Continuous Positive Airway Pressure in Severe Obstructive Sleep Apnea Reduces Pain Sensitivity

Study Objective: To evaluate effects of CPAP on pain sensitivity in severe OSA model of thermal nociception. Baseline FWL in seconds was obtained after the training session. CPAP pressure was titrated on a second night in the laboratory. Two nights after titration, patients returned to sleep in the laboratory on CPAP. FWL was tested in the morning after awakening, after 6-8 wks of CPAP use, and finally (within 6-8 weeks) after 2 nights of discontinuation of CPAP. Mean FWL in seconds (sec) was compared using MANOVAs with nights as the within subject variable. Results: Apnea-hypopnea index (AHI) decreased from 50.9 ± 14.5 to 1.4 ± 1.0 with CPAP, and sleep continuity improved. In parallel, FWL increased signifi cantly from a mean baseline of 9.8 ± 1.3 sec to 13.7 ± 5.1 sec (P = 0.01) and with continued CPAP use (5.1 ± 2.3 h nightly) for 6-8 weeks FWL remained elevated (21.1 ± 16.2 sec). After the 2-night CPAP discontinuation, apnea/hypopneas returned and sleep was fragmented (AHI = 32.6 ± 19.8). FWL decreased to 11.6 ± 5.9 sec relative to intermediate-term CPAP use (P = 0.03). Conclusion: CPAP treatment reduces pain sensitivity in OSA patients. Future studies will focus on patients with OSA and chronic pain and identify mediating mechanisms.


INTRODUCTION
Excessive daytime sleepiness is typically caused by a primary sleep disorder or inadequate sleep, whether due to insufficient time in bed or poor sleep quality. The poor sleep quality can be due to environmental and social factors, psychological or pathological disorders, or a medication/drug effect. Excessive sleepiness results in reduced quality of life, psychomotor impairment, adverse mood effects, accidents, decreased work productivity, and deficits in memory and decision making. [1][2][3][4] More recently, hyperalgesia is being recognized as a consequence of inadequate sleep and its resulting daytime sleepiness. 5 It is a well-known fact that pain associated with both acute conditions and chronic medical disorders like rheumatoid arthritis, fibromyalgia, and cystic fibrosis causes significant sleep disturbance. [6][7][8] On the other hand, the sleep disturbances in these disorders are also known to enhance pain perception. [6][7][8] So, the relationship between sleep and pain is bidirectional.
The effects of decreased sleep quantity on pain sensitivity, whether total sleep deprivation, sleep restriction or selective sleep stage deprivation, have been evaluated in terms of effects on pain. Hyperalgesia was found after 40 hours of total sleep to familiarize them with the pain testing equipment and procedures. Following the training, the threshold at which a finger withdrawal response was elicited for each individual participant was established. The threshold radiant heat intensity was defined as that intensity which produced a finger withdrawal in < 21 sec. This threshold and 4 incrementally greater intensities were then used throughout the experiment. These 5 different heat intensities were randomly presented: the threshold intensity and the 4 greater intensities as determined by units on a potentiometer controlling current to the radiant heat source. The test day pain assessments were done only after the training pain assessments were carried out and a given participant's basal pain threshold established.
In a FWLT session each participant was seated in a chair at a desk across from a research assistant who administered the pain testing procedures. The participant's finger was placed on top of a metal box which contained the heat source, and the pad of the index finger was centered over a 3-mm hole through which the heat radiated. The heat source used was a 100 watt projection bulb, located at a fixed distance from the subject's finger. A potentiometer controlled current delivered to the light bulb thereby varying potential heat intensity. The temperature measured after 10 sec for the lowest intensity used in the study was 83.4°F, and the highest was 101.6°F. The control device used to adjust the radiant heat intensities was operated by the research assistant and was hidden from the participant's view.
On each trial, participants were instructed to place the index finger on the hole through which the heat source radiates and withdraw their index finger when they felt pain. A photocell detected the finger withdrawal from the heat source and stopped the timer. Both index fingers were tested, and the heat intensities were adjusted on each trial, such that the 5 different heat intensities were presented in random order. The withdrawal latencies in seconds from left and right index fingers were then averaged to produce a stable estimate. It was further ensured prior to each test trial that finger temperature was between 88°F and 92°F before initiating a given trial. This was measured by a thermistor attached to the middle finger of one hand; if the temperature criterion was not met, the fingers were heated or cooled as necessary. The primary dependent measure for this test was mean finger withdrawal latency (FWL) in seconds averaged over the 5 heat intensities.

