REM rebound and CPAP compliance

REM rebound and CPAP compliance

Sleep Medicine 13 (2012) 864–868 Contents lists available at SciVerse ScienceDirect Sleep Medicine journal homepage: www.elsevier.com/locate/sleep ...

207KB Sizes 0 Downloads 7 Views

Sleep Medicine 13 (2012) 864–868

Contents lists available at SciVerse ScienceDirect

Sleep Medicine journal homepage: www.elsevier.com/locate/sleep

Original Article

REM rebound and CPAP compliance Brian B. Koo a,b,⇑, Roger Wiggins c, Carol Molina c a

Case Western Reserve University School of Medicine, United States Louis Stokes Veterans Affairs Medical Center, Department of Neurology, United States c Louis Stokes Veterans Affairs Medical Center, Department of Pulmonary, Critical Care and Sleep Medicine, United States b

a r t i c l e

i n f o

Article history: Received 6 October 2011 Received in revised form 27 February 2012 Accepted 5 March 2012 Available online 15 June 2012 Keywords: Sleep apnea CPAP Compliance REM rebound SWS rebound OSA

a b s t r a c t Objective/Background: The objective of this study was establish if rapid-eye-movement (REM) rebound on first exposure to continuous positive airway pressure (CPAP) is associated with CPAP compliance. A rebound or drastic increase in REM sleep in response to initial CPAP exposure is associated with improvement in the subjective quality of sleep. We wished to determine if REM rebound was also associated with increased CPAP compliance. Methods: Split night polysomnographic studies carried out in a one-and-a-half year period were examined for REM rebound and slow wave sleep (SWS) rebound. Compliance with CPAP according to percentage of days used and percentage of days used for more than 4 h was determined at 30, 60, and 120 days and compared between groups with and without REM rebound and then between groups with and without SWS rebound. Multivariate regression models were constructed to determine factors that were associated with increasing CPAP compliance. Results: CPAP compliance was greater for those with REM rebound than those without REM rebound at all time periods, but significantly so only for total percentage of days used at 30 days (86.7 ± 46.7, 96.7 vs. 56.7 [median ± 1st quartile, 3rd quartile] ± 32.5, 90.0; p = 0.04) and 60 days (78.3 ± 37.5, 93.4 vs. 50.0 ± 25.0, 80.9; p = 0.03). There was no difference in CPAP compliance for SWS rebound and there were no SWS rebound groups. Only the presence of REM rebound was associated with increased compliance with CPAP with neither SWS rebound nor diagnostic AHI being significantly associated with CPAP compliance. Conclusions: The presence of REM rebound, but not SWS rebound, on initial CPAP exposure is associated with early CPAP compliance. This increased compliance is not explained by severity of sleep apnea as measured by AHI. Published by Elsevier B.V.

1. Introduction Continuous positive airway pressure (CPAP) is very effective in the treatment of obstructive sleep apnea (OSA). CPAP acts as a pneumatic splint to maintain airway patency, allowing for normal breathing and uninterrupted sleep [1]. In up to 60–70% of the OSA population, effective treatment cannot be achieved because of poor CPAP compliance [2,3]. Studies looking at predictors of CPAP compliance have consistently identified perceived benefit as a factor associated with increased compliance [4–6]. While perceived benefit has most often been in the form of reduced daytime sleepiness, improved sleep quality has also been linked to improved CPAP compliance [6]. In untreated OSA, cortical arousal is often necessary to reopen the obstructed airway and to restore breathing. When this occurs ⇑ Corresponding author at: Louis Stokes Veterans Affairs Medical Center, Department of Neurology, 10701 East Blvd., Cleveland, OH 44106, United States. Tel.: +1 216 791 3800x5780; fax: +1 216 707 6401. E-mail address: [email protected] (B.B. Koo). 1389-9457/$ - see front matter Published by Elsevier B.V. http://dx.doi.org/10.1016/j.sleep.2012.03.019

