Journal of Sleep Disorders: Treatment and CareISSN: 2325-9639

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research Article, J Sleep Disor Treat Care Vol: 2 Issue: 1

Adaptive Servo-Ventilation Therapy in a Case Series of Patients with Co-morbid Insomnia and Sleep Apnea

Barry Krakow1-3*, Victor A Ulibarri1,2, Edward Romero4, Robert Joseph Thomas5 and Natalia McIver1,2
1Sleep & Human Health Institute, Albuquerque, NM, USA
2Maimonides Sleep Arts & Sciences, Ltd, Albuquerque, NM, USA
3Los Alamos Medical Center Sleep Laboratory, Los Alamos, NM, USA
4University of New Mexico School of Medicine, Albuquerque, NM, USA
5Division of Pulmonary, Critical Care & Sleep, Department of Medicine, Beth Israel Deaconess Medical Center, Boston, MA, USA
Corresponding author : Barry Krakow
Sleep & Human Health Institute, 6739 Academy N.E., Suite 380, Albuquerque, NM 87109, USA
Tel: (505) 998-7204; Fax: (505) 998-7220
E-mail: [email protected]
Received: November 27, 2012 Accepted: March 15, 2013 Published: March 18, 2013
Citation: Krakow B, Ulibarri VA, Romero E, Thomas RJ, McIver N (2013) Adaptive Servo-Ventilation Therapy in a Case Series of Patients with Co-morbid Insomnia and Sleep Apnea. J Sleep Disor: Treat Care 2:1. doi:10.4172/2325-9639.1000107


Adaptive Servo-Ventilation Therapy in a Case Series of Patients with Co-morbid Insomnia and Sleep Apnea

Study objectives: The study measured the effects and use of adaptive servo-ventilation (ASV) therapy among insomnia patients with co-morbid sleep-disordered breathing (complex insomnia) who failed standard positive airway pressure (PAP) therapy. Methods: A chart review was conducted on 56 consecutive patients with complex insomnia and self-reported anxiety or psychiatric distress to compare outcomes from their use of standard PAP therapy with their use of ASV. Measurements included polysomnographic changes in breathing events, sleep stages, sleep continuity markers, and adherence data downloads. Results: On standard PAP titrations, expiratory pressure intolerance and central apneas were highly prevalent, yielding a diagnosis of or subthreshold criteria for complex sleep apnea, after which patients received ASV therapy. Compared to standard PAP, ASV significantly improved objective breathing event indices. ASV also was associated with increased sleep efficiency, REM% sleep, and REM sleep consolidation as well as decreased awakenings, arousals, and time awake during the night. Among 39 of 43 current ASV users, adherence was significantly greater for nightly use and hours per night compared to prior use of standard PAP. Conclusions: Fifty-six complex insomnia patients failed standard PAP therapy and developed complex sleep apnea. ASV therapy was associated with improved sleep quality and increased device usage in this select sample.

Keywords: Adaptive servo-ventilation; Complex sleep apnea; Complex insomnia; Expiratory pressure intolerance; Insomnia; Sleep apnea


Adaptive servo-ventilation; Complex sleep apnea; Complex insomnia; Expiratory pressure intolerance; Insomnia; Sleep apnea

Main Findings

Emerging evidence shows the prevalence of insomnia and sleep apnea may be higher than expected and scant research addresses the use of PAP therapy in these patients. This study showed that ASV PAP therapy provided substantial improvements in a select sample of complex insomnia patients who previously failed standard PAP therapy devices.


