Abstract
Rationale: To establish a new approach to investigate the physiological effects of obstructive sleep apnea (OSA), and to evaluate novel treatments, during a period of continuous positive airway pressure (CPAP) withdrawal.
Objectives: To determine the effects of CPAP withdrawal.
Methods: Forty-one patients with OSA and receiving CPAP were randomized to either CPAP withdrawal (subtherapeutic CPAP), or continued CPAP, for 2 weeks. Polysomnography, sleepiness, psychomotor performance, endothelial function, blood pressure (BP), heart rate (HR), urinary catecholamines, blood markers of systemic inflammation, and metabolism were assessed.
Measurements and Main Results: CPAP withdrawal led to a recurrence of OSA within a few days and a return of subjective sleepiness, but was not associated with significant deterioration of psychomotor performance within 2 weeks. Endothelial function, assessed by flow-mediated dilatation, decreased significantly in the CPAP withdrawal group compared with therapeutic CPAP (mean difference in change, –3.2%; 95% confidence interval [CI], –4.5, –1.9%; P < 0.001). Compared with continuing CPAP, 2 weeks of CPAP withdrawal was associated with a significant increase in morning systolic BP (mean difference in change, +8.5 mm Hg; 95% CI, +1.7, +15.3 mm Hg; P = 0.016), morning diastolic BP (mean difference in change, +6.9 mm Hg; 95% CI, +1.9, +11.9 mm Hg; P = 0.008), and morning HR (mean difference in change, +6.3 bpm, 95% CI, +0.4, +12.2 bpm; P = 0.035). CPAP withdrawal was associated with an increase in urinary catecholamines but did not lead to an increase in markers of systemic inflammation, insulin resistance, or blood lipids.
Conclusions: CPAP withdrawal usually leads to a rapid recurrence of OSA, a return of subjective sleepiness, and is associated with impaired endothelial function, increased urinary catecholamines, blood pressure, and heart rate. Thus the proposed study model appears to be suitable to evaluate physiological and therapeutic effects in OSA.
Clinical trial registered with www.controlled-trials.com (ISRCTN 93153804).
Evaluating the physiological effects of obstructive sleep apnea (OSA) and continuous positive airway pressure (CPAP) by recruiting untreated patients in randomized controlled trials is time-consuming and expensive. A more efficient approach to investigate the physiological effects of OSA and to evaluate novel treatments could be during CPAP withdrawal, but the effects of CPAP withdrawal on sleep-disordered breathing, symptoms, and measures of cardiovascular risk need to be defined.
CPAP withdrawal leads to a rapid recurrence of OSA, a return of subjective sleepiness, and is associated with impaired endothelial function, increased urinary catecholamines, increased blood pressure, and heart rate. Thus the proposed study model appears to be suitable to evaluate physiological and therapeutic effects in OSA.
It has been estimated that between 2 and 4% of the adult population in Western countries are affected by moderate to severe obstructive sleep apnea syndrome (OSAS) (1). In patients with OSAS, excessive daytime sleepiness leads to impaired quality of life and is associated with an increased risk of traffic accidents (2, 3). OSAS has also been implicated as an important factor in the pathogenesis of hypertension and cardiovascular disease (4–8). Obstructive sleep apnea (OSA) can be treated effectively with continuous positive airway pressure (CPAP), which has been shown to improve daytime symptoms, psychomotor performance, quality of life, as well as measures of cardiovascular risk in randomized-controlled trials (2, 9).
Evaluating the consequences of OSAS, and any response to novel therapies, in conventional randomized-controlled trials is often compromised by the low recruitment rate of eligible study participants. In addition, the power of conventional randomized-controlled parallel trials of CPAP therapy is reduced by studying previously untreated patients with OSAS; such patients may have considerable difficulties becoming established on CPAP within the intervention period, thus diluting any treatment effect. A promising approach to investigate the consequences of OSAS, and to identify any response to novel therapies, is during a period of CPAP therapy withdrawal. This study model uses selectively chosen subjects from databases of patients who are already being treated successfully with CPAP, and in whom a maximal possible treatment effect can be assumed. The concept of studying patients already being treated may also have the advantage, compared with a conventional trial design, of eliminating potential secondary confounding long-term effects of OSAS, such as pharyngeal dilator muscle fatigue and pharyngeal mucosal edema (10), which may interfere with the response to other forms of OSAS treatment.
The basic concept of this model has been used in some uncontrolled studies investigating the effects of CPAP withdrawal on respiration during sleep and surrogate markers of cardiovascular risk (11–15). Because of limitations in the design and the short withdrawal time it is difficult to draw definitive conclusions from the findings of these studies. The current authors have addressed this uncertainty by exploring the effects of a 2-week withdrawal of CPAP on sleep-disordered breathing, symptoms, psychomotor performance, and various markers of vascular function and metabolic disease in a randomized-controlled trial.
Patients previously diagnosed with OSAS and treated with CPAP who were registered in a database of the Sleep Disorders Centre, University Hospital Zurich (Zurich, Switzerland) were eligible for the trial if they were between 20 and 75 years of age, had an oxygen desaturation index (ODI, ≥4% dips) greater than 10/hour in their initial sleep study, an ODI greater than 10/hour currently during an ambulatory nocturnal pulse-oximetry study performed on the last night of a four-night period without CPAP, and if they had been treated with CPAP for more than 12 months with an average compliance of at least 4 hours per night.