repetitively, light non-rapid-eye-movement (NREM) sleep increases at the expense of both slow wave (SWS) and rapid-eyemovement (REM) sleep, resulting in relative REM and SWS deprivation. By restoring normal breathing, CPAP reduces arousal and permits normal sleep cycling. On initial exposure, CPAP can precipitate an increase or rebound in either REM or SWS and a decrease in stage N1 sleep [7,8]. This rebounding of REM and SWS during an initial night of in-laboratory CPAP titration has been shown to correlate with improvement in subjective sleep quality [9]. Since increases in both SWS and REM sleep tend to occur simultaneously in response to CPAP initiation, it is difficult to assess how much symptomatic relief is attributable to restoration in each of the sleep stages separately. Some studies have shown that REM increases to a greater extent than SWS in response to first CPAP exposure [7,10], suggesting that REM restoration may play a more significant role in sleep quality improvement. One might also expect that individuals who display the greatest improvement in objective sleep architecture would also be more compliant with CPAP usage. This raises the question of whether an improvement

B.B. Koo et al. / Sleep Medicine 13 (2012) 864–868

in objective sleep measures during initial CPAP treatment might be a factor associated with better compliance with CPAP. If an initial rebound to REM sleep as opposed to SWS is associated with greater symptomatic benefit, then one might expect that individuals experiencing REM rebound more strongly than SWS rebound would have greater CPAP compliance. Such a finding would have implications about why daytime symptoms in obstructive sleep apnea emerge and to what extent do disruptions in NREM versus REM sleep play a role in this. We decided to study this question and expected to see that patients with a rebound of REM sleep but not slow wave sleep in their initial treatment with CPAP would have better CPAP compliance in subsequent months.

2. Methods All split-night studies performed at the Louis Stokes Veterans Affairs Medical Center (LSVAMC) from January 1, 2008 to July 1, 2010 were examined. The Institutional Review Board of LSVAMC approved the conduct of this study. Overnight polysomnography (Embla; Denver, CO) was conducted with the use of electroencephalography, bilateral electrooculography, submental electromyography, thoracic/abdominal respiratory inductance plethysmography, naso-oral thermistry, nasal pressure transduction, oximetry, and EKG. By protocol, patients were split to CPAP after a minimum of 2 h if there was an apnea-hypopnea index of at least 30 events per hour. Apneas were identified by the near absence of airflow determined by thermistry for at least 10 s, while hypopneas were identified by 50% or greater reduction in nasal pressure transduction for at least 10 s plus a 3% or greater desaturation or arousal. The apnea–hypopnea index (AHI) was calculated as the total number of apneas and hypopneas per hour of sleep. Subjects were excluded from the study if any of the following criteria were present: Cheyne-Stokes breathing or central sleep apnea on diagnostic portion of study, history of congestive heart failure, death, no compliance data available, or no appropriate CPAP pressure found. The records of 241 subjects were reviewed. Of these subjects, 146 were excluded, 13 for congestive heart failure, three for death, 23 for Cheyne–Stokes breathing or central sleep apnea, 19 for no optimal CPAP pressure found, and 88 for no follow up or no compliance data available. Demographic characteristics of the excluded subjects were examined and were found to be similar to the included subjects. The following demographic and patient information was collected: age, sex, body mass index in kg/m2 (BMI), Epworth Sleepiness Scale (ESS) score, and use of REM suppressant medication. From the diagnostic portion of the studies, we examined the apnea-hypopnea index (AHIDx), slow wave sleep percentage (%SWSDx), and REM sleep percentage (%REMDx). From the treatment portion of the studies the following variables were examined: REM latency, slow wave sleep percentage (%SWSRx), REM sleep percentage (%REMRx), and duration of the longest REM period. Compliance data was obtained using downloaded information from CPAP machines that were later entered into a central compliance tracker. CPAP compliance was examined at 30, 60, and 120 days. Two metrics were used to determine CPAP compliance: total percent days used and percent days used for P4 h. Compliance in percent days used in total was measured for 30 days (complianceT30), 60 days (complianceT60), and 120 days (complianceT120), and percent days used for P4 h were also measured at 30 days (complianceF30), 60 days (complianceF60), and 120 days (complianceF120). REM rebound was defined using the following criteria: (1) At least one REM period of P30 min duration in the treatment portion and (2) An increase in REM in the treatment portion of