Study protocol
The practice of the sleep center is to have a preliminary evaluation of the overnight NPSG by a sleep physician before the patient leaves that morning. Patients who were found to have severe OSA were screened by the investigators and those who fulfilled the study entry criteria were approached regarding the study. Those who consented were then taken to a separate room where the training session for pain assessment was done. After the training, the first pain assessment testing was done between 08:00 and 09:00 the morning after the initial diagnostic NPSG. The patient was then discharged and scheduled for their CPAP titration study night. Once a CPAP pressure was identified to treat the OSA, the patient was instructed to use the prescribed CPAP pressure nightly and come back after 2 nights. In the next visit, the subject slept in the sleep laboratory with the CPAP at the predetermined therapeutic pressure. The morning after tertiary care hospital between January and June 2009. The study participants were patients who were referred to the sleep laboratory by their physicians with a suspicion of OSA and underwent a diagnostic NPSG the preceding night of enrollment. In order to be included in the study, subjects had to be consenting adults, with an Epworth Sleepiness Scale (ESS) score ≥ 10/24, documented severe OSA (i.e., AHI > 30) on the NPSG , no neurological sensory or motor deficits, willing to undergo the pain testing procedure and complete the scheduled follow-up visits for pain testing. Patients were excluded from the study if they had a history of acute or chronic pain, were taking any analgesics, had a history of tobacco use, alcohol or drug abuse, had a coagulation disorder or were on anticoagulant medications, including ASA, had periodic limb movement disorder and did not meet all the inclusion criteria. After enrollment in the study, the patients had a physical examination including BMI measurements and a medical history was taken.
The study protocol was approved by the institutional review board, and subjects were compensated monetarily for their participation. All participants signed an informed consent and had the opportunity to withdraw after experiencing the pain testing procedures during the training sessions. Further, the consent clearly indicated that their CPAP treatment would be withdrawn for 2 nights and they would not be able to drive during those 2 nights and days.

Sleep recordings
Standard methods for NPSG assessment were used. Central and occipital electro-encephalogram (EEG) leads were placed to measure EEG activity. Left and right horizontal electroculograms (EOG) were used to measure eye movements. A V5 electrocardiogram lead and a submental electromyogram (EMG) lead were placed. Airflow (by pressure transducer), body position via video recording, thoracic and abdominal excursion (inductance plethysmography), oxygen saturation (finger pulse oximetry), leg movement (with electrode over left anterior tibialis muscle), and sound were also recorded. Total time in bed was 8 h. Minimum accepted sleep efficiency was 50%.
All recordings were scored in 30-sec epochs. Scoring was based on published standards for sleep 18 and arousal scoring, 19 by technicians with established intra-laboratory scoring concordance and verified by independent sleep physician. New AASM recommended criteria were used to score apneas and hypopneas. 20 Hypopneas were scored if there was ≥ 30% reduction in nasal pressure signal excursions from baseline and associated ≥ 4% desaturation from pre-event baseline.
In the CPAP titration night, the CPAP pressures were titrated to relieve the apneas and hypopneas to a minimum of < 5/hour. The face masks were used based on the patient's preference and included full face masks, nasal masks or nasal pillows. The recording parameters were the same as described above.

Pain sensitivity assessment
A radiant heat methodology was used to assess pain sensitivity, 21 referred to as Finger Withdrawal Latency Testing (FWLT). Participants initially underwent a training pain assessment the morning after their diagnostic NPSG. This was indicating improvement in the pain sensitivity ( Figure 2). FWL increased in 10 of the 12 subjects. With continued CPAP use (5.1 ± 2.3 h nightly) for 6-8 weeks, FWL remained elevated (21.1 ± 16.2 sec). After only 2 nights of CPAP discontinuation, severe apnea returned (AHI = 32.6 ± 19.8, P < 0.01) and FWL decreased signifi cantly to 11.5 ± 5.9 sec relative to chronic CPAP use (P = 0.03; Figure 2).

DISCUSSION
The results of this study show that in patients with severe obstructive sleep apnea including excessive sleepiness, pain sensitivity is reduced (i.e., FWL is increased) with CPAP treatment. The effect persisted over the 6-8 weeks of CPAP use. After discontinuing CPAP for even two nights, the pain sensitivity increased in these individuals (i.e., FWL was reduced to baseline levels). In this study pain sensitivity was assayed via FWLT procedures. This method of pain assessment has been used and validated in previous studies. 5,21-23 However, as an internal validation in the current study, all the patients showed a signifi cant reduction in fi nger withdrawal latency as a function of the fi ve 2 nights of CPAP usage, between 08:00 and 09:00, the subject underwent a second FWLT measurement. The third FWLT pain assessment was done 6-8 weeks later between 08:00 and 09:00 while using their therapeutic CPAP. To further test treatment effects, the patients were then taken off the CPAP again for 2 nights and underwent the fi nal FWLT assessment again between 08:00 and 09:00. For those 2 days without nightly CPAP use, patients were advised not to drive and avoid operating machinery.