Adaptive servo-ventilation (ASV) therapy is indicated for various types of central apnea events observed in sleep-disordered breathing (SDB) patients [1-5]. With ASV therapy, ventilation or flow or both are tracked, with averaged algorithmic compensation to counter the tidal volume oscillations of central apneas and periodic breathing [6]. ASV therapy may provide definitive treatment not only for obstructive breathing events but for central events as well and has been proven effective for patients suffering congestive heart failure with heightened chemoreflexes and circulatory delay [7-11] and for opiate patients with dampened respiratory drive [12,13].
Increased loop gain/heightened chemosensitivity that ultimately reduces arterial CO2 below the NREM sleep CO2 apnea/hypopnea threshold may trigger central apnea and periodic breathing events [14]. Loop gain is also implicated in “complex sleep apnea” (CompSA) in which exposure to CPAP or BPAP leads to iatrogenic (treatmentemergent) central apneas and other forms of respiratory control abnormality. Presumably, improved ventilation from positive airway pressure therapy leads to hypocapnia and ventilatory instability [15-18]. ASV is a proven therapy for CompSA [5,12,13,18,19]; possibly through the capacity to support the airway without inducing hypocapnia; this stabilization of CO2 levels fosters a relatively constant CO2-based respiratory drive [20].
Overall, the causes of treatment-emergent central apneas among PAP users are uncertain, and the above pathophysiological mechanisms may only reflect a subset of possible etiologies. In addition to this list, we speculate that centrally-mediated breathing events may be triggered by a simpler or mechanical process involving a feeling or sensation of intolerance to pressurized airflow on expiration subjectively while awake or objectively while asleep [21,22]. In this theory, a more direct irritant or mechanical process emerges wherein “excess” pressure triggers a central apnea. In its basic manifestation, we note “expiratory pressure intolerance” appears as subtle irregularities on the expiratory limb of the airflow curve. Although there is no scientifically validated or generally accepted waveform to denote expiratory pressure intolerance (EPI), we have observed that patients who report discomfort on exhalation with CPAP or BPAP while awake also frequently exhibit the waveform pattern seen in figure 1a while asleep. Thus, clinically we define EPI as the subjective report or objective finding of expiratory intolerance coupled with persistent low-frequency oscillatory waveform irregularities on the expiratory limb not otherwise explained by mouth breathing, mask leaks, or ballistocardiogenic effects. Persistent waveform indicates the pattern is present more often than not. As intolerance worsens objectively, central-like phenomena appear in the form of respiratory pauses of 3 to 9 seconds or less than the standard criteria for a central event (Figure 1b). In the extreme, the patient exhibits central apneas (Figure 1c) albeit it remains unknown whether this effect is mechanical, related to loop gain/hypocapnia, or both. Nakazaki has postulated a related process with respect to increased nasal resistance as a risk factor for complex sleep apnea in patients with normal heart function [23].
Figure 1: Thirty second epochs from titration polysomnography performed on a 43 y.o female (BMI 31.0; diagnostic sleep breathing indexes: AHI 9.7 events/hr, RDI 43.7 events/hr, and CAI 3.8 events/hr) showing expiratory intolerance manifesting as irregular deflections* and variability of expiratory duration on the expiratory limb of the airflow curve (Figure 1a), central-like pauses manifesting as extended end-expiratory phase pauses of the airflow curve lasting less than 10 seconds (Figure 1b), and central apneas (Figure 1c) in patients experiencing difficulty adapting to CPAP pressure. Also note the subtle variations in expiratory duration in Figure 1a, somewhat similar to but less overt than that seen in opiate-induced/associated sleep apnea. *A ballistocardiogenic effect may be a factor in these irregularities, but they disappear with ASV therapy.
a) Expiratory pressure intolerance. b) Central-like pauses. c) Central apnea.
Clinically, insomnia patients with sleep-disordered breathing (“complex insomnia”) [22,24] might be at risk for positive pressure intolerance for several reasons. These patients possess an increased level of somatic tension, which could manifest in the upper airway. Increases in sympathetic tone can exaggerate reflexes; and, it is plausible increased central adrenergic drive has effects at the brainstem level, altering the integration of chemoreceptor information. Theoretically, anxious insomnia patients may hyperventilate [25], especially if the mask feels intrusive, bringing end-tidal CO2 closer to the apnea threshold and amplifying the above effects. Insomnia patients have fragmented sleep and increased sleep-wake transitions. These periods may induce excess CO2 blow-off that also triggers central sleep apnea [16]. Decreased arousals with zolpidem use in high altitude-induced sleep apnea patients have also been shown to reduce susceptibility to central apneas [26]. Finally, in light of recent findings showing that complex insomnia patients may manifest a sizeable proportion of respiratory effort-related arousals (RERAs), the factors described above may come rapidly into play during a titration protocol that raises pressures in accordance with AASM standard to eliminate RERAs. In sum, insomnia patients with co-morbid sleep apnea may experience multiple risks for the development of complex sleep apnea when exposed to PAP therapy.
To investigate this phenomenon in complex insomnia (cooccurring insomnia and SDB), we studied a consecutive series of patients who failed standard CPAP or BPAP, that is they could not initiate treatment or they received limited or no benefits. All patients were eventually treated with ASV therapy for the diagnosis of or subthreshold criteria for complex sleep apnea. Compared to standard PAP therapy, we hypothesized:
1. ASV would decrease expiratory pressure intolerance as well as obstructive and central breathing events.
2. ASV would be associated with improvement in objective sleep quality.
3. ASV would be associated with greater device usage compared to their usage of CPAP or BPAP.


