ISSN 1550-9389 iN this issue · ISSN 1550-9389 Volume 8, Number 6 / December 15, 2012 Pages 627-734...

ISSN 1550-9389

Volume 8, Number 6 / December 15, 2012Pages 627-734

iN this issue:

Official Publication of the American Academy of Sleep Medicine

www.aasmnet.org

(Mis) Perceptions and Interactions of Sleep Specialists and Generalists: Obstacles to Referrals to Sleep Specialists and the Multidisciplinary Team Management of Sleep DisordersHayes; Murray; Castriotta; Landrigan; Malhotra

The Impact of Body Posture and Sleep Stages on Sleep Apnea Severity in AdultsEiseman; Westover; Ellenbogen; Bianchi

The Impact of Posttraumatic Stress Disorder on CPAP Adherence in Patients with Obstructive Sleep ApneaCollen; Lettieri; Hoffman

Use of Relaxation Techniques and Complementary and Alternative Medicine by American Adults with Insomnia Symptoms: Results from a National SurveyBertisch; Wells; Smith; McCarthy

Vitamin D, Race, and Excessive Daytime SleepinessMcCarty; Reddy; Keigley; Kim; Marino

Commentary on McCarty et al.Vitamin D as an Underlying Factor in Sleep-Related IssuesAndersen; Tufik

Journal of Clinical Sleep Medicine

December 15, 2012

Volume 8(6) 2012 Pages 627-734

Scope

JCSM Journal of Clinical Sleep Medi-cine focuses on clinical sleep medicine. Its emphasis is publication of papers with direct applicability and/or relevance to the clinical practice of sleep medicine. This includes, clinical trials, clinical re-views, clinical commentary and debate, medical economic/practice perspec-tives, case series and novel/interesting case reports. In addition, the journal will publish proceedings from conferences, workshops and symposia sponsored by the American Academy of Sleep Medi-cine or other organizations related to im-proving the practice of sleep medicine.

JCSM Journal of Clinical Sleep Medi-cine (Online 1550-9397; Website: www.aasmnet.org/jcsm) is published on-line 6 times per year: February, April, June, August, October, and December by the American Academy of Sleep Medicine, 2510 North Frontage Road, Darien, IL 60561-1511, phone (630) 737-9700 and fax (630) 737-9790.

ANNUAL SUBSCRIPTION RATES: Subscription rates for Volume 8, 2012: Individual Online (US and International):

$75.00; Institutional Online (US and Inter-national): $140.00. Mid-year subscriptions are not available. Subscriptions begin with the February issue of the current year. Re-newals should be secured as early in the year as possible to avoid uninterrupted service. Questions about subscriptions (including payments, billing procedures, or policy matters) should be directed to the AASM office at (630) 737-9700.

ADVERTISING: Digital advertis-ing is available on www.aasmnet.org/

jcsm. Please contact the National Sales Account Executive at [email protected] for complete information.

PERMISSION TO REPRODUCE: Writ-ten permission to reproduce, in print or electronically, whole articles or any parts of works, figures or tables published in JCSM must be obtained prior to pub-lication. Permission for republication must be arranged through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, phone (978)

750-8400 or fax (978) 646-8600 or URL http://www.copyright.com. There are royalty fees associated with such permissions.

REPRINTS: For author reprints contact the AASM office. For commercial re-print orders contact Pete Brown, Cadmus Printing, 500 Cadmus Lane, Easton, MD 21601. [email protected]

DISCLAIMER: The statements and opinions contained in editorials and ar-ticles in this journal are solely those of

the authors thereof and not of the Ameri-can Academy of Sleep Medicine, or of its officers, regents, members or employees. The Editor-in-Chief, the American Acad-emy of Sleep Medicine and its officers, regents, members and employees dis-claim all responsibility for any injury to persons or property resulting from any ideas or products referred to in articles contained in this journal.