865

P20% (e.g., for a diagnostic REM percent of 2% and a treatment REM percent of 45%, 45% 2% = 43%, which is greater than 20%). Slow wave sleep rebound was defined as an increase in SWS in the treatment portion of P10%. Of our final 95 included subjects, 35 had REM rebound and 60 did not have REM rebound; 17 subjects had SWS rebound and 78 did not have SWS rebound. 2.1. Statistical Analysis Subjects were separated into REM rebound vs. no REM rebound categories and SWS rebound vs. no SWS rebound based upon the above criteria. Mean and standard deviation were determined for age, BMI, ESS, AHIDx, %REMDx, %REMRx, %SWSDx, %SWSRx, and each of the compliance measures for the entire group and then the REM rebound, no REM rebound, SWS rebound, and no SWS rebound groups. Frequency was determined for use of REM suppressant medication. Student’s t tests used to determine if there were differences in means between the REM rebound and no REM rebound groups, and then the SWS rebound and no SWS rebound groups in age, BMI, ESS, AHIDx, %REMDx, %REMRx, %SWSDx, and %SWSRx. Because of a non-normal distribution in the compliance measures, non-parametric two sample Wilcoxon tests were used to compare medians of each of the compliance measures: complianceT30, complianceT60, complianceT120, complianceF30, complianceF60, and complianceF120. Additionally, the cohort was split by severity using the AHIDx to determine if compliance differed by severity with AHI of 30 and then 60 used as cut-off points. Three groups were used AHI < 30, AHI between 30–60, and AHI > 60. Kruskal–Wallis tests were used to determine if median values differed in each of the compliance measures among the three different AHI severity groups. Finally, multivariate linear regression was performed to determine if age, BMI, AHIDx, REM rebound, or SWS rebound was associated with compliance at 30, 60, or 120 days using the two metrics of compliance. Statistics were carried out using the R statistical package (Auckland, New Zealand) version 2.13.1. 3. Results Reflecting the Veterans Affairs clinical population, there were 87 males and eight females in the overall cohort of 95 subjects. Of this total number, 35 subjects had REM rebound and 60 did not have REM rebound; 17 subjects had SWS rebound and 78 did not have SWS rebound. For the overall group, the duration (total recording time) of the diagnostic and treatment portions of the split night study are as follows: 2.8 ± 0.7 and 4.5 ± 0.6 h. These duration measures did not differ for REM rebound and no REM rebound groups. Table 1 summarizes characteristics of the overall cohort and cohort by REM and SWS rebound status. In general, the cohort was middle-aged, obese, and had severe sleep apnea. Subjects with REM rebound tended to be younger (54.1 ± 11.1 vs. 58.2 ± 10.9; p = 0.08), trended toward taking a REM suppressing medication more often (45.7% vs. 30.0%; p = 0.12), and had significantly more severe OSA (54.1 ± 29.6 vs. 39.0 ± 27.1; p = 0.02) than subjects without REM rebound. The REM rebound and no REM rebound groups did not differ in the degree of obesity or in the level of subjective sleepiness. As expected, the REM rebound group had significantly less REM sleep on the diagnostic portion and significantly more REM sleep on the treatment portion of the study. SWS percentages were no different between the two groups on either portion of the study. Data are also displayed in Table 1 for the SWS rebound and no SWS rebound groups. Age, BMI, and taking REM suppressing medications were no different. The SWS rebound group trended toward being less sleepy subjectively by ESS (9.3 ± 5.5 vs. 12.3 ± 5.4;

866

B.B. Koo et al. / Sleep Medicine 13 (2012) 864–868

Table 1 Baseline characteristics for the overall cohort and cohort by REM rebound and SWS rebound status.