Sample and consent
This protocol was a retrospective chart review of a defined group of 56 consecutive complex insomnia patients titrated with and prescribed ASV at Maimonides Sleep Arts and Sciences (MSAS), a private, community based sleep medical center in Albuquerque, NM, operating at one mile above sea level. Per standard protocol at MSAS, all patients provide verbal and written consent for their medical information to be used anonymously for research and educational purposes in the context of subsequent chart and data reviews. This project has been reviewed and approved by the Los Alamos Medical Center (LAMC) Institutional Review Board.
Inclusion and exclusion criteria
A consecutive series of patients were included who met research diagnostic criteria for an insomnia disorder [27] at intake; an objective diagnosis of SDB (obstructive sleep apnea [OSA]: AHI > 5 or upper airway resistance syndrome [UARS]: RDI > 15 and AHI < 5); completed a full night or split-night titration PSG with standard PAP therapy; and completed a full night or split-therapy titration with ASV therapy. All titration testing occurred from June, 2008 through November, 2009. Additional inclusion criteria included: 18 years of age or older; and spoke English. A history of congestive heart failure or daily opiate use-two common risks for central apneas-were exclusions.
To qualify for a diagnosis of complex sleep apnea, a patient was required to demonstrate a CAI/AHI ratio > 50% coupled with a CAI > 5/hr. Subthreshold criteria for complex sleep apnea included the presence of treatment-emergent central apneas on a titration, CAI < 5/hr, and failure to maintain use or receive benefit from a standard PAP device.
Self-Report Measures: All patients completed a web-based intake questionnaire set (2008) before any PSG testing. The questionnaires are based on sleep medicine nosology and mirror a standard sleep medicine interview. The Epworth Sleepiness Scale (ESS) [28] and the Insomnia Severity Index (ISI) [29] measured sleepiness and insomnia respectively while the Functional Outcomes of Sleep Questionnaire (FOSQ) [30] and visual analog scales (VAS) for fatigue [31] and sleepiness [32] measured impairment. Patients were queried on psychiatric history and related questions to determine additional insomnia or psychiatric symptoms or disorders.
Polysomnography (diagnostic & titration) sensors and scoring: Parameters recorded were electroencephalography (EEG: C3 to A2, C4 to A1, O1 to A2, O2 to A1), left and right eye electro-oculography (EOG), chin electromyography (EMG), electrocardiography (EKG 1, EKG 2), blood oxygen saturation and pulse rate via pulse oximetry probe, left and right anterior tibialis EMG, airflow was monitored through built in pressure transducer inside the PAP device with output of flow and pressure recorded by DC channel in polygraph, (for diagnostic: airflow measured with nasal cannula pressure transducer and thermistor/thermocouple), chest and abdomen respiratory effort via two piezo-electric plethysmography sensor belts, and snoring via snore sensor.
PSG tracings were scored by registered polysomnography technologists. Staging of sleep was done using current American Academy of Sleep Medicine scoring guidelines. Respiratory-related events were of four types: obstructive apneas were a > 90% reduction in airflow lasting > 10 seconds; central apneas were a > 90% reduction in airflow lasting > 10 seconds and associated with an absent inspiratory effort throughout the entire period of absent airflow; hypopneas were a > 50% reduction in baseline nasal pressure signal excursion lasting > 10 seconds with a ≥ 3% desaturation from pre-event baseline or terminated with arousal or awakening; and respiratory effort-related arousals (RERAs) were a sequence of breaths lasting > 10 seconds characterized by increased respiratory effort or flattening of the nasal pressure waveform leading to an arousal or awakening in EEG activity when the sequence of breaths does not meet criteria for an apnea or hypopnea. The AI is the number of obstructive and central apneas per hour of sleep, the AHI is the number of apneas plus hypopneas per hour of sleep, and the CAI is the number of central apneas per hour of sleep. The RDI is the AHI added to the number of respiratory effortrelated arousals (RERAs) per hour of sleep.