Data Analysis
The baseline FWL latencies at each of the 5 heat intensities were compared using a MANOVA to validate the testing procedure. The primary dependent measure to assess treatment effects was the mean FWL in seconds averaged over the 5 intensities. FWL was compared between the baseline short-term CPAP treatment, after 6-8 wks of CPAP use, and fi nally after 2 nights of discontinuation (within 6-8 wks) of CPAP use. Mean FWLs were compared using MANOVAs with nights as a repeated variable, followed by post hoc t-tests comparing nights.

RESULTS
During the study period, about 90 patients were screened. Of them, 28 fulfi lled the inclusion criteria; 22 were approached regarding the study, and 12 consented. The 6 patients not approached were missed as they left early after their initial NPSG, and 10 did not consent for the study. None of the patients withdrew from the study after consenting.
The demographic characteristics of the 12 patients are shown in Table 1. All the patients were sleepy based on the Epworth Sleepiness Scale (ESS) score and had severe obstructive sleep apnea. On CPAP, sleep continuity was improved and AHI and ESS scores normalized (Table 1). After discontinuation of CPAP for 2 nights, sleep became fragmented again associated with a return of their apnea/hypopneas (AHI = 32.6 ± 19.8, P < 0.01; Figure 1).
At baseline FWL varied signifi cantly as a function of heat intensity, varying from 14.1 ± 2.1 to 6.1 ± 0.9 sec at the lowest to highest intensities (P < 0.001). Average FWL over the 5 intensities increased signifi cantly with short-term CPAP usage from a mean baseline of 9.8 ± 1.3 sec to 13.7 ± 5.1 sec (P = 0.01),  In summary, this study shows that severe OSA patients are hyperalgesic and their pain sensitivity improves with CPAP treatment, and the improvement disappears immediately with discontinuation of CPAP therapy.
increasing heat intensities at baseline, even the two patients that failed to show reduced pain sensitivity in response to CPAP.
The changes in pain sensitivity with variations in degree of sleep fragmentation parallel studies of sleep restriction, that produce sleepiness and increase pain sensitivity in healthy study volunteers. 5,9,12,13 An epidemiological study in the United States also provided similar evidence in a sample from the general population that sleep duration (especially sleep times less than 6 h) was associated with greater next-day pain. 24 Finally, in a group of inpatients with major depressive disorder, sleep deprivation resulted in decreased heat pain thresholds and augmentation of their pain complaints. 25 In contrast to the effects of sleep deprivation, the association of sleep fragmentation and pain is less well defined. In a study of healthy women, it was suggested that sleep fragmentation increases spontaneous pain reports, supporting a etiological role of sleep disturbance in chronic pain. 10 In another study of middle-aged women, slow wave sleep disruption, without reducing total sleep, for several consecutive nights was associated with decreased pain threshold and increased discomfort. 12 Severe obstructive sleep apnea is associated with numerous arousals causing significant sleep fragmentation. A study of a small group of elderly patients with moderate obstructive sleep apnea found that CPAP treatment with high vs low CPAP pressures (4 cm vs 5-10 cm H 2 O) showed trends in correlations between electrical pain tolerance and apnea indicies. 26 The present study is the first study to definitively demonstrate statistically significant effects of short-term and intermediate-term CPAP therapy, as well as its discontinuation on pain sensitivity.
The exact mechanism(s) and specificity of this analgesic effect of CPAP in OSA patients is unknown and requires further investigation. Is the effect the result of improved sleep continuity or improved oxygenation? How does the improved sleep and/ or breathing alter pain physiology and is that modulation of pain sensation occurring peripherally or centrally? Does the reduced radiant heat sensitivity extend to other pain modalities (i.e., cold or mechanical stimulation) and to other sensory modalities?
The results of this study also raise other clinical questions needing investigation. Patients with untreated severe sleep apnea are usually fatigued during the day and have headaches.
It is unclear what percentages of these severe OSA patients have comorbid pain conditions. But, usually these patients are morbidly obese and are prone to have musculoskeletal pain. Untreated OSA, based on our study results, would have the potential to exacerbate any such underlying pain condition. Studies including an obese, non-apneic comparison group would be important, as would studies that explore gender, age, and comorbid disease effects. This study is limited in its sample size, which negates its power to explore these various questions. The sample of this study was predominately non-Caucasian, which is representative of our patient population, but is also a limitation of this study.
It is also unknown whether untreated OSA patients, who are taking analgesics for their chronic pain conditions, would see a reduction in usage of their pain medications with CPAP treatment, although this needs to be demonstrated. Future studies should also evaluate OSA patients with chronic pain conditions to see the impact of the CPAP therapy on the likelihood of pharmacological dependence in them.