© 2012 American Academy of Sleep Medicine

ASSociAte editorS

Richard B. Berry, M.D., F.A.A.S.M., Gainesville, FL

Lee K. Brown, M.D., F.A.A.S.M., Albuquerque, NM

Rohit Budhiraja, M.D., F.A.A.S.M., Tucson, AZ

Andrew L. Chesson, M.D., F.A.A.S.M., Shreveport, LA

David Dinges, Ph.D., Philadelphia, PA

Susan M. Harding, M.D., F.A.A.S.M., Birmingham, AL

Birgit Högl, M.D., Innsbruck, Austria

Conrad Iber, M.D., Minneapolis, MN

Vishesh Kapur, M.D., F.A.A.S.M., Seattle, WA

Douglas B. Kirsch, M.D., F.A.A.S.M., Brighton, MA

Michael Littner, M.D. F.A.A.S.M., Sepulveda, CA

W. Vaughn McCall, M.D., F.A.A.S.M., Winston-Salem, NC

Amy L. Meoli, M.D., F.A.A.S.M., Kansas City, MO

Sairam Parthasarathy, M.D., F.A.A.S.M., Tucson, AZ

Stephen H. Sheldon, D.O., F.A.A.S.M., Chicago, IL

Nathaniel F. Watson, M.D., F.A.A.S.M., Seattle, WA

editorStuart F. Quan, M.D., F.A.A.S.M., Boston, MA/Tucson, AZ

executive director

Jerome A. Barrett, Darien, IL

MAnAging editor

Andrew Miller, Darien, IL

Candice A. Alfano, Ph.D., Houston, TX

Fernanda R. Almeida, D.D.S., M.Sc., Ph.D., A.B.D.S.M., Vancouver, BC, Canada

Monica L. Andersen, M.Sci., Ph.D., Sao Paulo, Brazil

Donna Arand, Ph.D., F.A.A.S.M.,Dayton, OH

Kristen Archbold, Ph.D., R.N., Tucson, AZ

Carol M. Baldwin, Ph.D., R.N., Tempe, AZ

Mary Susan Esther Banks, M.D., Charlotte, NC

Kenneth R. Casey, M.D., F.A.A.S.M.,Cincinnati, OH

Armando Castorena-Maldonado, M.D.Mexico City, Mexico

Alejandro D. Chediak, M.D., F.A.A.S.M., Miami Beach, FL

Ronald D. Chervin, M.D., Ann Arbor, MI

Madeline Grigg-Damberger, M.D., Albuquerque, NM

Sally L. Davidson Ward, M.D., F.A.A.S.M., Los Angeles, CA

William C. Dement, M.D., Ph.D., F.A.A.S.M., Palo Alto, CA

Lawrence J. Epstein, M.D., F.A.A.S.M., Bedford, MA

Birgit Frauscher, M.D., Innsbruck, Austria

Geoffrey S. Gilmartin, M.D., F.A.A.S.M., Boston, MA

Leila K. Gozal, M.D., Chicago, IL

Jason P. Kirkness, Ph.D., Baltimore, MD

Meir H. Kryger, M.D., F.A.A.S.M.,West Haven, CT

Clete A. Kushida, M.D., Ph.D., F.A.A.S.M., Stanford, CA

Peretz Lavie, Ph.D., Haifa, Israel

Teofilo L. Lee-Chiong, M.D., F.A.A.S.M., Denver, CO

Ching-Chi Lin, M.D., Taipei, Taiwan

Matthew T. Naughton, M.D., Prahran, Australia

James M. Parish, M.D., Scottsdale, AZ

Thomas Penzel, Ph.D., Berlin, Germany

Barbara A. Phillips, M.D., Lexington, KY

Daniel Picchietti, M.D., F.A.A.S.M.,Urbana, IL

Giora Pillar, M.D., Ph.D., Haifa, Israel

Naresh M. Punjabi, M.D., F.A.A.S.M.,Baltimore, MD

Daniel O. Rodenstein, M.D., Ph.D., Brussels, Belgium

Wolfgang W. Schmidt-Nowara, M.D., F.A.A.S.M., Dallas, TX

Shirin Shafazand, M.D., F.A.A.S.M., Miami, FL

Aaron E. Sher, M.D., Albany, NY

Michael H. Silber, M.B., Ch.B., F.A.A.S.M., Rochester, MN

Edward J. Stepanski, Ph.D., F.A.A.S.M., Memphis, TN

Patrick J. Strollo, Jr., M.D., F.A.A.S.M., Pittsburgh, PA

David P. White, M.D., F.A.A.S.M.,Boston, MA

B. Tucker Woodson, M.D., F.A.A.S.M., Milwaukee, WI

editoriAl BoArd

deputy editor

Daniel J. Buysse, M.D., F.A.A.S.M., Pittsburgh, PA

Official publication of the American Academy of Sleep Medicine

Volume 8, Number 6

December 15, 2012

Pages 627-734

table of contents Vol. 8, No. 6

The current issue podcast and instructions to authors are available online at www.aasmnet.org/jcsm.

CME Opportunities in the Issue630

Editorial631The Journal of Clinical Sleep Medicine—Moving ForwardStuart F. Quan

Scientific Investigations633(Mis) Perceptions and Interactions of Sleep Specialists and Generalists: Obstacles to Referrals to Sleep Specialists and the Multidisciplinary Team Management of Sleep DisordersSean M. Hayes; Suzanne Murray; Richard J. Castriotta; Christopher P. Landrigan; Atul Malhotra

643Lower Frequency of Obstructive Sleep Apnea in Spondyloarthritis Patients Taking TNF-InhibitorsJessica A. Walsh; Kristina Callis Duffin; Julia Crim; Daniel O. Clegg

649Association between QRS Duration and Obstructive Sleep ApneaShuchita Gupta; Beatriz Cepeda-Valery; Abel Romero-Corral; Abu Shamsuzzaman; Virend K. Somers; Gregg S. Pressman

655The Impact of Body Posture and Sleep Stages on Sleep Apnea Severity in AdultsNathaniel A. Eiseman; M. Brandon Westover; Jeffrey M. Ellenbogen; Matt T. Bianchi

667The Impact of Posttraumatic Stress Disorder on CPAP Adherence in Patients with Obstructive Sleep ApneaJacob F. Collen; Christopher J. Lettieri; Monica Hoffman

673Sleep Related Expiratory Obstructive Apnea in ChildrenMark E. Haupt; Denise M. Goodman; Stephen H. Sheldon

681Use of Relaxation Techniques and Complementary and Alternative Medicine by American Adults with Insomnia Symptoms: Results from a National SurveySuzanne M. Bertisch; Rebecca Erwin Wells; Michael T. Smith; Ellen P. McCarthy

693Vitamin D, Race, and Excessive Daytime SleepinessDavid E. McCarty; Aronkumar Reddy; Quinton Keigley; Paul Y. Kim; Andrew A. Marino

699Commentary on McCarty et al.Vitamin D as an Underlying Factor in Sleep-Related IssuesMonica Levy Andersen; Sergio Tufik

701Evaluating Sleepiness-Related Daytime Function by Querying Wakefulness Inability and Fatigue: Sleepiness-Wakefulness Inability and Fatigue Test (SWIFT)R. Bart Sangal

Case Reports713Effect of Laser Arytenoidectomy on Respiratory Stridor Caused by Multiple System AtrophyShun-ichi Chitose; Atsushi Kikuchi; Keiko Ikezono; Hirohito Umeno; Tadashi Nakashima

717“Myxedema Madness” Associated with Newly Diagnosed Hypothyroidism and Obstructive Sleep ApneaJ. Matthew Neal; Rodney Joe O. Yuhico

719Sleep Abnormalities in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis: A ReviewMelinda L. Jackson; Dorothy Bruck

table of contents Vol. 8, No. 6

The current issue podcast and instructions to authors are available online at www.aasmnet.org/jcsm.

Sleep Medicine Pearls730“Why Did My CPAP Beat Me Up?” Bilateral Periorbital Ecchymosis Associated with Continuous Positive Airway Pressure TherapyLourdes DelRosso; David E. McCarty; Romy Hoque

Letter to the Editor733Gabapentin Efficacy in Reducing Nighttime Awakenings in Premenopausal Women: A Class Effect of GABAergic Medications or Unique Property of Gabapentin Only?Gautam Ganguly

630Journal of Clinical Sleep Medicine, Vol. 8, No. 6, 2012

Continuing MediCal eduCation offeringsinstructions for earning creditParticipants may read the selected continuing medical education (CME) articles in this issue of JCSM and complete the online CME post-test and evaluation at http://www.aasmnet.org/JCSM/ within one year of the date of publication to receive AMA PRA Category 1 CreditsTM. Each activity – journal article, post-test, and evaluation – is estimated to take approximately 30 minutes to complete.

The ACCME mandates that accredited providers only offer AMA PRA Category 1 CreditsTM to physicians. Non-physicians will be provided with a letter of participation indicating the number of AMA PRA Category 1 CreditsTM awarded for the activity in which they participated. Non-physicians requesting letters of participation will be assessed the same fees as physicians requesting AMA PRA Category 1 Credit™, if applicable; visit the journal web site for information on applicable fees.

Each journal article designated for CME is worth 0.5 AMA PRA Category 1 CreditsTM.. To earn the credit, physicians must read the article and then correctly answer a minimum of 3 out of 5 post-test questions. Upon successful completion of the post-test, the learner may download a participation letter indicating the number of credits earned.

Accreditation StatementThe American Academy of Sleep Medicine is accredited by the ACCME to provide continuing medical education for physicians.The American Academy of Sleep Medicine designates this educational activity for a maximum of 0.5 category 1 credits toward the AMA PRA Category 1 CreditsTM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Statement of educational purpose/overall education objectivesJCSM is a peer-reviewed clinical journal addressing sleep, circadian rhythms, and the diagnosis and treatment of the broad spectrum of sleep disorders. Its mission and educational purpose is to promote the science and art of sleep medicine and sleep research. Sleep disorders medicine draws clinical and scientific applications from a wide variety of primary disciplines, including pulmonology, neurology, psychiatry, psychology, otolaryngology, and dentistry. Readers of JCSM should be able to: 1) appraise sleep research in basic science and clinical investigation; 2) interpret new information and updates on clinical diagnosis/treatment and apply those strategies to their practice; 3) analyze articles for the use of sound scientific and medical problems; and 4) recognize the inter-relatedness/dependence of sleep medicine with primary disciplines.

The following articles in this issue may be read for CME credit: Beginning page #

the impact of Body posture and Sleep Stages on Sleep Apnea Severity in Adults 655Objective: Recognize the importance of considering body position and sleep stage in the interpretation of single-night sleep studies.

the impact of posttraumatic Stress disorder on cpAp Adherence in patients with obstructive Sleep Apnea 667Objective: Understand the relationship between PTSD and sleep disturbances, in particular the adverse impact of PTSD on CPAP adherence.