Polysomnography collection data: All patients met screening criteria for diagnostic PSG testing based on the medical director’s review of patient intakes, and all 56 patients completed one or more PSG tests at MSAS. Data for use in this retrospective chart review were collected from three types of tests: (1) diagnostic information was captured from a full night or split-night diagnostic/titration study; (2) qualifying information for the diagnosis of complex sleep apnea (CompSA) was captured from either a full night titration with standard PAP therapy or a split-therapy protocol in which the first half of the study qualified the patient for ASV therapy [18]; and (3) ASV data were captured from their most recent ASV titrations, whether it was a full night of ASV therapy or their original and only use of ASV during a split-therapy protocol.
Polysomnography Titration Protocols: All PSG were performed at MSAS according to the guidelines from the American Academy of Sleep Medicine. In addition, we used a protocol developed at our center (Figure 2) to specifically address more complex patients who show persistent difficulty adjusting to expiratory pressures, residual RERAs, and related signs of poor sleep consolidation (e.g. low sleep efficiency, excess arousals, awakenings, and stage shifts, and persistent REM sleep fragmentation). In general, expiratory pressure intolerance [33] appears commonly in anxious or insomnia patients with co-morbid sleep breathing problems when they are exposed to PAP therapy [34-38]. Although rarely described in the sleep medicine literature, expiratory pressure intolerance (EPI) can be detected both subjectively and objectively during the pre-sleep desensitization period or a PAP-NAP [22]; and, in a large proportion of these types of patients, the problem is severe enough to establish a degree of “CPAP failure” that requires switching to an expiratory pressure relief device [(EPR) e.g. CPAP w/EPR, APAP w/EPR, BPAP] at the outset of a titration test [39].
Figure 2: Five-step titration algorithm to eliminate residual RERAs, Expiratory pressure intolerance (EPI), Iatrogenic central apneas, and persistent sleep fragmentation.
1 – Persistent RERAs: Point where obstructive apneas and hypopneas are eliminated on a standard, fixed pressure PAP device per AASM Clinical Guidelines (see reference #40).
T1 – Follow AASM titration guidelines for elimination of RERAs: CPAP: increase CPAP > 1 cm H2O if > 5 RERAs present.
2 – Expiratory Pressure Intolerance (EPI) Emerges: Initial appearance of EPI during titration. Follow AASM Guidelines to decrease pressure to a level that patient reports as comfortable and allows return to sleep.
T2 – Continue titration to treat residual RERAs.
3 – EPI Persists: EPI cannot be eliminated despite following AASM guidelines. Switch to advanced pressure device that provides clear expiratory relief [CPAP w/ EPR, Auto-CPAP (APAP) w/ EPR or BPAP].
T3– Resume titration at pressures similar to highest reached prior to onset of EPI. Focus on relief of EPI through expiratory relief adjustments while continuing efforts to eliminate residual RERAs. Work in incremental pressure increases or decreases of 0.2–0.4 cm H20 as effective and tolerated.
4 – Persistent Sleep Fragmentation: Intractable EPI, RERAs, or signs of unresolved sleep fragmentation, including excess arousals, awakenings, stage shifts, or failure to generate or consolidate REM sleep. Switch to Auto- Bilevel (ABPAP) device.
T4– Manually titrate ABPAP device while in auto mode to resolve EPI, RERAs, or sleep fragmentation using specific technology to fine tune settings (e.g. RESMED Easy Breathe Algorithm, Cycle/Trigger/Exhale Settings, TiMin, TiMax) while using incremental pressure increases or decreases of 0.2–0.4 cm H20 as effective and tolerated to adjust Max IPAP, Min EPAP, and PS (Pressure Support).
5 – Complex Sleep Apnea (CompSA): Iatrogenic central apneas appear to arise when EPI does not resolve or when higher pressures are required to generate or consolidate REM. If Medicare criteria attained [(CAI) > 5 and CAI/ AHI > 50%] switch to adaptive servo-ventilation (ASV) or consider ASV for subthreshold CompSA in a patient who fails all other devices.
T5– Manually titrate ASV device by increments of 0.2–0.4 cm H2O units to normalize the airflow signal focusing on Min PS for inspiratory flow limitation (RERAs), MaxPS for central apneas, and EPAP settings to normalize expiratory flow curve without triggering EPI.
To normalize the airflow curve by eliminating RERAs (per AASM guidelines), we have observed the utility of raising inspiratory pressures with an advanced device that provides some form of expiratory pressure relief. Sequential pressure delivery mode changes ensue, usually commencing with CPAP (if CPAP failure did not occur during pre-sleep desensitization), then moves to either CPAP or APAP with an EPR setting or to BPAP, then to ABPAP (auto-bilevel) and finally ASV when all else fails or the patient develops complex sleep apnea (CompSA). Titration changes tend to be incremental in an attempt to eliminate RERAs without aggravating expiratory pressure intolerance; gradual changes in settings from 0.2 to 0.4 cm H20 are common throughout the night (Figure 2).