Patients with ventilatory failure, Cheyne-Stokes breathing, unstable and untreated coronary or peripheral artery disease, severe and inadequately controlled arterial hypertension, or a history of any sleep-related accident, or who were current professional drivers, were excluded from the study. The trial was approved by the University Hospital Zurich research ethics committee (EK-1600) and registered (ISRCTN 93,153,804). Written informed consent was obtained from participants.
After confirming the persistence of OSA by home overnight pulse oximetry at the end of a four-night period without CPAP, eligible patients returned to therapy with CPAP for at least 7 days. After baseline assessments patients were randomized to either continue with CPAP therapy or to switch to subtherapeutic CPAP for 2 weeks by a series of presealed and numbered envelopes. Follow-up assessments were performed at 2 weeks, endothelial function measurements were also performed at 1 week. Patients remained blinded concerning whether they were receiving therapeutic or subtherapeutic CPAP, as did the investigators. The member of the research team who randomized the patients did not take part in outcome assessments.
A sample size estimation was performed on the basis of the assumption that a clinically relevant difference in ODI (≥4% drops in oxygen saturation/h) between active treatment and placebo is 10/hour (SD, 10) and that a clinically relevant difference in endothelial function as assessed by flow-mediated dilatation of the brachial artery is 2% (SD, 2.3) (16, 17). On the basis of these assumptions, power calculation indicated that 41 patients were required in total so as not to miss a clinically relevant difference in the ODI, with a power of 90%, and in endothelial function, with a power of 80%.
Attended polysomnography was performed and analyzed according to standard methods (see the online supplement) (18). In addition, pulse oximetry was performed every night at home during the 2-week study period, using the S8 Res-Link device (ResMed Ltd, Basel, Switzerland) from which the ODI was derived and the apnea–hypopnea index (AHI) was downloaded from the CPAP device (S8; ResMed Ltd). In patients randomized to subtherapeutic CPAP, the subtherapeutic pressure was achieved by previously described methods (see the online supplement) (19, 20).
Subjective sleepiness was assessed by Epworth Sleepiness Scale (ESS) score at baseline and at 2 weeks (21, 22).
Objective sleepiness was measured in the evening with one sleep resistance challenge (Osler test) at baseline and at 2 weeks (23).
Divided attention was evaluated in the evening by the divided attention driving simulator (DADS) test during one 30-minute session at baseline and at 2 weeks (24). The deviation from the center line (expressed as standard deviation) was recorded and analyzed by dedicated software (Stowood Scientific Instruments, Oxford, UK).
Behavioral alertness was assessed in the evening, using the psychomotor vigilance task (PVT) at baseline and at 2 weeks (25). Reaction times were recorded and analyzed by dedicated software (Stowood Scientific Instruments).
Wrist actigraphy was performed during the 2-week study period to record sleep/wake patterns (Actiwatch; Cambridge Neurotechnology, Cambridge, UK) (26).
Flow-mediated dilatation (FMD) measurements were performed by ultrasound according to the method originally described by Celermajer (see the online supplement) (27). FMD measurements were performed at baseline and at 1 and 2 weeks.
During the 2-week study period, participants measured their blood pressure and heart rate in triplicate three times every day (morning, noon, and evening) with a standard digital automatic monitor (Omron Healthcare Co., Kyoto, Japan) in the sitting position after a period of rest of 5 minutes. The average blood pressure at each time of the day was used for further analysis.
Blood was drawn from all patients in the morning at baseline and at 2 weeks. Measurements of high-sensitivity C-reactive protein (hsCRP), IL-6, IL-10, and tumor necrosis factor (TNF)-α in plasma samples were performed as previously described (see the online supplement).
Blood samples were collected at baseline and at 2 weeks for determination of lipids and insulin resistance by homeostatic model assessment (HOMA) in the morning after an overnight fast as previously described (28).
Urine was collected for 12 hours from 7:00 p.m. to 7:00 a.m. at baseline and at 2 weeks. Urinary catecholamines were measured as previously described (see the online supplement) (29).
All values are presented as means (SD) unless otherwise stated. Statistical analyses were performed with software from STATA Corporation (College Station, TX) (version 11.1 for Windows). Differences in baseline characteristics between groups were assessed by independent t tests and χ2 tests as appropriate. Comparisons of changes between groups were assessed by independent t tests and Mann-Whitney U tests as appropriate. To further explore the time effect on changes in blood pressure and heart rate across the 14 days, multilevel regression analysis, stratified according to the allocated group, was performed, including the two levels of patient and time (i.e., Day 0 to Day 14), thus taking into account that observations in an individual patient are correlated (i.e., are not independent). To further investigate a group effect over time an additional interaction term (i.e., group × time) was modeled. A two-sided P value less than 0.05 was considered to be statistically significant.