Age BMI (kg/m2) ESS AHIDx REM suppress (%) %REMDx (%) %REMRx (%) %SWSDx (%) %SWSRx (%) * **

Overall (n = 95)

REM Rebound

No REM Rebound

*

56.7 ± 11.1 35.0 ± 7.0 11.7 ± 5.5 44.5 ± 28.8 35.8 6.7 ± 9.3 23.3 ± 15.1 18.4 ± 16.9 14.9 ± 12.6

54.1 ± 11.1 36.0 ± 6.5 11.8 ± 5.3 54.1 ± 29.6 45.7 2.3 ± 6.5 35.8 ± 13.1 15.4 ± 15.1 14.9 ± 12.0

58.2 ± 10.9 34.4 ± 7.3 11.6 ± 5.7 39.0 ± 27.1 30.0 9.2 ± 9.8 16.0 ± 10.9 20.1 ± 17.8 14.9 ± 13.1

0.08 0.29 0.90 0.02 0.12 0.001 0.001 0.18 0.99

P-value

SWS Rebound

No SWS Rebound

**

56.1 ± 12.6 36.3 ± 6.7 9.3 ± 5.5 60.6 ± 30.6 41.2 2.2 ± 6.5 21.7 ± 13.7 4.5 ± 6.1 29.1 ± 12.4

56.9 ± 10.8 34.7 ± 7.1 12.3 ± 5.4 41.0 ± 27.4 34.6 7.7 ± 9.6 23.6 ± 14.5 21.4 ± 17.0 11.8 ± 10.4

0.81 0.40 0.06 0.02 0.73 0.007 0.62 0.001 0.001

P-value

Student’s t test for comparison of REM rebound and no REM rebound. Student’s t test for comparison of SWS rebound and no SWS rebound.

Fig. 1. Compliance by REM rebound. Displayed is CPAP compliance, percentage of total days, at 30, 60, and 120 days for those with and without REM rebound. Compliance at all time points was greater for the REM rebound group. Significantly greater compliance was seen for percentage of days used at 30 and 60 days for the REM rebound group.

p = 0.06) and had significantly more severe OSA (60.6 ± 30.6 vs. 41.0 ± 27.4; p = 0.02). Diagnostic REM percentage was significantly decreased for the SWS rebound group but no different on the treatment portion, and, as expected, SWS was significantly less, and then more, on the diagnostic and treatment portions, respectively, for the SWS rebound group. Compliance was measured at 30, 60, and 120 days as percentage of days used in total and percentage of days used for more than 4 h. Compliance data are presented as median ± 1st and 3rd quartile values. In general, all compliance measures are greater for the REM rebound group than for the no REM rebound group at all time periods, but more for 30 and 60 days than 120 days. For the REM rebound group, significantly better compliance was seen at 30 days (86.7 ± 46.7, 96.7 vs. 56.7 ± 32.5, 90.0; p = 0.04) and 60 days (78.3 ± 37.5, 93.4 vs. 50.0 ± 25.0, 80.9; p = 0.03) for percentage of days used and trends for better compliance at 30 days (40.0 ± 8.4, 81.7 vs. 26.7 ± 6.1, 50.8; p = 0.08) and 60 days (45.0 ± 7.5, 82.5 vs. 20.9 ± 3.7, 51.7; p = 0.10) for percentage of days used P4 h. Percent days used in the REM rebound and no REM rebound groups are presented in Fig. 1 and in tabular format along with percent days used more than 4 h in Table 2. Compliance did not differ for SWS rebound and no SWS rebound groups at any time or any compliance measure with pvalues all exceeding 0.25 and generally being above 0.5. Compliance was also determined for groups stratified by severity with AHI > 30 and AHI > 60 and is shown in Table 3. Compliance was

significantly greater for those with AHI > 30 than those with AHI < 30 for many of the compliance measures. Having an AHI > 60 conferred no additional benefit to compliance at any time period and, in fact, trended toward having worsening compliance at the 120 day time point. Multivariate linear regression was carried out for each of the compliance measures at 30, 60, and 120 days using age, BMI, AHIDx, REM rebound (factor), and SWS rebound (factor) as covariates. Of all the time points and measures, only the presence of REM rebound was significantly associated with compliance, and only for total usage at 30 days. There were trends for the other measures of compliance at 30 and 60 days, again only for REM rebound. None of these factors was associated with compliance at 120 days. ComplianceT30 increased significantly by 15.3% when there was REM rebound. Coefficients and p-values are shown in Table 4.