Last, in applying our protocol to eliminate RERAs as well as EPI, we have also noticed a set of objective markers that tend to correlate with a normal airflow signal-normal on both inspiration and expiration. Specifically, a normalized airflow signal appears linked to fewer arousals, awakenings, and sleep stage shifts [40] as well as a greater degree of sleep efficiency, REM rebound [41] and possibly REM sleep consolidation (uninterrupted periods of REM sleep) [42]. On this latter point, Insana and colleagues have described the first use of a REM sleep fragmentation metric and noted its correlation with worse psychiatric distress in a cohort of posttraumatic stress patients (exposed to early life trauma) [43]. Our model represents the converse approach: we score the degree of REM consolidation by counting the total number of REM minutes and dividing it by the number of discrete REM periods no matter how small the duration of each interval of REM sleep (Figure 3), which is not dissimilar to Lavie’s rendering in 2001 of REM fragmentation and consolidation comparing normal sleepers to trauma survivors [44].
Figure 3: Comparison of: a) diagnostic; b) standard PAP titration; and c) ASV titration hypnograms showing progression from REM fragmentation to REM consolidation as well as progression from excess sleep stages shifts to marked decrease in sleep stage shifts.
a) Diagnostic- REM Fragmentation (REM Consolidation Index* = 2.83). b) Standard PAP- Persistent NREM and REM Fragmentation (REM Consolidation Index = 8.79). c) ASV- REM Consolidation (REM Consolidation Index = 15.5).
CPAP Failure determinations standard PAP failure was defined as follows: (1) Prescribed CPAP never initiated at home; (2) prescribed CPAP device initiated at home for a brief trial, then discontinued and not re-initiated; and (3) prescribed CPAP device initiated at home but use was sporadic, minimal or without benefit. Categories #1 and #2 comprised the failure type designated “noninitiators,” and category #3 comprised the failure type designated “poor responders”. Coupled with this initial determination of CPAP failure, all patients re-confirmed objective CPAP failure during the pre-sleep desensitization (n = 24) or at the outset of their titrations (n=32) in the sleep lab. Accordingly, for the cohort of 56 patients in this chart review, the bulk of titration time applied various BPAP devices, including the Puritan Bennett 425 GoodKnight, the ResMed VPAP/ST, or the Respironics BiPAP Synchrony. All ASV titrations utilized the ResMed VPAP Adapt SV device.
ASV Prescriptions and Follow-up: All patients were prescribed ASV therapy by the medical director and were scheduled for set up with the device. As per standard protocol at MSAS, follow-up was attempted with each patient in order to assess adherence with ASV therapy via objective data downloads (ODD) of the devices. The chart review was completed in 2011.
Data analyses: There were no systematic differences between those who presented as non-initiators (initial failure) (n=23) and those who adapted poorly (subsequent failure) (n=33) when comparing sociodemographics, subjective sleep symptoms, and objective diagnostic data; therefore, the sample was collapsed into one group of 56 patients for final analyses. The primary analysis examined objective data with repeated measures ANOVA (using SPSS v. 11.0.1) of the 3 time points for diagnostic, standard PAP titration qualifying patients for ASV, and the most recent ASV titration data, for which we assessed the following objective parameters: sleep stages N1%, N2%, N3%, REM%, AI, AHI, RDI, CAI, arousal index, awakenings index, sleep stage shifts, REM sleep consolidation, wake after sleep onset time, and sleep efficiency. As applicable, per hour of sleep metrics were used for variables to combine data from full night and splitnight or split-therapy protocols. Mann-Whitney U test was used for unequal sample sizes comparing users and non-users of ASV therapy. For patients who provided data downloads after established on ASV therapy, we conducted a two-point repeated measures ANOVA for relevant variables comparing standard PAP to post-ASV. P values ≤ 0.05 were statistically significant. Cohen’s d measured effect sizes for within sample comparisons; and, Hedge’s g measured effect sizes for unequal samples. Standard effect sizes are typically gauged small (0.20), medium (0.50) and large (0.80 or above).