Figure 1 shows the trial profile. Eleven patients (6.5%) who underwent the prestudy oximetry at the end of a 4-night CPAP withdrawal period did not have a return in ODI of more than 10, and thus were not included in the study. The two groups were similar regarding patient characteristics at baseline (Table 1). Data on AHI downloaded from the patients’ CPAP machine, as well as an ESS score within the normal range, proved successful treatment of patients with CPAP before the study (Table 1). One patient in the subtherapeutic CPAP group withdrew from the study 4 days after randomization because of intolerable daytime symptoms.
| Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 21) | P Value | |
| Age, yr | 63.6 (5.1) | 61.8 (7.5) | 0.393 |
| Males/females | 19/1 | 21/0 | 0.300 |
| Body mass index, kg/m2 | 32.9 (6.5) | 33.1 (4.4) | 0.904 |
| Waist/hip circumference ratio | 1.0 (0.1) | 1.0 (0.0) | 0.106 |
| Neck circumference, cm | 46.4 (3.7) | 46.1 (4.2) | 0.838 |
| Current smokers, % | 5.0 | 19.1 | 0.170 |
| Ex-smokers, % | 25.0 | 38.1 | 0.368 |
| Hypertension, % | 70.0 | 80.9 | 0.414 |
| Diabetes, % | 20.0 | 23.8 | 0.768 |
| CAD, % | 10.0 | 4.8 | 0.520 |
| Antihypertensive medication, % | 65.0 | 76.2 | 0.431 |
| Cholesterol-lowering medication, % | 35.0 | 33.3 | 0.910 |
| Glucose-lowering medication, % | 15.0 | 9.5 | 0.592 |
| AHI original sleep study | 36.0 (17.3) | 45.3 (22.3) | 0.155 |
| ODI original sleep study | 26.6 (13.5) | 37.3 (22.7) | 0.141 |
| ODI 4-d withdrawal | 25.4 (8.6) | 28.9 (16.2) | 0.401 |
| AHI on CPAP* | 5.1 (2.7) | 4.3 (2.3) | 0.329 |
| CPAP compliance, min | 373.1 (67.9) | 362.8 (72.3) | 0.642 |
| CPAP use relative to time in bed, %† | 80.8 (14.5) | 79.2 (20.0) | 0.774 |
| ESS score before therapy | 13.8 (2.6) | 15.3 (3.5) | 0.188 |
| ESS score on CPAP | 7.4 (3.1) | 6.6 (2.7) | 0.399 |
| Time from screening to randomization, d | 75.9 (29.0) | 68.3 (29.9) | 0.410 |
Withdrawal of CPAP was associated with a significant increase in AHI, ODI, and the number of arousals at 2 weeks, compared with continuation of CPAP (Table 2). Because of technical reasons there were no data on polysomnography in one patient at 2 weeks.
| Baseline | Change from Baseline at 2 wk | |||||
| Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 21) | Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 20) | 95% CI between Groups | P Value | |
| Total sleep time, min | 304 (64) | 306 (82) | +27.7 (98.7) | +20.3 (116.2) | −77.2/+62.4 | 0.833 |
| % stage 1 | 25.5 (11.2) | 26.6 (13.4) | +9.3 (11.1) | +12.8 (19.8) | −6.8/+13.8 | 0.489 |
| % stage 2 | 48.4 (16.0) | 43.2 (15.9) | +1.1 (14.4) | +13.0 (44.3) | −9.3/+33.0 | 0.265 |
| % stage 3 | 7.7 (7.4) | 13.4 (11.6) | +0.6 (4.4) | −2.7 (8.4) | −7.6/+1.0 | 0.125 |
| % stage 4 | 0.6 (1.2) | 2.8 (4.6) | −0.1 (0.8) | −0.4 (2.9) | −1.7/+1.1 | 0.630 |
| % REM | 17.8 (10.1) | 14.0 (12.0) | +3.3 (9.7) | +0.3 (11.6) | −9.9/+3.9 | 0.396 |
| Arousals, events/h | 7.3 (3.4) | 8.5 (7.5) | −0.6 (3.6) | +20.7 (23.9) | +10.2/+32.2 | <0.001 |
| AHI, events/h | 1.7 (1.8) | 2.2 (2.5) | +0.4 (2.8) | +33.8 (24.3) | +22.3/+44.5 | <0.001 |
| ODI, events/h | 0.5 (0.8) | 0.9 (2.0) | −0.2 (1.0) | +26.3 (22.9) | +16.1/+36.9 | <0.001 |
| ESS score | 7.4 (3.1) | 6.6 (2.7) | −0.7 (2.2) | +2.0 (2.7) | +1.1/+4.3 | 0.001 |
| Osler, s | 2,100 (577) | 2,331 (229) | −42.9 (330.8) | −213.3 (711.7) | −544.0/+167.0 | 0.290 |
| PVT, ms | 237 (41) | 238 (54) | −9.8 (29.6) | +25.0 (172.5) | −44.4/+114.0 | 0.380 |
| DADS | 0.38 (0.21) | 0.32 (0.09) | −0.1 (0.2) | 0.0 (0.1) | −0.004/+0.204 | 0.200 |
Withdrawal of CPAP was associated with a rapid return of sleep-disordered breathing within a few days (Figure 2). Mean CPAP compliance at follow-up was 381.8 (80.4) minutes in the therapeutic and 280.2 (190.8) minutes in the subtherapeutic CPAP group.
Figure 2. Data from ambulatory pulse-oximetry and continuous positive airway pressure (CPAP) machine download showing the mean (±SE) nightly apnea–hypopnea index (AHI) (top) and oxygen desaturation index (ODI) (bottom) of patients withdrawn from CPAP (Placebo-CPAP, solid circles) and of patients continuing CPAP (open circles). Withdrawal of CPAP was associated with a rapid return of sleep-disordered breathing within the first nights off CPAP. (P < 0.001 for all comparisons between groups during CPAP withdrawal.)
[More] [Minimize]Withdrawal of CPAP was not associated with a change in mean activity, time spent in bed, or estimated sleep time, at either time point during the intervention period (P > 0.05 for all comparisons between groups).
Subjective sleepiness as assessed by ESS score increased significantly at 2 weeks, in the subtherapeutic CPAP group compared with the therapeutic CPAP group (Table 2).