631 Journal of Clinical Sleep Medicine, Vol. 8, No. 6, 2012

http://dx.doi.org/10.5664/jcsm.2250

ED

ITO

RIA

L

Beginning in 2013, our 9th volume, there will changes com-ing to the Journal. The major change will be monthly pub-

lication which is an additional landmark in the evolution of the Journal. By doing so, manuscripts will be published sooner and each issue will contain slightly fewer papers. The latter will allow the reader to focus on a reduced amount of information, hopefully resulting in greater transfer of knowledge. In addi-tion, your editorial staff is considering enhancement of items directly relevant to clinical practice, such as more board review questions and a journal club.

During the eight years I have had the privilege of being your editor, the Journal has made major strides into becoming the leading scientifi c publication for Sleep Medicine clinicians. In our 1st year, there were fewer than 100 original manuscripts or case report submissions. For this our 8th year, there will be close

The Journal of Clinical Sleep Medicine—Moving ForwardStuart F. Quan, M.D., F.A.A.S.M., Editor-in-Chief

Boston, MA and Tucson, AZ

to 250 submissions in these categories. Many are from interna-tional institutions, demonstrating world-wide recognition of the Journal. Although, I have not emphasized achieving a high im-pact factor, the Journal’s fi rst impact factor was 3.232, ranking it 3rd among the seven non-review sleep journals receiving impact factors. Many papers have been highly cited. Despite the in-creasing number of submissions, the quality of papers has been high allowing greater selectivity in what is ultimately published.

As the Journal enters its 9th year, it will remain clinically focused with its primary mission to be a venue for transfer of clinically relevant knowledge to Sleep Medicine clinicians, as well as to clinical investigators. Your editorial staff is open to suggestions on how to make the Journal better and relevant to the practice of Sleep Medicine as the Journal embarks on a new era.


Study Objectives: This study assessed generalists’ percep-tions and challenges in providing care to sleep disorders patients and the role of sleep specialists in improving gaps in care.Methods: A mixed-method approach included qualitative (semi-structured interviews, discussion groups) and quantita-tive (online surveys) data collection techniques regarding care of patients with obstructive sleep apnea (OSA) and shift work disorder (SWD).Results: Participants: OSA: generalists n = 165, specialists (internists, neurologists, psychiatrists, pulmonologists) n = 12; SWD: generalists n = 216, specialists n = 108. Generalists re-ported challenges in assessing sleep disorders and diagnosing patients with sleep complaints. Generalists lacked confi dence (selected ≤ 3 on a 5-pt Likert scale) in managing polyphar-macy and drug interactions (OSA: 54.2%; SWD: 62.6%), ad-diction (OSA: 61.8%), and continuous positive airway pressure (OSA: 66.5%). Generalists in both studies reported defi cits in knowledge of monitoring sleep disorders (OSA: 57.7%; SWD:

78.7%), rather relying on patients’ subjective reports; 23% of SWD generalists did not identify SWD as a medical condition. Challenges to generalist-specialist collaboration were report-ed, with 66% of generalists and 68% of specialists in the SWD study reporting lack of coordination as a barrier. Generalists reported lack of consistency in sleep medicine and a perceived lack of value in consulting with sleep specialists.Conclusions: Knowledge and attitudinal challenges were found in primary care of patients with sleep disorders. Sleep specialists need to clarify and educate practitioners regarding primary care’s approach.Keywords: Sleep disorders, obstructive sleep apnea, shift work disorder, primary care, qualitative research, physician competence, multidisciplinary careCitation: Hayes SM; Murray S; Castriotta RJ; Landrigan CP; Malhotra A. (Mis) perceptions and interactions of sleep special-ists and generalists: obstacles to referrals to sleep specialists and the multidisciplinary team management of sleep disorders. J Clin Sleep Med 2012;8(6):633-642.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Inadequate sleep is widespread in the United States,1,2 affect-ing approximately 50-70 million Americans.1 Chronic lack

of sleep and insomnia have important physiological and health effects as well as psychological, cognitive, safety, social, and economic implications.3 Obstructive sleep apnea (OSA) and shift work disorder (SWD) are two common sleep disorders with potentially serious social impact. OSA affects roughly 5% to 10% of the U.S. population or 18 million Americans,4 in-cluding an estimated 11.6% of the shift work population.5 Shift work disorder, interruption of the normal sleep/wake cycle by the need to work shifts, or non-traditional daytime hours, and is referred to as circadian rhythm sleep disorder, shift work type, or shift work disorder (SWD).6 Approximately 20% of workers are estimated to work non-traditional schedules.7 Of these, 10% have been found to experience SWD.8

In spite of their prevalence, substantial impact, and the availability of effective treatment strategies, sleep disorders are generally underdiagnosed and undertreated by healthcare providers.1,2 The annual Sleep in America survey2 reported that 86% of respondents’ generalists had never discussed sleep with them. Six of ten healthcare professionals reported not having enough time to discuss sleep problems during offi ce visits.2

(Mis) Perceptions and Interactions of Sleep Specialists and Generalists: Obstacles to Referrals to Sleep Specialists and the

Multidisciplinary Team Management of Sleep DisordersSean M. Hayes, Psy.D.1; Suzanne Murray1; Richard J. Castriotta, M.D., F.A.A.S.M.2; Christopher P. Landrigan, M.D., M.P.H.3,4;

Atul Malhotra, M.D., F.A.A.S.M.3,4

1AXDEV Group Inc., Quebec, Canada; 2University of Texas Medical School at Houston, Houston, TX; 3Brigham and Women’s Hospital, Boston, MA; 4Harvard Medical School, Boston, MA

Even when patients were asked about sleep issues, providers neither managed them directly nor referred patients to special-ists. In another study, 90% of generalists rated their knowledge of sleep disorders as fair or poor.9 Patients of sleep specialists exhibited greater awareness of the OSA management process,

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Sleep disorders are generally under-diagnosed and undertreated by primary care providers and not optimally referred to sleep specialists. However, there has been limited in depth assessment of the etiology of barriers faced by generalists in their assessment of sleep disorders, and how to optimize the role of sleep specialists in patient care.Study Impact: Knowledge, skill, and attitudinal challenges and gaps were identifi ed among a national sample of primary care providers in-volved in the care of patients with sleep disorders, resulting in patients being under-diagnosed, undertreated, stigmatized, and under-priori-tized. Challenges in understanding and enhancing the role and value of the sleep specialist to the primary care community, as well as incor-porating them to the interdisciplinary team for optimal care, were also identifi ed. This study reveals a defi cit in generalist physician knowledge about sleep medicine and a large gap between generalist and specialist attitudes concerning the diagnosis and management of sleep disorders.


SM Hayes, S Murray, RJ Castriotta et although no difference was seen in their acceptance or compli-ance with CPAP.10 This pattern of diagnostic delay and misdi-agnosis has been seen with narcolepsy as well.11

Collaboration between generalists and sleep specialists has been recommended to alleviate the gaps in primary care provid-ers’ knowledge of sleep disorders,12 with guidelines for OSA recommending multidisciplinary team care.13 However, true multidisciplinary, shared care—while logical—can be chal-lenging to achieve. Specialists reported that while some pri-mary care physicians have been receptive to sleep education in the past, they were reluctant to take the responsibility for in-terpreting polysomnography results and for treatment.9 Further, the healthcare system as a whole has not adapted to the care of sleep disorders patients and does not provide the resources or infrastructure needed to provide effective care across the con-tinuum of the patient experience for sleep disorders.9

In light of these and related data, the Institute of Medicine called for efforts to expand the awareness of sleep disorders among healthcare professionals through education and train-ing.1 However, barriers to primary providers’ diagnosis and management of sleep disorders remain largely unknown. More-over, we wished to gain a better understanding of the issues and challenges between generalists and sleep specialists that might undermine optimal patient care. Consequently, we conducted a national applied behavioral performance needs assessment with the following goals:

1. To identify challenges, as well as educational and perfor-mance gaps of generalist primary care physicians in pro-viding care to sleep disorders patients, focusing on OSA and SWD.