4. Discussion This study provides evidence that rebounding of REM sleep on the first exposure to CPAP is associated with increased CPAP compliance at 30 and 60 days but not at 120 days. CPAP compliance was also increased for those with severe OSA with an AHI greater than 30 compared to those with AHI less than 30, but it was no different for persons with or without a rebound of SWS on first exposure to CPAP. Finally, when looking at factors associated with CPAP

867

B.B. Koo et al. / Sleep Medicine 13 (2012) 864–868 Table 2 CPAP compliance by REM rebound status.

ComplianceT30 (%) ComplianceF30 (%) ComplianceT60 (%) ComplianceF60 (%) ComplianceT120 (%) ComplianceF120 (%)

Overall

REM Rebound

No REM Rebound

P-value

73.3 ± 40.0, 96.7 33.3 ± 6.7, 60.0 66.7 ± 30.0, 88.3 33.3 ± 5.0, 61.7 50.9 ± 20.4, 82.7 27.5 ± 4.6, 64.3

86.7 ± 46.7, 96.7 45.0 ± 7.5, 82.5 78.3 ± 37.5, 93.4 40.0 ± 8.4, 81.7 67.5 ± 21.7, 85.0 41.7 ± 4.2, 69.2

56.7 ± 32.5, 90.0 20.9 ± 3.7, 51.7 50.0 ± 25.0, 80.9 26.7 ± 6.1, 50.8 43.3 ± 20.0, 80.8 26.7 ± 5.8, 43.3

0.04 0.08 0.03 0.10 0.39 0.29

ComplianceT30: percent overall days used in the first 30 days; complianceF30: percent days used P4 h in the first 30 days; complianceT60: percent overall days used in the first 60 days; complianceF60: percent days used P4 h in the first 60 days; complianceT120: percent overall days used in the first 120 days; complianceF30: percent days used P4 h in the first 120 days.

Table 3 Compliance by OSA severity.

ComplianceT30 (%) ComplianceF30 (%) ComplianceT60 (%) ComplianceF60 (%) ComplianceT120 (%) ComplianceF120 (%)

AHI < 30 (n = 33)

AHI 30–60 (n = 40)

AHI > 60 (n = 22)

*

53.3 ± 36.7,83.3 23.3 ± 6.7, 50.0 44.2 ± 22.9, 78.3 15.0 ± 6.3, 48.0 42.8 ± 12.5, 57.5 9.2 ± 5.8, 28.3

86.7 ± 46.7, 96.7 40.0 ± 9.2, 75.8 76.7 ± 37.9,95.8 39.2 ± 6.3, 80.4 79.2 ± 41.3, 93.0 50.8 ± 9.2, 75.9

71.7 ± 34.2, 95.9 40.0 ± 6.6, 66.7 65.0 ± 20.0,90.0 40.0 ± 3.3, 68.3 46.3 ± 15.0, 74.2 27.5 ± 2.3, 57.3

0.09 0.12 0.08 0.22 0.18 0.06

P-value

**

P-value

0.03 0.04 0.03 0.08 0.01 0.02

***

P-value

0.27 0.69 0.22 0.57 0.06 0.17

ComplianceT30: percent overall days used in the first 30 days; complianceF30: percent days used P4 h in the first 30 days; complianceT60: percent overall days used in the first 60 days; complianceF60: percent days used P4 h in the first 60 days; complianceT120: percent overall days used in the first 120 days; complianceF30: percent days used P4 h in the first 120 days. * Kruskal–Wallis test; comparison among three groups. ** Wilcoxon test for comparison of AHI < 30 and AHI 30–60. *** Wilcoxon test for comparison of AHI 30–60 and AHI > 60.