Objective sleepiness as assessed by sleep resistance time in the Osler test did not increase significantly after 2 weeks, in the subtherapeutic CPAP group compared with the therapeutic CPAP group (Table 2).
Withdrawal of CPAP was not associated with a significant deterioration in performance on the DADS test or in behavioral alertness as assessed by the PVT at 2 weeks when compared with the therapeutic CPAP group (Table 2).
Endothelial function as assessed by FMD decreased significantly at 1 and 2 weeks in the CPAP withdrawal group compared with the therapeutic CPAP group (mean difference in FMD change, –1.7% [95% CI, –2.8 to –0.6%] and –3.2% [95% CI, –4.5 to –1.9%]; P = 0.002 and P < 0.001, respectively; Figure 3). Endothelial independent vasodilation induced by glycerol trinitrate did not change significantly with CPAP withdrawal at 1 and 2 weeks when compared with the therapeutic CPAP group (mean difference, –1.6% [95% CI, –4.8 to +1.6%] and –2.3% [95% CI, –4.9 to +0.3%]; P = 0.296 and P = 0.080, respectively).
Figure 3. Endothelial function as assessed by flow-mediated dilatation of the brachial artery progressively decreased in the continuous positive airway pressure (CPAP) withdrawal group (solid columns) whereas there was no significant change in endothelial function in patients who continued with CPAP therapy (open columns), thus resulting in significant differences in changes over time. Data are presented as means and standard deviations. *P = 0.002 and #P < 0.001 for comparisons of changes between groups at Week 1 and at Week 2, respectively.
[More] [Minimize]Withdrawal of CPAP was associated with a significant increase in domiciliary morning systolic and diastolic blood pressure, as well as heart rate, at 2 weeks when compared with the therapeutic CPAP group (Table 3 and Figure 4).
| Baseline | Change from Baseline at 2 wk | |||||
| Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 21) | Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 20) | 95% CI between Groups | P Value | |
| BP sys morning, mm Hg | 133.3 (16.6) | 129.2 (12.7) | −2.3 (9.8) | +6.2 (11.1) | +1.7/+15.3 | 0.016 |
| BP dia morning, mm Hg | 82.3 (7.8) | 82.2 (8.2) | −2.5 (7.7) | +4.4 (7.8) | +1.9/+11.9 | 0.008 |
| HR morning, bpm | 61.8 (9.2) | 65.0 (14.1) | 0.0 (7.8) | +6.3 (10.2) | +0.4/+12.2 | 0.035 |
| BP sys noon, mm Hg | 132.1 (9.1) | 127.8 (11.9) | −2.5 (10.1) | +3.6 (13.3) | −2.1/+14.3 | 0.143 |
| BP dia noon, mm Hg | 79.8 (6.9) | 80.4 (8.6) | +0.5 (9.1) | +0.3 (8.9) | −6.3/+5.9 | 0.944 |
| HR noon, bpm | 64.4 (10.3) | 69.7 (14.4) | +2.8 (9.6) | +5.5 (8.3) | −3.4/+8.8 | 0.375 |
| BP sys evening, mm Hg | 131.6 (16.7) | 130.4 (12.7) | −4.6 (13.1) | +5.0 (8.7) | +2.3/+16.9 | 0.012 |
| BP dia evening, mm Hg | 78.2 (8.6) | 81.6 (7.7) | −4.0 (8.2) | +2.0 (7.3) | +0.9/+11.1 | 0.022 |
| HR evening, bpm | 67.4 (11.1) | 72.8 (17.9) | +0.9 (14.8) | +0.4 (9.5) | −8.7/+7.7 | 0.362 |
| Norepinephrine, nmol/mmol | 22.1 (11.5) | 26.3 (10.2) | +0.9 (5.4) | +11.5 (15.5) | +2.5/+18.7 | 0.012 |
| Epinephrine, nmol/mmol | 2.7 (1.6) | 3.3 (3.1) | +0.1 (3.4) | +2.4 (13.0) | −4.3/+8.9 | 0.480 |
Figure 4. (A) Daily ambulatory systolic and diastolic blood pressures measured in the mornings in the continuous positive airway pressure (CPAP) withdrawal group (solid circles/squares) and in the group that continued CPAP therapy (open circles/squares). There was a significant time effect on diastolic blood pressure in the withdrawal group (β = 0.21, P = 0.003) but not in the CPAP group (β = –0.05, P = 0.46). The additional modeling of an interaction term (group × time) was highly significant (β = 0.27, P = 0.008), indicating an increasing difference in diastolic blood pressure over time between groups. Although there was no statistically significant time effect on systolic blood pressure within groups, the interaction term (group × time) indicated a trend toward higher systolic blood pressure over time in the withdrawal group (β = 0.24, P = 0.10). (B) Daily ambulatory heart rate measured in the mornings in the CPAP withdrawal group (solid circles) and in the group which continued therapeutic CPAP therapy (open circles). Heart rate increased significantly over time in the withdrawal group (β = 0.40, P < 0.001) whereas no change has been observed in the CPAP group (β = –0.05, P = 0.6). The interaction term (group × time) was highly significant in the withdrawal group (β = 0.45, P < 0.001), indicating an increasing difference in heart rate over time between groups.
[More] [Minimize]Withdrawal of CPAP for 2 weeks was associated with a statistically significant increase in overnight urine norepinephrine, but not in epinephrine, compared with the therapeutic CPAP group (Table 3).