2. To assess the roles, perceptions, attitudes, and interactions of generalists and sleep specialists in providing care for patients with sleep disorders.

3. To provide a baseline to evaluate the impact and outcomes of future educational and performance improvement in-terventions.

METHODS

Following IRB approval, we conducted educational and performance needs assessments among nationwide samples of participants that focused on challenges and gaps in the treat-ment and management of OSA (data collection May – August 2008) and SWD (data collection March – June 2009). Mixed-methods approach that included both qualitative and quantita-tive data collection techniques in a triangulated research design were employed.14,15 A triangulated research design involves ex-amination of several data sources using multiple data collection methods to examine the same phenomena,15,16 enhancing the trustworthiness and validity of the findings.

ParticipantsPurposive sampling was used to ensure that the sample was

representative of the target audience.15,16 Participants were drawn from existing subject pools and recruited via telephone, fax, and e-mail in collaboration with the American Thoracic Society (ATS, www.thoracic.org), the New Jersey Academy of Family Physicians (NJAFP, www.njafp.org), the Office of Continuing Medical Education of the University of Virginia

School of Medicine (UVA-OCME, www.medicine.virginia.edu/education/more/cme/home-page), and the University of Wisconsin School of Medicine and Public Health Office of Continuing Professional Development (UW-SMPH-OCPD, www.ocpd.wisc.edu). The nature of the distribution of the invitations precluded identifying and following up with non-responders. Institutional review board approval was obtained for both studies. Financial compensation was provided to par-ticipants for their time.

Generalist participants were family physicians and internists who reported seeing at least one patient with sleep disorders (OSA or SWD) per month or whose patient population included ≥ 2% of patients who worked > 8 h/day and/or outside tradition-al daytime hours (07:00-18:00).17 Specialists (internists special-izing in sleep, neurologists, psychiatrists, pulmonologists) were included who saw ≥ 1 patient/month with sleep disorders (OSA or SWD) or who reported having ≥ 5% of their patients who worked shifts. The inclusion criteria were determined in col-laboration with faculty and a review of the literature.

Data CollectionBest practices and challenges in the care of patients with OSA

and SWD were identified through a comprehensive literature re-view. Key concepts provided a framework to guide the design of qualitative data collection instruments. Topics to emerge from this process included contextual issues, gaps throughout the con-tinuum of care, inter-professional collaboration and referral gaps, and specific educational needs in sleep disorders. Based on this framework, comprehensive discussion group and semi-structured interview guides were developed to explore the practice and expe-riences of generalists and specialists. The discussion groups were lead by an expert-facilitator who asked participants questions re-lated to the developed framework. Participants were encouraged by the facilitator to share their thoughts and opinions, and engage with their peers on the topics discussed. For in-depth understand-ing, probes consisting of open-ended questions addressing issues around knowledge, skill, attitudes, healthcare team and system, and current and desired practice were used in the discussion groups and interviews. Discussion groups were approximately 3.5 h and interviews were approximately 60 minutes. Both discussion groups and interviews were audio-recorded.

Quantitative surveys were developed based on substantive qualitative findings and key concepts identified in the literature review. The OSA survey consisted of 48 items, and the SWD survey of 75-78 items, consisting of rating statements on a 5-point scale (where 1 = low, 5 = high; see tables 6 and 7). For example, participants were asked to select the number that best describes how they currently evaluate their knowledge relative to the statement presented. Next, they were asked to indicate their desired level of knowledge, defined as the level the partici-pants would like to have or feel they need to attain.

AnalysisQualitative analysis was through open coding.18 Coders were

experienced qualitative researchers, including co-authors SH, SM, and KC. Coding categories were then grouped into related themes and subthemes, such as: Knowledge—lack of knowl-edge of diagnostic testing, and Attitude—lack of prioritiza-tion of sleep disorders. Themes were validated among coders


Sleep Specialists and Generalists: Obstacles to Referralsthrough review of selected data excerpts and discussion of cod-ing. Discrepancies were resolved through discussion until con-cordance was achieved. Concordance was achieved in all cases. Selective coding was then conducted,18 whereby data were sys-tematically coded with respect to core concepts identified in the literature review and analysis of interview data.

Quantitative analysis consisted of descriptives (means, fre-quencies), (SPSS 12.0, SPSS, Chicago, IL). Gap analysis was carried out, in which participants provided a self-assessment of both their current level and desired level of knowledge.14 The difference, or gap, between current and desired levels provides an indicator of specific areas of educational need. ANOVA validated the statistical significance of these gaps. A 5-point Likert agreement scale was used throughout each of the two surveys.

RESULTS

Findings revealed contextual and attitudinal barriers as well as clinical challenges across the continuum of care of patients with sleep disorders, regardless of specific diagnosis, and indi-cated key challenges relevant to inter-professional collabora-tion between generalists and sleep specialists.

Sample (Table 1 and 2)Five discussion groups (n = 32), 24 interviews, and 445

surveys were completed for a total sample of 401 participants. Forty physicians (7.7%) participated in both OSA and SWD studies. Seventy-three percent of physicians participating in the OSA study saw > 10 patients/month, compared to 45% of phy-sicians in the SWD study.

Knowledge and Skill in Care of Patient with Sleep Disorders

Screening and DiagnosisOne-third of generalists in the SWD and OSA studies ex-

pressed hesitancy in addressing both OSA and SWD, neither systematically nor proactively screening for sleep disorders, and were reluctant to raise the topic because they perceived sleep disorders as a complex issue to manage. Participants in the OSA study identified substantive knowledge gaps on as-sessment and differential diagnosis across multiple sleep disor-

ders (table 3). Physicians in the SWD study identified an even greater lack of knowledge of diagnosis of SWD (table 4), lacking knowledge of key questions to pose to make an ac-curate diagnosis and relying on subjective impressions and patient reports to support their diagnosis, rather than formal tools, guidelines, or criteria.

Treatment and ManagementChallenges were also identified in the treatment of sleep

disorders. Generalists in both studies articulated a lack of con-fidence in dealing with polypharmacy and evaluation of drug interactions, describing reluctance to prescribe stimulants for daytime due to fear of addiction, and lack of knowledge of treat-ment cessation. Generalists in the OSA study described knowl-edge gaps with regard to initiating CPAP therapy (table 3) as well as supporting patients on CPAP. Generalists in both studies reported relying on patients’ subjective reports in monitoring patients and treatment outcomes. Generalists did not character-ize guidelines as useful in determining their care

Attitudes toward Sleep DisordersGeneralist participants generally did not prioritize discussing

sleep disorders. They further demonstrated a lack of understand-ing of the impact of sleep disorders on patients’ daily living and comorbid conditions. While the majority of physicians believed in SWD as a medical condition, 23% of generalists and 16% of specialists did not. Some generalists in the OSA study reported characterizing sleep disorders as a symptom rather than as a primary diagnosis.