Sample characteristics
The sample comprised predominantly male (58.9%), marginally overweight [BMI mean (SD)=25.9 (2.1)], Caucasian (78.5%), married (67.9%) individuals with bachelor degrees or higher (57.1%). Selfreported psychiatric conditions were reported in 93% of the sample, predominantly anxiety or mood disorders. Racing thoughts in bed (55.4%), anxiety or fear in bed (35.7%), and the use of near nightly or nightly prescription sleep medications (25%) or over the counter sleep aids (17.9%) were highly prevalent.
Intake surveys showed mean (SD) ISI of 15.2 (4.5), self-reported sleep efficiency (SE) 77.7% (20.7) and subjective wake-after-sleeponset (WASO) 73.9 minutes (89.7), indicative of moderate insomnia. Despite mean (SD) baseline ESS being in the mild range for sleepiness [10.6 (4.8)], the baseline FOSQ [13.8 (3.6)] and VAS scales for sleepiness and tiredness [6.2 (2.2) and 6.6 (2.0) respectively] were in the moderate to severe range for impairment. All 56 total patients met criteria for an insomnia disorder, and insomnia was their chief complaint and primary sleep problem on intake. We also estimated the prior duration of PAP usage in our sample prior to initiating ASV therapy. Among the 23 non-initiators, use of a PAP device ceased in nearly all cases within the first month for the few patients who actually tried the device. Among the 33 poor responders, use of a PAP device averaged 20 months of traditional PAP therapy use before initiating ASV.
Diagnostic data and follow-up clinical encounters
During diagnostic testing, AHI averaged in the moderate to moderately severe range; and all patients had a clinical diagnosis of SDB (54 patients OSA; 2 patients UARS) with an abnormal but low mean rate of central apneas (2.4 events/hr) (Table 1). Prior to reaching the stage in which an ASV titration was conducted, the patients averaged 7.85 clinical encounters (clinic appointments, telephone follow-up appointments, mask fittings, data downloads, outcome measurements, and email exchanges), 3.47 sleep tests (diagnostic, titration, split night tests, and PAP-NAPs), and spent 11.5 months from the point of presenting to our sleep medical center to the ASV titration.
Table 1: Diagnostic, Standard PAP, and ASV titration sleep, Sleep-Disordered Breathing (SDB), and sleep continuity indexes (n=56).*Cohen’s d used to determine effect size for Diagnostic vs ASV (d1) and Standard PAP vs ASV (d2) comparisons.
Expiratory intolerance and complex sleep apnea: Figure 4 shows the pattern of longitudinal titrations, leading to the diagnosis of complex sleep apnea or subthreshold CompSA whereby many patients required retitration studies to eventually qualify for the use of ASV therapy. For example, note in the BPAP column of figure 4, 28 total titrations (white bar) were completed by our sample at various points in their clinical care of which 7 studies (black bar) resulted in a diagnosis of CompSA. In the BPAP-ASV column, 8 patients (white bar) underwent BPAP titrations that resulted in a same night CompSA diagnosis, which permitted the use of ASV for the remainder of the study in all 8 patients (black bar). Overall, 31 patients had full night standard PAP titrations leading to a diagnosis of CompSA and subsequent full night ASV titration; whereas, 25 patients had multi-modality titrations, during which CompSA (or subthreshold CompSA) was diagnosed in the first half of the study with standard PAP, allowing ASV therapy to be tested in the second half [18]. Standard PAP titrations resulted in expiratory pressure intolerance in all patients as well as a 600% increase in central apneas from the diagnostic study (Table 1). Fifty patients met conventional diagnostic criteria (CAI > 5; CAI > 50% of AHI) for CompSA (CAI = 17.8) and 6 patients were sub-threshold (mean CAI = 3.7) but were prescribed ASV due to failure with all other devices. When using ASV therapy, the average EEP or EPAP setting was 7.59 + 2.27 cm H20, which was somewhat lower than the 8.61 + 2.13 cm H20 CPAP or the EPAP setting of BPAP for standard PAP therapy devices.
Figure 4: Total titrations completed (N=96) (white bar) and total titrations resulting in diagnosis of CompSA (N=56) (black bar) based on pressure delivery mode(s) utilized during these titrations.
Impact of ASV therapy titration on breathing events
The use of ASV therapy markedly decreased all respiratory breathing event indexes (AI, AHI, RDI, and CAI). All ASV effect sizes for breathing event indexes were medium to large compared to either diagnostic (mean d=1.18) or standard PAP values (mean d=1.06) (Table 1). Notwithstanding, RDI data demonstrate the difficulty in eliminating RERAs in these patients as the contrasts between AHI and RDI in table 1 showed numerous breathing events (apneas and hypopneas) being “converted” to RERAs with ASV use, yet the RERA index itself was higher than either diagnostic or standard titration studies.
Additional Objective Markers of Sleep Stages and Continuity
Additional objective benefits measured on changes in sleep continuity were associated with ASV therapy. For REM% of sleep, the repeated measures test was highly significant (p=.0001) with medium to large effects for ASV compared to both diagnostic (d=0.62) and standard PAP (0.68) studies. However, these effects were partially explained by the use of split therapy protocols and the natural increase in REM sleep during the second half of the night. All six sleep continuity measures (arousals, awakenings, excess stage shifts, REM consolidation, WASO, and sleep efficiency) were significantly improved with ASV therapy with small to large effects in comparison to diagnostic results and generally medium effects in comparison to standard PAP tests (Table 1).
ASV Users vs. Non-Users
Of the total 56 patients, 8 were lost to follow-up, and 5 were no longer using ASV (2 used oral appliances), thus using a conservative intent-to-treat approach for ASV, the final breakdown was 43 current users and 13 non-users of ASV therapy at the conclusion of the chart review. All four sleep breathing indexes (AHI, AI, CAI, RDI) as measured on their latest ASV titrations were lower for current users compared to non-users with a significant difference and a large effect on the AHI metric (2.1 vs. 7.0; p=0.03; g=0.72). Each type of breathing event [apnea (0.36 vs. 0.91), hypopnea (1.7 vs. 6.1) and RERA (22.8 vs. 24.8)] trended lower for users compared to non-users, but these changes were not statistically significant.
Standard PAP vs. ASV Adherence Rates
Among 43 users currently on ASV, 39 provided adherence information on data downloads that were compared to traditional PAP adherence data or surrogate data (no prior use or estimates from clinical encounters). As seen in table 2, adherence for ASV compared to standard PAP for these 39 patients showed significantly more nights used, average hours on nights used, nights used > 4 hrs, and median hours used, suggesting improved adaptability. To highlight these effects, 13 patients who never initiated PAP therapy were now regular users of ASV therapy.
Table 2: Comparison of Standard PAP vs. ASV Objective Data Downloads (ODD) (n = 39).
Treatment outcomes and intercurrent treatment factors
At the point of this chart review for the 39 ASV users with data downloads, the average duration of adherence to this advanced device was 1.31 (SD 0.21) years. Among these 39 ASV users, insomnia and sleepiness questionnaires compared intake to post ASV implementation. ISI scores decreased significantly (16.8 vs. 9.7; p=.001; d=1.40), and ESS scores showed a trend toward a mediumsized decrease (11.2 vs. 9.3; p=0.11; d=0.37). Both these adherence and outcome findings must be viewed cautiously, however, due to intercurrent treatment influencing results. Specifically, among the 56 patients, 50 (89.3%) received coaching for insomnia, 35 (62.5%) utilized education materials for insomnia, 26 (46.4%) were started on an intensive nasal hygiene regimen, 22 (39.3%) were treated with medication for leg jerks, and 4 (7.1%) attended our nightmare treatment clinic.


In the current study, 56 complex insomnia patients with anxiety or other psychiatric disturbances manifested various forms of CPAP and BPAP failure. The most apparent physiological barrier to PAP adaptation was expiratory pressure intolerance, which appeared to culminate in or at minimum was associated with central sleep apneas upon exposure to positive airway pressure. This barrier was overcome with adaptive servo-ventilation using the ResMed Adapt SV device. Prior to initiation of ASV therapy, only 13% of our small, select sample reported any benefit, albeit marginal, from standard PAP therapy. No one reported an optimal response. Quantitatively assessed or estimated adherence among a subset of 39 patients prior to starting ASV therapy showed average standard PAP use at 31.8% of the time (~2 days/week) for an average of 2.4 hrs/night on nights used; whereas, once this subset of patients were prescribed ASV, average use increased to 71.8% (~5 days/week) for an average of 5.7 hrs/night on nights used. Thus, standard PAP usage equaled 5 hours/ week compared to ASV usage of 29 hours/week. Moreover, ASV titrations were associated with enhanced sleep consolidation in the form of greater objective sleep efficiency, REM sleep consolidation, possibly greater time spent in REM sleep, as well as fewer arousals, awakenings, and stage shifts, all of which are important prognostic markers for adherence when assessing the objective response to PAP therapy [45]. In sum, to our knowledge, this is the first study to demonstrate a potential utility and a theoretical explanation for the use of ASV therapy in challenging patients with the combination of insomnia and sleep-disordered breathing (“complex insomnia”).
Conventional wisdom argues many of our patients should have experienced fewer not more central apneas over time. Several papers indicate overall rates of complex sleep apnea are low among long term CPAP users, [46-48] albeit none of these studies or any other protocols appear to have evaluated or reported on the persistence of expiratory pressure intolerance in their samples. Thus, central apneas themselves may be the veritable tip of an iceberg in analyzing how some insomnia patients respond to PAP therapy. The more vexing problem may be the persistence of expiratory pressure intolerance in these patients with psychiatric co-morbidity. Indeed, given the difficulties observed in titrating RERAs in our sample, it could be argued central apneas may have been triggered by efforts to attain the guidelines set by AASM titration goals to normalize the airflow [49]. If similar findings were replicated by other research investigations, it would indicate a need for a more thorough inspection of expiratory pressure intolerance in PAP users, irrespective of treatment-emergent central apneas. To our knowledge, there have been no studies to define, measure or treat expiratory pressure intolerance, despite the frequent usage of the term, “pressure intolerance” in current literature on non-compliant PAP therapy patients [49,50].
Along the same lines, it is clinically noteworthy that several studies have been published to test whether or not advanced modes of PAP therapy such as those with expiratory pressure relief systems (CPAP with EPR or CFLEX), dual pressure symptoms (BPAP) or auto-adjusting devices (APAP or ABPAP) produce more effective treatment of breathing events or improved compliance rates in select samples [51-55]. While the outcomes have been mixed and often report minimal if any improvements above fixed pressure CPAP, the most commonly reported benefit has been greater comfort among patients using more advanced devices irrespective of adherence rates. Speculatively, none of these studies assessed or clearly treated expiratory pressure intolerance, and failing to do so may have accounted for a lack of improvement in adherence beyond that achieved with standard CPAP devices.
Moreover, among those studies using auto-adjusting devices, the researchers opted for constrained protocols in which the patient was only subjected to an “auto” mode instead of undergoing a manual titration by a sleep technologist in a sleep lab, that is, with a device set for “auto” mode and with a sleep tech present to override and adjust the system. As we have recently commented, such protocols are highly flawed because they make the unfounded assumption that the “auto” mode would eliminate RERAs and yield a normalized airflow curve [39]. To reiterate, the problem of persistent RERAs appears deeply intertwined with the emergence of expiratory pressure intolerance, which in the past few years has led us to examine the utility of the ASV devices in such patients. Unfortunately, others who report on the use of ASV in complex sleep apnea generally do not report on either RERAs or EPI, so comparisons cannot be drawn between their efforts to treat complex sleep apnea patients and those treated in this study. Nonetheless, future studies must investigate these relationships to provide information about sleep-disordered breathing and PAP therapy beyond issues related to complex sleep apnea.
Such research must also look at cost-effectiveness. As one cost-oriented example, it would behoove us to learn whether or not the “comfort” attributed to advanced devices would lead to fewer polysomnograms and more rapid achievement of adherence milestones. If so, an argument may be proffered that more expensive ASV devices are more cost-effective than standard CPAP devices due to more rapid adherence and therefore subsequent savings from fewer subsequent titrations in the sleep lab and greater use of the device. A cost-effectiveness model must also calculate whether we should rely less on the phenomenology of complex sleep apnea and pay more attention to how well a particular PAP device resolves RERAs, prevents expiratory pressure intolerance, and increases adherence rates. These speculations may only apply to patients like those in our sample who are both complex and challenging; it remains unknown whether other types of complex or challenging noncompliant PAP users would benefit from advanced ASV therapy. As several studies have shown, residual breathing events are a concern in non-adherent patients or those reporting suboptimal results [56-62].
Finally, emerging evidence indicates that sleep-disordered breathing may co-occur more frequently than previously recognized in patients with insomnia disorders [63]. Treatment experiences with complex insomnia cases show mixed results on how readily these patients adapt to PAP therapy [22,64,65] and whether or not insomnia influences adaptation to PAP therapy [22,66]. From a pathophysiological perspective, the response to ASV in these complex insomnia patients may be a function of respiratory control alterations, albeit this form of PAP therapy, as above, may simply feel more comfortable or natural for a patient with anxiety or somatized tension. Additional speculative mechanisms may include increased central adrenergic modulation of upper airway reflexes in response to positive pressure sensing in the upper airway, given the increased activity in the locus ceruleus and its target sites in anxiety disorders [67]. There may also be relevant associations due to increased sensitivity of the respiratory chemoreflex and anxiety-type symptoms [68,69]. Research must explore these speculations as well as develop clearer guidelines on how to best treat the OSA or UARS component of “complex insomnia”. Toward that end, greater application of polysomnography in the evaluation of appropriate chronic insomnia patients seems warranted in efforts to answer the questions raised by our findings.
This retrospective case series suffers limitations that restrict generalizability. It was a small sample and not a prospective randomized controlled design from which causality can be discerned. Our findings must be viewed with the same caution as with any case series that precludes validation of clinical evidence. While the work appears clinically relevant for patients in various stages of PAP failure, their improved adherence with ASV therapy only indicates a meaningful association. Moreover, the small sample offers no guidance as to the proportion of chronic insomnia patients with comorbid sleep-disordered breathing who suffer similar compliance problems with standard PAP therapy. Also, many other factors influence adaptability to PAP therapy such as mask changes, addition or removal of medications that influence sleep, or treatments that concurrently improve insomnia. The potential influence of an elevation of one mile above sea level may add unique risks for complex sleep apnea. Finally, the study also points to the difficulty in eliminating RERAs during a titration study, suggesting that technological barriers still exist in our efforts to normalize airflow in patients with sleep breathing disorders, albeit decreasing all sleep breathing events is still presumed to be associated with greater usage of a PAP device [49].
In sum, the literature implies that patients who maintain use of standard PAP devices should expect adjustment problems to subside; specifically, those with initial treatment-emergent central apneas should experience decreased central apneas over time [47]. This effect did not occur for our select sample of insomnia patients with anxiety and other psychiatric disturbances, who were diagnosed with comorbid SDB. Instead, expiratory pressure intolerance was pervasive over a long interval and may have been a critical factor leading to the diagnosis of complex sleep apnea or subthreshold complex sleep apnea. Regardless of the underlying pathophysiology, ASV therapy may confer special benefits to this cohort of patients, even if the essential therapeutic effect is one of optimized pressure comfort [70]. Prospective studies are needed to more precisely understand the barriers and solutions to standard PAP therapy adherence among complex insomnia patients.


Track Your Manuscript

Recommended Conferences

8th International Conference on Epilepsy & Treatment

Amsterdam, Netherlands