None of the measures of systemic inflammation, insulin resistance, or cholesterol changed significantly after 2 weeks of CPAP withdrawal compared with the CPAP group (Table 4). CPAP withdrawal was associated with a significant decrease in blood triglycerides at 2 weeks when compared with the therapeutic CPAP group (Table 4).
| Baseline | Change from Baseline at 2 wk | |||||
| Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 21) | Therapeutic CPAP (n = 20) | Subtherapeutic CPAP (n = 18) | 95% CI between Groups | P Value | |
| Glucose, mmol/L | 5.9 (5.4,7.3) | 6.2 (5.5,6.8) | −0.1 (−0.4,0.2) | 0.0 (−0.1,0.3) | −0.2/+0.5 | 0.373 |
| Insulin, mlU/L | 11.0 (4.1,17.6) | 11.6 (5.4,17.3) | +0.1 (−2.3,4.0) | +0.5 (−7.4,2.6) | −5.7/+2.8 | 0.838 |
| HOMA, mIU/L * mmol/L | 3.5 (1.1,4.7) | 3.1 (1.4,4.9) | +0.2 (−1.0,1.7) | +0.2 (−2.4,0.9) | −1.8/+0.8 | 0.579 |
| Cholesterol, mmol/L | 4.8 (4.1,5.7) | 5.1 (4.3,5.8) | 0.0 (−0.2,0.2) | −0.2 (−0.5,0.5) | −0.4/+0.2 | 0.511 |
| HDL, mmol/L | 1.3 (1.0,1.4) | 1.1 (0.9,1.3) | 0.0 (0.0,0.1) | 0.0 (0.0,0.1) | −0.1/+0.1 | 0.942 |
| LDL, mmol/L | 3.1 (2.4,3.6) | 3.2 (2.4,3.9) | −0.1 (−0.2,0.1) | −0.2 (−0.3, 0.3) | −0.2/+0.3 | 0.930 |
| Triglycerides, mmol/L | 1.2 (1.0,1.5) | 1.7 (1.2,2.3) | +0.2 (0.0,0.4) | −0.2 (−0.5,−0.2) | −0.7/−0.2 | 0.013 |
| hsCRP, mg/ml | 1.2 (0.7,2.4) | 2.4 (1.0,4.5) | −0.1 (−0.3,0.6) | −0.3 (−1.0,0.1) | −1.7/+0.1 | 0.077 |
| IL−6, pg/ml | 0.5 (0.3,1.0) | 0.3 (0.3,0.8) | 0.0 (−0.2,0.2) | 0.0 (−0.2,0.1) | −0.2/+0.2 | 0.838 |
| IL−10, pg/ml | 0.8 (0.5,1.3) | 0.6 (0.5,1.1) | 0.0 (−0.2,0.1) | 0.0 (−0.1,0.2) | −0.1/+0.2 | 0.413 |
| TNF−α, pg/ml | 0.7 (0.5,0.7) | 0.6 (0.5,0.7) | 0.0 (−0.1,0.1) | 0.0 (−0.1,0.1) | −0.1/+0.1 | 0.661 |
This is the first randomized-controlled study investigating the effects of a 2-week CPAP withdrawal on OSA severity, sleepiness, psychomotor performance, and measures of cardiovascular risk. We found that CPAP withdrawal was associated with a return of OSA by the first night, an increase in subjective sleepiness, progressively impaired endothelial function, increased urinary catecholamines, increased blood pressure, as well as increased heart rate at 2 weeks. CPAP withdrawal was not associated with significant deterioration in psychomotor performance, systemic inflammation, insulin resistance, or blood lipids within 2 weeks. Because of the observed rapid return of OSA, sleepiness, and deterioration in well-established measures of cardiovascular risk, the proposed study model appears to be suitable to evaluate treatment effects during a period of CPAP withdrawal.
Withdrawal of CPAP was associated with a rapid return of OSA, with statistical significance already achieved on the first night off CPAP (Figure 2), which is in agreement with previously published data from uncontrolled studies (30–33). The severity of OSA did not increase further after the first week of CPAP withdrawal. This suggests that the duration of any future trial using this model, and purely investigating the treatment effects on the severity of sleep-disordered breathing, could be as short as a few nights; a sample size calculation based on our findings indicates that, for instance, 36 patients are required in total to exclude a difference of 15 in ODI between any treatment and placebo on the first night off CPAP (α, 0.05; 90% power).
With CPAP withdrawal in this study, subjective sleepiness (ESS score) increased statistically significantly by 2.7 units at 2 weeks. This is in agreement with previously published data from uncontrolled studies (30–33). The relatively slow increase in ESS score may be explained by the fact that this questionnaire was not designed to assess changes in sleepiness over such a short time scale, but rather focuses on the recent past.
Interestingly, in the current trial, CPAP withdrawal for 2 weeks was not associated with a significant increase in objective sleepiness as measured by the Osler test, or with deterioration in performance as assessed by the driving simulator test (DADS) and psychomotor vigilance task (PVT). This finding confirms the results of previously published uncontrolled studies, which also found no significant change in PVT after one night and no significant deterioration in driving performance after 1 week of CPAP withdrawal (30, 33). However, our results on objective sleepiness are in contrast to the findings of an uncontrolled study reporting increased sleepiness, using the Osler test, after 1 week of CPAP withdrawal (32). Possible explanations for this discrepancy may be differences regarding the characteristics of the study populations and different times of the day that the tests were done. However, as we did not find a statistically significant deterioration in the DADS, we are confident that the use of the proposed study model is not limited by issues regarding exposure of patients to a high risk of road accidents. In fact, this may be an additional advantage of the withdrawal study model, as the intervention period in conventional randomized placebo-controlled studies evaluating treatment effects is usually longer and patients are sleepier, and thus may be at a higher cumulative risk for road accidents during any placebo treatment period. Despite this, it should be mentioned that patients taking part in randomized-controlled studies investigating the effects of OSA treatments should be discouraged from driving.