Table 2—Sample distribution of physicians participating in the online surveysOSA SWD

Specialty (total n = 145) n (%) Specialty (total n = 300) n (%)Generalists 75 (52) Family Physicians 127 (42)

70 (48) Internists 73 (24)Internal Medicine with sub-specialty in sleep 11 (4)

Neurologists 8 (3)Psychiatrists 13 (4)

Pulmonologists 68 (23)

Patients with OSA Per Month n (%) Patients with Sleep Disorders Per MonthGeneralists

n (%)Specialists

n (%)1-5 patients 43 (30) 1-5 patients 76 (38) 38 (38)

6-10 48 (33) 6-10 32 (16) 12 (12)11+ 54 (37) 11+ 89 (45) 47 (47)

Table 1—Sample distribution by data collection activityOSA SWD

Generalists Specialists Generalists SpecialistsQuantitative

Online Survey

145 N/A 200 100

QualitativeFocus Groups

20 12 N/A N/A

Interviews N/A N/A 16 8


SM Hayes, S Murray, RJ Castriotta et al

The Healthcare Team: Roles and ValueHealthcare team roles and responsibilities regarding OSA

and SWD were described as unclear. Participants in both studies reported a lack of clarity about generalists’ roles and responsibilities in regards to sleep disorders, with respect to which patients to refer to a sleep specialist, when referral was indicated, and to whom they should refer. This was particu-larly true of SWD, with less uncertainty related to OSA re-ferrals (table 5). In the SWD study, 66% of generalists and 66% of specialists described lack of role clarity as a barrier to optimal patient care (Figure 1A), as well as lack of coor-dination between generalists and specialists in treatment and management of SWD patients (generalists 66%, specialists 68%; Figure 1B). They described managing sleep disorders in isolation: 38.5% of generalists reported not referring to specialists for SWD.

A majority of generalists did not recognize sleep medicine as a specialty, characterizing it as a poorly defined area of expertise. Furthermore, all participant groups reported a lack of uniformity in training for sleep specialists, further hin-dering their full recognition of sleep medicine as a credible subspecialty.

“I don’t think they [generalists] see sleep as a unifying subspecialty. Someone may have snoring, they don’t see it as a referral to a sleep lab as they would for someone with insomnia or restless leg syndrome.”—Specialist

A majority of participants in both studies questioned whether specialists know much more than generalists about sleep disor-ders; and perceived a lack of value for patients in referring to specialists and sleep centers. Thus, only a minority of generalists reported referring patients to a limited group of sleep specialists.

“I find it somewhat difficult to diagnose OSA and the experts don’t do much better. I think they over-diagnose.”—Generalist

Generalists further expressed reluctance to refer patients to sleep specialists because they perceived that specialists assess and diagnose but do not treat or manage OSA patients.

“It seems that the emerging group of sleep specialists are more than willing to do the test and make the diagnosis, but not to follow with the treatment, compliance, etc. Spe-cialists make the money and leave the hard stuff for the primary care physician.”—Generalist

Sleep laboratories were also viewed with some skepticism:

“And there are a lot of shabby labs, some of them run by medical device companies.”—Generalist

Because of the perceived lack of value of sleep lab reports, generalists expressed reluctance to refer patients to sleep labo-ratories. Since generalists were not convinced of the value of

Table 3—Gap analysis in care of patients with OSAGeneralists (n = 145)

DesiredLevel

Percentage of Participants Gap T1

Low2 3 4 5

High XSD p

DiagnosisAdminister and interpret tests correctly assessing patients with sleeping disorders

Current 10.3 17.9 43.4 23.4 4.8 1.3* -17.02Desired 2.1 4.2 8.5 37.3 47.9 0.93 0.000

Diagnose circadian rhythm sleep disorders Current 8.3 23.6 36.8 27.8 3.5 1.5* -18.00Desired 1.4 0.7 5.6 39.2 53.1 0.98 0.000

Diagnose narcolepsy Current 6.9 27.1 40.3 19.4 6.3 1.4* -16.78Desired 2.1 0.7 7 44.4 45.8 1.0 0.000

Diagnose substance-induced sleep issues Current 2.1 12.5 47.2 34.7 3.5 1.3* -18.54Desired 0 0.7 4.3 32.6 62.4 0.85 0.000

TreatmentEvaluate and manage potential drug interactions in SD Current 2.1 17.4 34.7 39.6 6.3 1.3* -17.18

Desired 0 0 4.2 30.3 65.5 0.91 0.000Initiate CPAP treatment for OSA Current 14.7 23.1 28.7 25.9 7.7 1.3* -13.93

Desired 3.5 4.3 15.6 25.5 51.1 1.1 0.000

ManagementMonitor OSA status and progression Current 6.3 15.3 36.1 36.8 5.6 1.3* -16.96

Desired 0.7 0.7 4.9 32.9 60.8 0.94 0.000Manage patients with OSA according to current clinical guidelines

Current 6.3 16 38.9 31.9 6.9 1.4* -17.45Desired 0.7 0 4.9 28 66.4 0.98 0.000

Online survey. Gaps between current level of knowledge of appropriate care in providing care for patients with OSA and desired knowledge, rated on a 5-point Likert scale (1 = Low, 5 = High). Gray shading indicates substantive gaps (Generalists > 1.00). *Statistically significant (p ≤ 0.001).


Sleep Specialists and Generalists: Obstacles to Referrals

sleep laboratory studies, they described difficulties in convinc-ing patients of the importance of such testing.

DISCUSSION

Sleep disorders—including both obstructive sleep apnea and shift work disorder—were unrecognized and/or not prioritized by generalists. Few generalists reported adequate knowledge in the area, characterizing care of patients with poor sleep as not urgent, frustrating, and de-motivating—attitudes that con-tributed to under- and misdiagnosis, under- and mistreatment, and stigmatization of patients with these low priority condi-tions. In 2002, 90% of generalists surveyed evaluated their knowledge of sleep disorders as fair or poor,9 as compared to 5.5% in the OSA study and 35.2% in SWD study. This study also demonstrated a shift in this knowledge gap, as only 20%

of generalists questioned sleep disorders as a real diagnosis, with large gaps identified not only by generalists but also by specialists. Findings described in the results of these two stud-ies suggest that there have been gains in the past seven years, yet much remains to be done in primary medical, specialty, and public education.

The role of specialists in sleep disorders care was questioned by generalists in these studies. This reiterates the examination of 69 qualified pulmonologists’ expertise carried out in 1998,19 which found poor performance when asked to evaluate non-pulmonary sleep disorder cases. At that time, chest physicians themselves ex-pressed a need for more formal training in sleep disorders. Current findings suggest that this remains a pressing need, with lack of qualified experts resulting in lack of interdisciplinary support for generalists in the care of their sleep disorders patients as well as generating lack of credibility for the sleep disorders specialty. Yet

Table 4—Gap analysis in care of patients with SWDGeneralist/PCP (n = 200)Specialists (n = 100)

GroupDesiredLevel

Percentage of Participants Gap t1

Low2 3 4 5

HighXSD p

DiagnosisAdministering and interpreting tests assessing patients with sleeping disorders

Generalists Current 38.9 25.8 25.7 8.6 1 1.7* -20.87Desired 10.2 7.1 18.8 26.9 37.1 1.1 0.000

Specialists

Current 7.1 11.2 17.3 24.5 39.8 0.85* -7.68Desired 0 4.3 6.4 16 73.4 1.1 0.000

Diagnosing shift work disorder Generalists Current 11.7 23.5 41.3 19.4 4.1 1.6* -22.06Desired 0.5 1 10.9 32.1 55.4 1.0 0.000

Specialists

Current 3.1 8.2 16.3 37.8 34.7 0.78* -8.10Desired 0 2.2 4.3 18.3 75.3 0.92 0.000

Differentiating between shift work disorder and depression

Generalists Current 5.6 20.8 35.5 29.9 8.1 1.4* -20.00Desired 0.5 0.5 6.6 32.7 59.7 0.95 0.000