Table 4 Predictors of CPAP compliance. Dx AHI ComplianceT30 (%) ComplianceF30 (%) ComplianceT60 (%) ComplianceF60 (%) ComplianceT120 (%) ComplianceF120 (%)

0.03 0.07 0.02 0.04 0.10 0.17

(0.67) (0.59) (0.88) (0.74) (0.55) (0.30)

REM rebound 15.3 14.1 14.3 13.8 7.0 8.7

(0.04) (0.06) (0.06) (0.08) (0.44) (0.34)

SWS rebound 5.0 1.2 5.0 2.7 17.7 17.0

(0.58) (0.89) (0.60) (0.78) (0.16) (0.17)

⁄ Shown are multivariate coefficients with p-values in parentheses. ComplianceT30: percent overall days used in the first 30 days; complianceF30: percent days used P4 h in the first 30 days; complianceT60: percent overall days used in the first 60 days; complianceF60: percent days used P4 h in the first 60 days; complianceT120: percent overall days used in the first 120 days; complianceF30: percent days used P4 h in the first 120 days.

compliance in multivariate modeling, only the presence of REM rebound was associated with increased compliance; neither SWS rebound nor the diagnostic AHI were associated with CPAP compliance. CPAP can be a challenge for patients with OSA to use, and encouraging CPAP compliance is a priority for clinicians treating patients with sleep apnea. Identifying factors that are associated with CPAP compliance may help the clinician understand patterns of CPAP usage and non-usage and inform decision-making in an effort to optimize CPAP compliance. A modest association between OSA severity and CPAP compliance has been demonstrated by several studies [11,12]. Symptom severity in the form of daytime sleepiness may be a modulator of this interaction between OSA severity and CPAP compliance as subjective sleepiness predicts CPAP compliance regardless of OSA severity and CPAP usage in persons with severe sleep apnea, and little daytime sleepiness is low [13,14]. In OSA, the emergence of daytime symptoms may be more related to the degree of sleep disruption than to AHI. The degree to which sleep changes with first exposure to CPAP is in some way a measure of pre-existing sleep disruption with large changes in REM and SWS, likely reflecting significant prior sleep fragmentation. This is the basis for the rebounding of both REM and SWS.

We thought that REM rebound and not SWS rebound would be associated with CPAP compliance due to the simple observation that rebounding of REM with CPAP exposure is more common and certainly more impressive. When one considers the symptomatic consequences of sleep disruption, sleepiness is the most obvious, but not the only, outcome. In seminal studies carried out by Clemes and Dement, subjects deprived selectively of REM sleep for six nights showed changes in psychological testing suggesting a higher intensity of feeling and need [15]. The psychological effects of REM sleep deprivation in animals include a decreased ability to cope with stress with selective REM deprivation [16]. Alternatively, REM sleep deprivation does not seem to impact sleepiness, as one to two nights of selective REM sleep deprivation does not decrease mean sleep latency on multiple sleep latency testing [17]. That REM sleep plays a role in mood and its dysfunction is supported by the association between depression and a reduced REM sleep latency [18]. Furthermore, depression is seen with increasing frequency in persons suffering from OSA [19]. In cases of OSA where REM is decreased and subsequently rebounds with CPAP exposure, it could be an elevation in mood rather than alertness that drives compliance with CPAP. It is interesting that there was a trend toward an increase in being on an REM suppressant medication in the REM rebound group. This finding suggests either that REM sleep deprivation is associated with depression, precipitating antidepressant prescription, or that the REM-suppressing effect of the antidepressant medications work synergistically with sleep fragmentation to cause REM sleep deprivation. REM sleep is more easily aborted by sleep disruption than SWS. Early studies of selective sleep deprivation demonstrated that subjects needed five to seven times more frequent arousal to cause deprivation of SWS compared to REM sleep [20]. It is perhaps for this reason that REM sleep deprivation and REM sleep rebound is more common than SWS deprivation in the setting of sleep apnea. One would think that, since it is relatively more difficult to deprive SWS compared to REM, SWS deprivation would provoke more symptoms than REM sleep deprivation. At least on the measure of sleepiness, there was no evidence for this. In fact, there was a