Return of OSA was associated with progressively impaired endothelial function, which reached statistical significance at 1 week in the current trial and deteriorated further by the second week of CPAP withdrawal. This finding strengthens the evidence from previous studies (13, 34, 35) that OSA causes endothelial dysfunction independently from obesity.
With CPAP withdrawal in this study, domiciliary systolic and diastolic blood pressure measured in the morning increased statistically significantly by 8.5 and 6.9 mm Hg, respectively, at 2 weeks. This rise in morning blood pressure was accompanied by a marked increase in heart rate (6 bpm) and increased urinary catecholamine excretion, suggesting augmented sympathetic activity as the underlying cause. However, the results of the current study show that the effects of CPAP withdrawal on blood pressure and heart rate were less pronounced at noon and in the evening (Table 3), which was corroborated by the lack of a time effect in the multilevel regression analysis (data not shown) and may be due to the stronger influence of confounding factors during daytime, such as the patients’ lifestyle. Our findings are in concordance with the report of Phillips and colleagues (15), who found a trend toward higher morning blood pressure after 1 week of CPAP withdrawal in an uncontrolled study.
There is some evidence from in vitro and animal model studies, as well as from controlled studies in humans, suggesting that systemic inflammation may be an important mechanism linking OSA with endothelial dysfunction and vascular disease, but there are sparse data from randomized-controlled trials on this postulated association (4, 28, 36, 37). In the current trial, 2 weeks of CPAP withdrawal was not associated with an increase in hsCRP and proinflammatory cytokines. The 95% confidence intervals of the difference in change of hsCRP, TNF-α, and IL-6 indicate that we have excluded increases as small as 0.1 mg/ml, 0.1 pg/ml, and 0.2 pg/ml, respectively, which are well below the reported differences in uncontrolled studies looking at the effect of CPAP on these inflammatory markers (38, 39). Our findings confirm the results of an uncontrolled study in which no significant increases in hsCRP, IL-6, and TNF-α after 1 week of CPAP withdrawal were found (14). Possible methodological explanations for this negative finding are that 2 weeks of CPAP withdrawal may not be long enough to induce systemic inflammation or that the inclusion of patients with cardiovascular comorbidities might have masked an effect on systemic inflammation. However, as CPAP withdrawal was associated with clinically relevant changes in endothelial function, urinary catecholamines, blood pressure, and heart rate in the current study, it may be assumed that sympathetic activation is a more likely mechanism of vascular damage in typical patients with OSA with comorbidities than systemic inflammation, at least in the short term.
OSA has also been implicated as a causal factor of insulin resistance and metabolic disease, but there are still few data from randomized-controlled trials evaluating the impact of CPAP therapy on glucose and lipid metabolism (40). In this study we found no statistically significant changes in glucose metabolism as assessed by HOMA or in blood cholesterol levels at the end of the 2 weeks of CPAP withdrawal. A possible methodological explanation for this negative finding may be that we did not specifically include patients with abnormal glucose or lipid metabolism, and thus the proportion of patients with diabetes mellitus was quite small (approximately 20%).
The CPAP withdrawal model seems to be a useful addition to the armamentarium of study paradigms investigating the physiological consequences of OSA such as the introduction of “simulated OSA” in healthy humans. Using the latter paradigm in humans, it is extremely difficult to simulate all physiological features of OSA (e.g., intermittent hypoxia, arousals, and pleural pressure swings) at the same time and thus it has been used mainly to simulate intermittent hypoxia (41–43), whereas withdrawal of CPAP therapy reintroduces the full disease with all its characteristics. On the other hand, by selectively simulating intermittent hypoxia it is possible to explore specific physiological consequences of intermittent hypoxia within a few hours or days, which is clearly more difficult using the CPAP withdrawal model. Thus the CPAP withdrawal model seems to be suitable when the short-term physiological effects of OSA and responses to novel treatments need to be investigated rather than when specific physiological effects of one feature of OSA are explored.
The CPAP withdrawal model has limitations; some patients randomized to subtherapeutic CPAP may become aware of the group allocation, which may influence subjective outcomes. Using this approach in its current form it is of course not possible to clarify whether there is a causal relationship between OSA and any proposed long-term consequences, for example, cardiovascular events and increased mortality. Because of ethical issues it may be difficult to prolong the intervention time to more than 2 weeks, and thus the CPAP withdrawal model is suitable to assess only short-term effects of OSA and responses to novel treatments. In addition, the findings of a study that selects patients with high CPAP compliance cannot be generalized without caution to the untreated OSA population usually seen in a sleep clinic.