Specialists

Current 1 8.2 33 37.1 20.6 1.1* -11.17Desired 0 0 4.3 15.1 80.6 0.93 0.000

TreatmentCurrent clinical guidelines for treating shift work disorder

Generalists Current 24.7 27.3 37.4 9.1 1.5 2.0* -29.29Desired 0.5 0 14.3 32.1 53.1 0.96 0.000

Specialists

Current 5.1 9.2 24.5 42.9 18.4 1.1* -11.85Desired 0 0 6.4 21.3 72.3 0.91 0.000

Balancing treatment for shift work disorder with treatment for other non-sleep related comorbidities (e.g., diabetes or osteoporosis)


Specialists Current 6.1 10.2 41.8 29.6 12.2 1.2* -11.64Desired 0 3.2 9.6 26.6 60.6 0.98 0.000

Providing self-management education to my patients with shift work disorder


Specialists

Current 2 11.2 35.7 30.6 20.4 1.1* -11.13Desired 0 0 9.7 23.7 66.7 0.91 0.000

Monitoring shift work disorder status and progression

Generalists Current 18.8 27.9 32 16.8 4.6 1.6* -20.89Desired 1.5 2.6 14.4 34 47.4 1.1 0.000

Specialists

Current 2 10.2 36.7 32.7 18.4 1.1* -11.90Desired 0 0 8.6 22.6 68.8 0.88 0.000

Online survey. Gaps between current level of knowledge of appropriate care in providing care for patients with SWD and desired knowledge, rated on a 5-point Likert scale (1 = Low, 5 = High). *Statistically significant (p ≤ 0.001).



the value of this approach is not evident to the generalists who are often the first contact of patients with sleep disorders.

There is evidence of positive impact of a functional inter-professional healthcare team upon patient healthcare outcomes and system processes.20 Inter-professional collaboration (IPC) requires shared purpose, responsibility and goals, coordinated

efforts, interdependency, and recognition of the value of each team members and commitment to the value of team-based care.21,22 The Institute of Medicine 2003 report on education in the health professions23 identifies the need for cooperation, communication, and integration of healthcare. Inter-profes-sional education provides promise for development of such interdisciplinary care.24,25 Findings of these studies suggest the fundamental importance of addressing issues around interde-pendency and recognition of inter-professional value.

The importance of addressing the perception of sleep medi-cine is recognized within the sleep medicine community, in terms of credibility in demonstrating added value of sleep medicine within the larger medical community, and in terms of increasingly competitive and scarce funding for graduate medical education positions in sleep medicine.26 The Adult Obstructive Sleep Apnea Task Force of the American Acad-emy of Sleep Medicine recommended a multidisciplinary approach to the care of OSA, including primary and special-ist care, as well as other sleep resources.13 The American Thoracic Society has identified core competencies to guide sleep disorder competencies in pulmonary fellowship train-ing programs. The American Academy of Sleep Medicine has identified the importance of addressing these challenges in the medical community as a whole, to better establish the subspe-cialty of sleep medicine, particularly in the face of challenges posed by sleep therapists who might further impact the cred-ibility of sleep physicians.27

Performance improvement initiatives are required that will address not only knowledge and skill in diagnosing and manag-ing sleep disorders, but also attitudinal and team issues. Such initiatives would need to be clinically based in order to result in translation of information into actual practice, and would need to be iterative in order to continue to build and improve care of patients with sleep disorders.

• Challenges have been identified in the interdisciplinary care of patients with sleep disorders

• Educational programs on sleep medicine should be de-signed and provided to primary care generalist physi-

Table 5—Referral of patients with sleep disorders

ReferralGroup

DesiredLevel

Percentage of Participants Gap t1

Low2 3 4 5

HighXSD p

Determining when to refer patient back to generalist

Specialist (SWD)

Current 2.3 4.6 23 39.1 31 0.74 -8.53Desired 0 1.2 7.1 20.2 71.4 0.79* 0.000

Determining which patients with SWD should be referred to a specialist

Generalists (SWD)

Current 14.1 23.7 31.8 20.2 10.1 1.5 -16.90Desired 0.5 2.5 10.7 32 54.3 1.2* 0.000

Identifying which type of specialist I should refer my patient with SWD

Generalists (SWD)

Current 10.7 20.3 30.5 23.4 15.2 1.3 -14.76Desired 0.5 2 9.7 27 60.7 1.3* 0.000

Determining which patients with potential OSA should be referred to a specialist

Generalists (OSA)

Current 1.4 7.6 24.3 45.8 20.8 0.86 -11.32Desired 0 0.7 4.2 27.3 67.8 0.91* 0.000

Identifying which type of specialists I should refer my patients with OSA

Generalists (OSA)

Current 0.7 8.3 17.4 41.7 31.9 0.68 -8.52Desired 0 1.4 4.2 23.2 71.1 0.96* 0.000

Gap analysis. Gaps between current level of knowledge related to referring patients with sleep disorders to sleep specialists and desired knowledge, rated on a 5-point Likert scale. Gray shading indicates substantive gaps (Generalists > 1.00, Specialists > 0.75). *Statistically significant (p ≤ 0.001).

A

B

35%30%25%20%15%10%5%0%

35%30%25%20%15%10%5%0%

1 2 3 4 5Not at all A major barrier

1 2 3 4 5Not at all A major barrier

13%

11%

21%23%

23%20%

30%

35%

30%

29%

33%

26%

27%

32%

3%

5%

9%

7%

11%

12%

PCPs (n = 137) Specialists (n = 97)

PCPs (n = 136) Specialists (n = 98)

Figure 1—Generalist-specialist barriers to coordinated, interprofessional care

(A) Lack of role clarity as a barrier. (B) Lack of coordination as a barrier.



cians. These should include information about when referral to sleep specialists may be of benefit and the rea-sons for this.

• Sleep medicine training programs should include methods of integrating care with and providing useful consultation to primary care generalists.

Table 6—Excerpt of questionnaire on obstructive sleep apnea completed by generalistsUsing the scale provided (where 1 = Low and 5 = High), please indicate your current and desired level of knowledge concerning each issue in sleeping disorders.

My CURRENT Level of Knowledge

My DESIRED Level of Knowledge

Low High Low HighAdminister and interpret tests correctly assessing patients with sleeping disorders 1 2 3 4 5 1 2 3 4 5Identify co-morbid conditions in patients with sleeping disorders 1 2 3 4 5 1 2 3 4 5Differentiate between OSA and insomnia 1 2 3 4 5 1 2 3 4 5Diagnose insomnia 1 2 3 4 5 1 2 3 4 5Diagnose OSA 1 2 3 4 5 1 2 3 4 5Diagnose narcolepsy 1 2 3 4 5 1 2 3 4 5Diagnose circadian rhythm sleep disorders 1 2 3 4 5 1 2 3 4 5Diagnose restless leg syndrome 1 2 3 4 5 1 2 3 4 5Diagnose sleep issues secondary to mental health issues 1 2 3 4 5 1 2 3 4 5Diagnose sleep issues secondary to physical conditions 1 2 3 4 5 1 2 3 4 5Diagnose substance-induced sleep issues 1 2 3 4 5 1 2 3 4 5Determine which patients with potential OSA should be referred to a specialist 1 2 3 4 5 1 2 3 4 5Identify to which type of specialist I should refer my patient with OSA 1 2 3 4 5 1 2 3 4 5Initiate CPAP treatment for OSA 1 2 3 4 5 1 2 3 4 5Initiate and manage sedative pharmacological treatment for SD 1 2 3 4 5 1 2 3 4 5Initiate and manage stimulant pharmacological treatment for SD 1 2 3 4 5 1 2 3 4 5Evaluate and manage potential drug interactions in SD 1 2 3 4 5 1 2 3 4 5Provide lifestyle changes recommendation for patients with OSA 1 2 3 4 5 1 2 3 4 5Monitor OSA status and progression 1 2 3 4 5 1 2 3 4 5Manage patients with OSA according to current clinical guidelines 1 2 3 4 5 1 2 3 4 5

Using the scale provided (where 1 = Not a barrier at all and 5 = A major barrier, and N/A = Not applicable to my practice), please indicate to what extent you think each of the following is a barrier for you in seeking to provide optimal care to patients with OSA.