868

B.B. Koo et al. / Sleep Medicine 13 (2012) 864–868

trend toward being less sleepy when there was SWS rebound. Additionally, the presence of SWS rebound was not significantly associated with greater CPAP compliance at any time period when comparing groups with and without SWS rebound or in multivariate analyses. In this and several other studies, OSA severity as measured by AHI was significantly associated with CPAP compliance [21–23]. It was interesting that this was true when groups with AHI < 30 were compared to those with an AHI between 30–60. Having an AHI > 60 conferred no additional benefit to CPAP compliance at any time period and, in fact, was worse at the 120 day period. One might expect that, if initial OSA severity were important in later CPAP compliance, the difference in compliance would be even greater for the extremely severe cases. This was not the case. Additionally, when looking at factors predicting CPAP compliance in multivariate modeling, AHI was not, and the presence of REM rebound was, an independent predictor of CPAP compliance. These finding suggest that it may not be the OSA severity that drives the greater CPAP compliance, but instead the presence of REM rebound which itself is correlated with AHI. Symptomatically, individuals have a wide range of response to untreated sleep apnea with some displaying severe excessive sleepiness and inability to maintain wakefulness even when driving a vehicle, and others demonstrating little in terms of daytime symptoms. This varied response to the disease suggests that the degree to which sleep is disrupted differs from person to person even in a narrow AHI range. Examining the degree of sleep disruption in untreated disease likely provides important information regarding this varied symptomatic response to OSA. Looking at the architectural sleep response to initial CPAP treatment may provide additional insight into this response. REM rebound with first CPAP exposure may identify a certain sleep architectural response to untreated sleep apnea or identify a common profile in persons that are susceptible to REM decrement in the setting of sleep apnea. Limitations of this study include the small sample size, which limited the power of the study to detect increases in CPAP compliance with REM rebound. Also, this sample consisted of predominantly men with severe sleep apnea, so the results may not be generalizable to women or patients with milder disease. The use of the split night study also deserves some mention here. Certainly, both homeostatic and circadian aspects differ for both REM and NREM sleep depending on when CPAP is initiated with pressure for SWS occurring early and for REM sleep occurring late. This difference could have confounded the results of the study and should be considered when interpreting the results. Future studies could include larger sample sizes, more women, and patients with mild to moderate obstructive sleep apnea, and they could include CPAP initiation in the split night and full night settings. The phenomenon of REM rebound is certainly worthy of additional research not only to replicate an association with CPAP compliance, but also to potentially gain insight into different architectural responses of sleep to recurrent apnea. Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: doi:10.1016/j.sleep.2012.03.019.