In conclusion, we have shown that CPAP withdrawal leads to a return of OSAS within the first night off CPAP and is associated with an increase in subjective sleepiness. Two weeks of CPAP withdrawal is also associated with impairment of endothelial function, as well as an increase in urinary catecholamines, blood pressure, and heart rate. Therefore, the proposed study model, using CPAP withdrawal as the intervention in patients with OSA, seems to be a useful protocol to evaluate both short-term physiological effects of OSA and responses to novel treatments.
| 1. | Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med 2002;165:1217–1239. |
| 2. | Siccoli MM, Pepperell JC, Kohler M, Craig SE, Davies RJ, Stradling JR. Effects of continuous positive airway pressure on quality of life in patients with moderate to severe obstructive sleep apnea: data from a randomized controlled trial. Sleep 2008;31:1551–1558. |
| 3. | George CF. Reduction in motor vehicle collisions following treatment of sleep apnoea with nasal CPAP. Thorax 2001;56:508–512. |
| 4. | Kohler M, Stradling JR. Mechanisms of vascular damage in obstructive sleep apnea. Nat Rev Cardiol 2010;7:677–685. |
| 5. | Haentjens P, Van Meerhaeghe A, Moscariello A, De Weerdt S, Poppe K, Dupont A, Velkeniers B. The impact of continuous positive airway pressure on blood pressure in patients with obstructive sleep apnea syndrome: evidence from a meta-analysis of placebo-controlled randomized trials. Arch Intern Med 2007;167:757–764. |
| 6. | Marin JM, Carrizo SJ, Vicente E, Agusti AG. Long-term cardiovascular outcomes in men with obstructive sleep apnoea–hypopnoea with or without treatment with continuous positive airway pressure: an observational study. Lancet 2005;365:1046–1053. |
| 7. | Redline S, Yenokyan G, Gottlieb DJ, Shahar E, O'Connor GT, Resnick HE, Diener-West M, Sanders MH, Wolf PA, Geraghty EM, et al.. Obstructive sleep apnea–hypopnea and incident stroke. Am J Respir Crit Care Med 2010;182:269–277. |
| 8. | Gottlieb DJ, Yenokyan G, Newman AB, O'Connor GT, Punjabi NM, Quan SF, Redline S, Resnick HE, Tong EK, Diener-West M, et al.. Prospective study of obstructive sleep apnea and incident coronary heart disease and heart failure. Circulation 2010;122:352–360. |
| 9. | Kohler M, Pepperell JCT, Casadei B, Craig S, Crosthwaite N, Stradling JR, Davies RJO. CPAP and measures of cardiovascular risk in males with OSAS. Eur Respir J 2008;32:1488–1496. |
| 10. | Ryan CF, Lowe AA, Li D, Fleetham JA. Magnetic resonance imaging of the upper airway in obstructive sleep apnea before and after chronic nasal continuous positive airway pressure therapy. Am Rev Respir Dis 1991;144:939–944. |
| 11. | Grunstein RR, Stewart DA, Lloyd H, Akinci M, Cheng N, Sullivan CE. Acute withdrawal of nasal CPAP in obstructive sleep apnea does not cause a rise in stress hormones. Sleep 1996;19:774–782. |
| 12. | Bonsignore MR, Parati G, Insalaco G, Marrone O, Castiglioni P, Romano S, Di Rienzo M, Mancia G, Bonsignore G. Continuous positive airway pressure treatment improves baroreflex control of heart rate during sleep in severe obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2002;166:279–286. |
| 13. | Ip MSM, Tse HF, Lam B, Tsang KWT, Lam WK. Endothelial function in obstructive sleep apnea and response to treatment. Am J Respir Crit Care Med 2004;169:348–353. |
| 14. | Phillips CL, Yang Q, Williams A, Roth M, Yee BJ, Hedner JA, Berend N, Grunstein RR. The effect of short-term withdrawal from continuous positive airway pressure therapy on sympathetic activity and markers of vascular inflammation in subjects with obstructive sleep apnoea. J Sleep Res 2007;16:217–225. |
| 15. | Phillips CL, Yee B, Yang Q, Villaneuva AT, Hedner J, Berend N, Grunstein R. Effects of continuous positive airway pressure treatment and withdrawal in patients with obstructive sleep apnea on arterial stiffness and central BP. Chest 2008;134:94–100. |
| 16. | Gokce N, Keany JF, Hunter LM, Watkins MT, Menzoian MO, Vita JA. Risk stratification for postoperative cardiovascular events via noninvasive assessment of endothelial function: a prospective study. Circulation 2002;105:1567–1572. |
| 17. | Kohler M, Craig SE, Pepperell JCT, Nicoll D, Bratton D, Nunn A, Davies RJO, Stradling JR. CPAP improves endothelial function in minimally symptomatic OSA patients: results from the MOSAIC trial [abstract]. Thorax 2010;65:A9–A10. |
| 18. | Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. NIMH Publication 204. Washington, DC: U.S. Government Printing Office; 1968. |
| 19. | Jenkinson C, Davies RJ, Mullins R, Stradling JR. Comparison of therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnoea: a randomised prospective parallel trial. Lancet 1999;353:2100–2105. |
| 20. | Pepperell JCT, Ramdassingh-Dow S, Crosthwaite N, Mullins R, Jenkinson C, Stradling JR, Davies RJO. Ambulatory blood pressure after therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnoea: a randomised parallel trial. Lancet 2002;359:204–210. |
| 21. | Bloch KE, Schoch OD, Zhang JN, Russi EW. German version of the Epworth Sleepiness Scale. Respiration 1999;66:440–447. |
| 22. | Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540–545. |
| 23. | Bennett LS, Stradling JR, Davies RJ. A behavioural test to assess daytime sleepiness in obstructive sleep apnoea. J Sleep Res 1997;6:142–145. |
| 24. | Juniper M, Hack MA, George CF, Davies RJ, Stradling JR. Steering simulation performance in patients with obstructive sleep apnoea and matched control subjects. Eur Respir J 2000;15:590–595. |
| 25. | Dinges DF, Powell JW. Microcomputer analyses of performance on a portable, simple visual RT task during sustained operations. Behav Res Methods Instrum Comput 1985;17:652–655. |
| 26. | Cole RJ, Kripke DF, Gruen W, Mullaney DJ, Gillin JC. Automatic sleep/wake identification from wrist actigraphy. Sleep 1992;15:461–469. |
| 27. | Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI, Sullivan ID, Lloyd JK, Deanfield JE. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340:1111–1115. |
| 28. | West SD, Nicoll DJ, Wallace TM, Matthews DR, Stradling JR. The effect of CPAP on insulin resistance and HbA1c in men with obstructive sleep apnoea and type 2 diabetes. Thorax 2007;62:969–974. |
| 29. | Peaston RT. Routine determination of urinary catecholamines by high performance liquid chromatography with electrochemical detection. J Chromatogr A 1988;424:263–272. |
| 30. | Kribbs NB, Pack AI, Kline LR, Getsy JE, Schuett JS, Henry JN, Maislin G, Dinges DF. Effects of one night without nasal CPAP treatment on sleep and sleepiness in patients with obstructive sleep apnea. Am Rev Respir Dis 1993;147:1162–1168. |
| 31. | Sforza E, Lugaresi E. Daytime sleepiness and nasal continuous positive airway pressure therapy in obstructive sleep apnea syndrome patients: effects of chronic treatment and 1-night therapy withdrawal. Sleep 1995;18:195–201. |
| 32. | Yang Q, Phillips CL, Melehan KL, Rogers NL, Seale JP, Grunstein RR. Effects of short-term CPAP withdrawal on neurobehavioral performance in patients with obstructive sleep apnea. Sleep 2006;29:545–552. |
| 33. | Turkington PM, Sircar M, Saralaya D, Elliott MW. Time course of changes in driving simulator performance with and without treatment in patients with sleep apnoea hypopnoea syndrome. Thorax 2004;59:56–59. |
| 34. | Cross MD, Mills NL, Al-Abri M, Riha R, Vennelle M, Mackay TW, Newby DE, Douglas NJ. Continuous positive airway pressure improves vascular function in obstructive sleep apnoea/hypopnoea syndrome: a randomised controlled trial. Thorax 2008;63:578–583. |
| 35. | Alonso-Fernandez A, Garcia-Rio F, Arias MA, Hernanz A, De la Pena M, Pierola J, Barcelo A, Lopez-Collazo E, Agusti A. Effects of CPAP on oxidative stress and nitrate efficiency in sleep apnoea: a randomised trial. Thorax 2009;64:581–586. |
| 36. | Drager LF, Bortolotto LA, Figueiredo AC, Krieger EM, Lorenzi-Filho G. Effects of continuous positive airway pressure on early signs of atherosclerosis in obstructive sleep apnea. Am J Respir Crit Care Med 2007;176:706–712. |
| 37. | Kohler M, Ayers L, Pepperell JCT, Packwood KL, Ferry B, Crosthwaite N, Craig S, Siccoli MM, Davies RJO, Stradling JR. Effects of continuous positive airway pressure on systemic inflammation in patients with moderate to severe obstructive sleep apnoea: a randomised controlled trial. Thorax 2009;64:67–73. |
| 38. | Yokoe T, Minoguchi K, Matsuo H, Oda N, Minoguchi H, Yoshino G, Hirano T, Adachi M. Elevated levels of C-reactive protein and interleukin-6 in patients with obstructive sleep apnea syndrome are decreased by nasal continuous positive airway pressure. Circulation 2003;107:1129–1134. |
| 39. | Ryan S, Taylor CT, McNicholas WT. Selective activation of inflammatory pathways by intermittent hypoxia in obstructive sleep apnea syndrome. Circulation 2005;112:2660–2667. |
| 40. | Levy P, Bonsignore MR, Eckel J. Sleep, sleep-disordered breathing and metabolic consequences. Eur Respir J 2009;34:243–260. |
| 41. | Pialoux V, Hanly PJ, Foster GE, Brugniaux JV, Beaudin AE, Hartmann SE, Pun M, Duggan CT, Poulin MJ. Effects of exposure to intermittent hypoxia on oxidative stress and acute hypoxic ventilatory response in humans. Am J Respir Crit Care Med 2009;180:1002–1009. |
| 42. | Tamisier R, Gilmartin GS, Launois SH, Pepin JL, Nespoulet H, Thomas R, Levy P, Weiss JW. A new model of chonic intermittent hypoxia in humans: effect on ventilation, sleep, and blood pressure. J Appl Physiol 2009;107:17–24. |
| 43. | Louis M, Punjabi NM. Effects of acute intermittent hypoxia on glucose metabolism in awake healthy volunteers. J Appl Physiol 2009;106:1538–1544. |
Supported by a grant from the Swiss National Science Foundation (32003B_124915) and the Swiss Society of Pneumology.
Author Contributions: Design of the study—M.K., J.R.S.; acquisition of data—M.K., A.C.S., L.A.; data analysis and interpretation—M.K., A.C.S., L.A., O.S., K.E.B., E.W.R., J.R.S.; manuscript draft and revision for intellectual content and approved the final version—M.K., A.C.S., L.A., O.S., K.E.B., E.W.R., J.R.S.
This article has an online supplement, which is available from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201106-0964OC on August 11, 2011
Author disclosures