Not a barrier at all

A major barrier

Not applicable

Sleeping disorders assessment is not a priority during regular physical exams 1 2 3 4 5 N/APatients lack of awareness and tendency to ignore sleeping problems 1 2 3 4 5 N/AFear of opening up a “can or worms” when asking questions about sleep to patients 1 2 3 4 5 N/ADifficulty asking targeted questions to accurately assess patients regarding OSA 1 2 3 4 5 N/ADifferentiating between OSA and insomnia 1 2 3 4 5 N/AConcerns over legal repercussions when diagnosing patients with sleeping disorders 1 2 3 4 5 N/APatient refusal to accept diagnosis of sleeping disorders 1 2 3 4 5 N/AIdentifying and ruling out alternative underlying conditions for sleep issues 1 2 3 4 5 N/ALack of quality standardization of sleep labs 1 2 3 4 5 N/ACost of medication and patient’s lack of insurance coverage 1 2 3 4 5 N/APatients not complying with the CPAP therapy 1 2 3 4 5 N/AGetting patient to change their lifestyles 1 2 3 4 5 N/AConcerns over addiction issues related to sleep medication 1 2 3 4 5 N/AProviding education and self-management skills to my patients with SD 1 2 3 4 5 N/ALack of tools to monitor sleeping disorder progression and response to treatment 1 2 3 4 5 N/ALack of clarity of the roles and responsibilities of PCPs dealing with sleeping disorders 1 2 3 4 5 N/ACurrent clinical guidelines in SD not relevant nor useful to PCPs 1 2 3 4 5 N/A

Demographic questions and questions relating to educational format preferences have been removed. Survey was deployed online, therefore format presented does not reflect actual.



• Since sleep disorders such as OSA and SWD are chronic diseases requiring long-term follow-up and management, the importance of continuing care must be part of the educa-tional process for patients, generalists and sleep specialists.

• Performance improvement initiatives that address attitudes and team roles and responsibilities are needed.

• Development of a recognized, credible, competent, spe-cialist pool is required to support excellent primary care.

LimitationsOur study has several limitations. Results are based on self-

report, introducing the possibility of bias due to erroneous self-assessment. However, the objective of this research was to assess subjects’ perceptions of gaps, barriers, and attitudes, which can only be gathered through self-report. In this study, triangulation of findings across two disorders, two subject groups, focus groups, interviews, and survey data was used to strengthen the trustworthiness of the findings. In addition, it is possible that those who participated in our surveys may differ systematically from those who did not respond.

CONCLUSION

Knowledge, skill, and attitudinal challenges and gaps have been identified in primary care of patients with sleep disorders, resulting in sleep disorders being underdiagnosed, undertreat-ed, stigmatized, and under-prioritized. Challenges in under-standing and incorporation of the interdisciplinary team, and in

particular, enhancing the role and value of the sleep specialist to the primary care community, for optimal care were also identi-fied, impeding the contribution of sleep specialists to patient outcomes. Performance improvement initiatives are needed that address generalist knowledge and competence in providing care of patients with sleep disorders as well as competence in collaborating in an interdisciplinary team. Sleep medicine itself must also address gaps in its own training and practice as well as misconceptions of others.

REFERENCES1. Institute of Medicine of the National Academies. Washington: Institute of Medi-

cine; 2006. Sleep Disorders and Sleep Deprivation: An Unmet Public Health Problem. Available at: http://www.iom.edu/CMS/3740/23160/33668.aspx. Ac-cessed July 8, 2009.

2. Sleep in America National Sleep Foundation. Washington: National Sleep Foun-dation; 2008. Sleep in America Poll. Available at: http://www.sleepfoundation.org/site/c.huIXKjM0IxF/b.3933533/. Accessed July 8, 2009.

3. Banks S, Dinges DF. Behavioral and physiological consequences of sleep re-striction. J Clin Sleep Med 2007;3:519-28.

4. Hiestand DM, Goldman M, Phillips B. Prevalence of symptoms and risk of sleep apnea in the US population. Chest 2006;130:780-6.

5. Michigan Lung and Critical Care Specialists. Michigan: Michigan Lung and Criti-cal Care Specialists. Shift Work Sleep Disorder. Available at: http://omlcc.com/SWSD.html. Accessed February 6, 2009

6. American Academy of Sleep Medicine. The international classification of sleep disorders : diagnostic and coding manual (2nd ed). Westchester, IL: American Academy of Sleep Medicine; 2005.

7. Presser HB. Job, family, and gender: Determinants of nonstandard work sched-ules among employed Americans in 1991. Demography 1995;32:577-98.

Table 7 continues on the following page

Table 7—Excerpt of questionnaire on obstructive sleep apnea completed by generalists and specialistsIn the context of your clinical practice, how do you define shift work disorder (SWD)? _____________________________________________________________________________________

Using the scale provided (where 1 = I completely disagree, 5 = I completely agree, and N/A = Not applicable to my practice), please indicate to what extent you agree or disagree with the statements provided.

I completely

disagreeI somewhat

agree

I completely

agreeNot

applicableAssessing patients’ sleep is not a priority during regular physical exams 1 2 3 4 5 N/AI believe in shift work disorder as a medical condition 1 2 3 4 5 N/AI see it as my responsibility to treat and manage patients for shift work disorder 1 2 3 4 5 N/AI find it motivating to treat and manage patients with shift work disorder 1 2 3 4 5 N/AAchieving positive health outcomes for patients with shift work disorder is something I can control 1 2 3 4 5 N/AI think shift work disorder is a label used to promote pharmacotherapy 1 2 3 4 5 N/AI see value in patients with shift work disorder doing a sleep study 1 2 3 4 5 N/A

Who do you think should be involved in treating and managing patients with shift work disorder? (Please select all that apply.) ❑ Family Physician ❑ Internist ❑ Psychiatrist ❑ Psychologist ❑ Neurologist ❑ Pulmonologist ❑ Other (please specify): ________________________________________________

Who should be involved in providing care to patients with shift work disorder? Please select all that apply, for each activity.

Primary Care Physician

Sleep Specialist

(physician)

Other Allied Health

Provider Not ApplicableInitiating testing for suspected shift work disorder ❑ ❑ ❑ ❑Diagnosing shift work disorder ❑ ❑ ❑ ❑Recommending non-pharmacological treatment for shift work disorder (e.g. sleep hygiene measures) ❑ ❑ ❑ ❑Recommending pharmacological treatment for shift work disorder ❑ ❑ ❑ ❑Managing shift-work disorder patients long term ❑ ❑ ❑ ❑Providing counseling for shift work disorder ❑ ❑ ❑ ❑Providing patient education about shift work disorder ❑ ❑ ❑ ❑

To whom do you refer patients with shift work disorder? (Please select all that apply.) ❑ I do not refer patients with shift work disorder ❑ Family Physician ❑ Internist ❑ Psychiatrist ❑ Psychologist ❑ Neurologist ❑ Pulmonologist ❑ Other (please specify): _______________________________________________

Demographic questions and questions relating to educational format preferences have been removed. Survey was deployed online, therefore format presented does not reflect actual. Not all questions were answered by specialists and generalists.



8. Drake CL, Roehrs T, Richardson G, Walsh JK, Roth T. Shift work sleep disorder: prevalence and consequences beyond that of symptomatic day workers. Sleep 2004;27:1453-62.

9. Papp KK, Penrod CE, Strohl KP. Knowledge and attitudes of primary care physi-cians toward sleep and sleep disorders. Sleep Breath 2002;6:103-9.

10. Owens JA. The practice of pediatric sleep medicine: results of a community sur-vey. Pediatrics 2001;108:E51.

11. Kryger MH, Walid R, Manfreda J. Diagnosis received by narcolepsy patients in the year prior to diagnosis by a sleep specialist. Sleep 2002;25:36-41.

12. Lavie P. Sleep medicine – time for a change. J Clin Sleep Med 2006;2:207-11.13. Epstein LJ, Kristo D, Strollo PJ Jr, et al; Adult Obstructive Sleep Apnea Task

Force of the American Academy of Sleep Medicine. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med 2009;5:263-76.

14. Chatterji M. Evidence on “what works”: an argument for extended-term mixed method (ETMM) evaluation designs. Educ Res 2005;34:14-24.

15. Johnson RB, Onwuegbuzie AJ. Mixed methods research: a research paradigm whose time has come. Educ Res 2004;33:14-26.

16. Creswell JW. Research design: qualitative, quantitative, and mixed approaches. Thousand Oaks, CA: Sage Publications; 2003.

17. Caruso C, Waters TR. A review of work schedule issues and musculosketal dis-orders with an emphasis on the healthcare sector. Ind Health 2008;46;523-34.

18. Neuendorf KA. The content analysis guidebook. Thousand Oaks, CA: Sage; 2002.

19. Phillips B, Collop N, Goldberg R. Sleep medicine practices, training, and at-titudes: a wake-up call for pulmonologists. Chest 2000;117:1603-7.

20. Zwarenstein M, Goldman J, Reeves S. Interprofessional collaboration: effects of practice-based interventions on professional practice and healthcare outcomes. Cochrane Database Syst Rev 2009;(3):CD000072.

21. Johanson LS. Interprofesssional collaboration: Nurses on the team. Medsurg Nurs 2008;17:129-30.

Table 7 (continued )—Excerpt of questionnaire on obstructive sleep apnea completed by generalists and specialists

Please consider each issue on sleeping disorders listed in the table below. First, using the scale provided (where 1 = Low and 5 = High), select the number that best describes how you currently evaluate your level of knowledge concerning each issue. Next, please indicate your desired level of knowledge concerning each issue in sleeping disorders.

My CURRENT Level of Knowledge

My DESIRED Level of Knowledge

Low High Low HighCurrent clinical guidelines for treating shift work disorder 1 2 3 4 5 1 2 3 4 5Administering and interpreting tests assessing patients with sleeping disorders 1 2 3 4 5 1 2 3 4 5Diagnosing shift work disorder 1 2 3 4 5 1 2 3 4 5Differentiating between shift work disorder and other types of insomnia 1 2 3 4 5 1 2 3 4 5Differentiating between shift work disorder and obstructive sleep apnea 1 2 3 4 5 1 2 3 4 5Differentiating between shift work disorder and narcolepsy 1 2 3 4 5 1 2 3 4 5Differentiating between shift work disorder and depression 1 2 3 4 5 1 2 3 4 5Differentiating between shift work disorder and anxiety disorder 1 2 3 4 5 1 2 3 4 5Providing lifestyle change recommendations for patients with shift work disorder (e.g. sleep hygiene) 1 2 3 4 5 1 2 3 4 5Initiating and managing sedative pharmacological treatment for shift work disorder 1 2 3 4 5 1 2 3 4 5

Initiating and managing stimulant pharmacological treatment for shift work disorder 1 2 3 4 5 1 2 3 4 5

Initiating and managing anti-depressant pharmacological treatment for shift work disorder 1 2 3 4 5 1 2 3 4 5Balancing treatment for shift work disorder with treatment for other non-sleep related co-morbidities (e.g. diabetes or osteoporosis) 1 2 3 4 5 1 2 3 4 5Providing self-management education to my patients with shift work disorder 1 2 3 4 5 1 2 3 4 5Monitoring shift work disorder status and progression 1 2 3 4 5 1 2 3 4 5Determining which patients with shift work disorder should be referred to a specialist 1 2 3 4 5 1 2 3 4 5Identifying which type of specialist I should refer my patient with shift work disorder to 1 2 3 4 5 1 2 3 4 5Determining when to refer patients back to their primary care provider 1 2 3 4 5 1 2 3 4 5

Using the scale provided (where 1 = Not a barrier at all, 5 = A major barrier, and N/A = Not applicable to my practice), please indicate to what extent you think each of the following is a barrier for you in seeking to provide optimal care to patients with shift work disorder.

Not a barrier at all

A major barrier

Not applicable

Patients’ lack of awareness and tendency to ignore sleeping problems 1 2 3 4 5 N/AFear of opening up a “can of worms” when asking patients questions about sleep 1 2 3 4 5 N/ADifficulty asking targeted questions to accurately assess patients regarding shift work disorder 1 2 3 4 5 N/ADifficulty in defining shift work disorder 1 2 3 4 5 N/AConcerns over legal repercussions when diagnosing patients with shift work disorder 1 2 3 4 5 N/APatient lack of understanding of shift work disorder diagnosis 1 2 3 4 5 N/APatient refusal to accept diagnosis of shift work disorder 1 2 3 4 5 N/AIdentifying and ruling out alternative underlying conditions for sleep issues 1 2 3 4 5 N/ALack of objective criteria to diagnose shift work disorder 1 2 3 4 5 N/AMy skill in helping patients change their lifestyles 1 2 3 4 5 N/AAddiction issues related to sleep medication 1 2 3 4 5 N/AMy level of confidence in prescribing stimulants 1 2 3 4 5 N/AMy level of confidence in prescribing sedatives 1 2 3 4 5 N/AMy level of confidence in prescribing anti-depressants 1 2 3 4 5 N/AProviding self-management education to my patients with shift work disorder 1 2 3 4 5 N/ALack of tools to monitor shift work disorder progression and response to treatment 1 2 3 4 5 N/ACoordination between primary care providers and specialists for shift work disorder treatment and management 1 2 3 4 5 N/ALack of clarity of the roles and responsibilities of Primary Care Providers in dealing with shift work disorder 1 2 3 4 5 N/ARelevance of current clinical guidelines pertaining to shift work disorder 1 2 3 4 5 N/ACoding for reimbursement for shift work disorder 1 2 3 4 5 N/A

Demographic questions and questions relating to educational format preferences have been removed. Survey was deployed online, therefore format presented does not reflect actual. Not all questions were answered by specialists and generalists.



22. D’Amour D, Ferrada-Videla M, San Martin Rodriguez L, Beaulieu MD. The con-ceptual basis for interprofessional collaboration: core concepts and theoretical frameworks. J Interprof Care 2005;19 Suppl 1:116-31.

23. Grenier AC, Knebel (eds.). Health professions education: a bridge to quality. Committee on the Health Professions Education Summit Board on Health Care Services. