standard CPAP therapy in order to provide the rationale demonstrating their need to attempt an advanced PAP therapy device. CPAP failure occurred in three phases of their treatment course:

Prescribed CPAP: 24 of the 74 patients initially used a CPAP device at home and failed primarily due to:


  • - Noncompliance (insufficient use)
  • - Subjective reports of difficulty breathing out (expiratory pressure intolerance,EPI)
  • - Poor outcomes defined as persisting insomnia, sleepiness or fatigue symptoms


Presleep/Desensitization: 48 of the 74 patients attempting CPAP in the sleep lab and prior to starting the actual titration failed CPAP during the desensitization trial primarily due to:


  • - Subjective reports of EPI
  • - Note: Actual instances of EPI equaled 65 due to patients failing on multiple devices (e.g. APAP, BPAP) before attempting advanced PAP


Titration: 69 of 74 patients titrated with CPAP, APAP, or BPAP in the sleep lab failed primarily due to:


  • - Needing variable pressure settings (auto-adjusting technology) to stabilize airflow signal
  • - Residual breathing events including central apneas
  • - Diagnosis of complex sleep apnea (residual central apneas meeting sufficient criteria)
  • - Note: Objective EPI equaled 83 instances, including failure on more than one PAP mode.


Within the 74 patients, the most striking difference was that some patients clearly were using their advanced PAP device more than others, which led to a dichotomy involving 56 patients who were using PAP on average more than 42 hours per week or greater than 6 hours per night (designated PAP Users) compared a Partial User group averaging less than 12 hours per week or less than 2 hours per night. Using this dichotomy, we compared how the severity of global insomnia and sleep onset insomnia changed in the two groups.

In the PAP User group, the Insomnia Severity Index total scores (ISI-TOT) dropped from 21.8 (consistent with severe global insomnia to 13.0 (consistent with less than moderate levels of insomnia severity). This change reflects what is known as a very large effect size, a statistical term that demonstrates the clinical relevance of a treatment response. In other words, a large effect size, or in this case a very large effect size usually coincides with a very large clinical change, which then is very noticeable to the patient. Small effects are usually less noticeable to patients while medium sized effects are noticeable but not as pronounced.

Remarkably, even the Partial User group demonstrated a large effect size with ISI-TOT scores dropping from 21.6 to 16.3. The changes between the two groups demonstrated statistical significance, which in this study is a way of declaring that the superior results in the PAP Users compared to the Partial Users was a very reliable finding, likely to be replicated if the research were repeated. However, to reiterate, the research is only a chart review and not a randomized controlled trial, the latter the highest form of evidence when attempting to prove the validity of a particular treatment. 

The findings on the ISI-SOI single item score showed equally impressive results with the PAP User group demonstrating decreased ratings from midway between severe and very severe sleep onset difficulties to just under the moderate level of intensity. This change yielded a very large effect size double that observed in the Partial Users; yet even the latter group dropped their ratings from slightly more than severe to midway between moderate and severe. Again, both groups showed impressive improvements in sleep onset insomnia, but the PAP Users results were superior and statistically significant to the Partial Users.

An obvious question, one we could not answer in this study, was whether or not the two devices, ABPAP or ASV, were any different in their effects on SOI. The results were the same whether a patient used ABPAP (24 patients) or ASV (50 patients). And, regardless of the degree of use (PAP Users vs Partial Users), there were also no differences based on PAP mode. However, our findings on these points are not reliable given the numbers were relatively small when we divided up these patients by their status as Users vs Partial Users and by ABPAP vs ASV modes. A more sophisticated level of research design would be required to compare or contrast ABPAP and ASV effects. 

One of the main clinical findings of the study was that many patients were using prescription sleeping pills or over-the-counter sleep aids, yet they were clearly failing these treatment approaches, and undoubtedly this failure explains at least part of their motivation to seek treatment at a sleep center. Moreover, their use of these medications combined with their strong endorsements of psychological factors as the primary cause of their sleep onset insomnia attests to the perspective that these individuals did not imagine a sleep breathing condition would prove to be an integral factor in their problems. The epitome of this mental framework was the ubiquitous problem of racing thoughts at bedtime. And, yet anecdotally, some patients reported a decrease in mental activation upon initiating and regularly use of advanced PAP therapy. These reports do not constitute evidence, but future studies must look carefully at whether or not racing thoughts are effectively decreased with advanced PAP. If so, such findings would strongly support a new theory on the underlying factors that cause racing thoughts as we have described previously, in which a psychological vantage point shifts to a physiological standpoint, albeit both could still play major roles in the development of sleep onset insomnia.

Along the same lines, there is no effort on our part to negate the potential role of medications or other forms of treatment for sleep onset difficulties such as sleep hygiene instructions or cognitive-behavioral therapy for insomnia (CBT-I). Rather, our work in this realm is about finding definitive or comprehensive treatment for various types of insomnia. We would be surprised if PAP therapy proves curative for a large proportion of patients with severe sleep onset difficulties, despite the provocative findings in our research. More commonly, sleep onset insomnia patients describe the problem of unfinished business when they hit the sack, which leads them to bring their work and worries into the bedroom and into bed. 

If advanced PAP therapy ultimately proves a potent technique to decrease racing thoughts, then such information would be a game-changer in the treatment of insomnia. Nevertheless, it would still be likely such patients would benefit from other improvements in their coping skills to resist bringing unfinished business into the bedroom. These skills could be directly improved with sleep hygiene, CBT-I, and a host of psychotherapy methods. Finding comprehensive care for these patients is our primary goal, but we remain concerned that so many professionals in sleep medicine continue to overlook the potential relationships between OSA/UARS and chronic insomnia. We trust this new research will catalyze greater interest in more vigorous research protocols that use randomized controlled trials to learn how we can assist insomnia patients, and when relevant, how to treat that part of their insomnia properly addressed with advanced PAP therapy.

Barry Krakow MD


Dr Krakow’s 27 years of sleep research have focused on the complex relationship between physiological and psychological sleep disorders. Dr Krakow currently operates private sleep medical center, Maimonides Sleep Arts & Sciences, Ltd., and serves as Classic SleepCare’s paid Medical Director.

Leave a comment

Comments have to be approved before showing up.