Acknowledgements The authors would like to acknowledge Kingman Strohl for his help with the preparation of this manuscript. References [1] Sullivan CE, Issa FG, Berthon-Jones M, Eves L. Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet 1981;1(8225):862–5. [2] Kribbs NB, Pack AI, Kline LR, Smith PL, Schwartz AR, Schubert NM, et al. Objective measurement of patterns of nasal CPAP use by patients with obstructive sleep apnea. Am Rev Respir Dis 1993;147(4):887–95. [3] Lindberg E, Berne C, Elmasry A, Hedner J, Janson C. CPAP treatment of a population-based sample–what are the benefits and the treatment compliance? Sleep Med 2006;7(7):553–60. [4] Kingshott RN, Vennelle M, Hoy CJ, Engleman HM, Deary IJ, Douglas NJ. Predictors of improvements in daytime function outcomes with CPAP therapy. AJRCCM 2000;161:866–71. [5] Engleman HM, Asgari-Jirhandeh N, McLeod AL, Ramsay CF, Deary IJ, Douglas NJ. Self-reported use of CPAP and benefits of CPAP therapy. Chest 1996;109:1470–6. [6] Edinger JD, Carwile S, Miller P, Mayti C. Psychological status, syndromatic measures, and compliance with nasal CPAP therapy for sleep apnoea. Percept Mot Skills 1994;78:1116–8. [7] Aldrich M, Eiser A, Lee M, Shipley J. Effects of continuous positive airway pressure on phasic events of REM sleep in patients with obstructive sleep apnea. Sleep 1989;12:413–9. [8] Issa FG, Sullivan CE. The immediate effects of nasal continuous positive airway pressure treatment on sleep pattern in patients with obstructive sleep apnea. Electroenceph Clin Neurophysiol 1986;63:10–7. [9] Verma A, Radtke RA, VanLandingham KE, King JH, Husain AM. Slow wave sleep rebound and REM rebound following the first night of treatment with CPAP for sleep apnea: correlation with subjective improvement in sleep quality. Sleep Med 2001;2(3):215–23. [10] Espinar Sierra J. Immediate effect of nasal continuous positive pressure in the treatment of obstructive sleep apnea. Rev Clin Esp 1989;184(2): 69–74. [11] Krieger J. Long-term compliance with CPAP therapy in obstructive sleep apnoea patients and in snorers. Sleep 1996;19:S136–43. [12] Meurice J-C, Dore P, Paquereau J, Neau J-P, Ingrand P, Chavagnat J-J, et al. Predictive factors of long- term compliance with nasal continuous positive airway pressure treatment in sleep apnoea syndrome. Chest 1994;105:429–33. [13] McArdle N, Devereux G, Heidarnejad H, Engleman HM, Mackay TW, Douglas NJ. Long-term use of CPAP therapy for sleep apnea/hypopnea syndrome. Am J Respir Crit Care Med 1999;159(4 Pt. 1):1108–14. [14] Barbé F, Mayoralas LR, Duran J, Masa JF, Maimó A, Montserrat JM, et al. Treatment with continuous positive airway pressure is not effective in patients with sleep apnea but no daytime sleepiness. A randomized, controlled trial. Ann Intern Med 2001;134(11):1015–23. [15] Clemes SR, Dement WC. Effect of REM sleep deprivation on psychological functioning. J Nerv Ment Dis 1967;144(6):485–91. [16] Fu J, Li P, Ouyang X, Gu C, Song Z, Gao J, et al. Rapid eye movement sleep deprivation selectively impairs recall of fear extinction in hippocampusindependent tasks in rats. Neuroscience 2007;144(4):1186–92. [17] Nykamp K, Rosenthal L, Folkerts M, Roehrs T, Guido P, Roth T. The effects of REM sleep deprivation on the level of sleepiness/alertness. Sleep 1998;21(6):609–14. [18] Kupfer DJ. REM latency: a psychobiologic marker for primary depressive disease. Biol Psychiatry 1976;11(2):159–74. [19] Reynolds 3rd CF, Coble PA, Spiker DG, Neil JF, Holzer BC, Kupfer DJ. Prevalence of sleep apnea and nocturnal myoclonus in major affective disorders: clinical and polysomnographic findings. J Nerv Ment Dis 1982;170(9):565–7. [20] Agnew Jr HW, Webb WB, Williams RL. Comparison of stage four and 1-rem sleep deprivation. Percept Mot Skills 1967 Jun;24(3):851–8. [21] Engleman HM, Martin SE, Douglas NJ. Compliance with CPAP therapy in patients with the sleep apnoea/hypopnoea syndrome. Thorax 1994;49: 263–6. [22] Krieger J. Long-term compliance with nasal continuous positive airway pressure (CPAP) in obstructive sleep apnea patients and nonapneic snorers. Sleep 1992;15:S42–46. [23] Popescu G, Latham M, Allgar V, Elliott MW. Continuous positive airway pressure for sleep apnoea/hypopnoea syndrome: usefulness of a 2 week trial to identify factors associated with long term use. Thorax 2001;56:727–33.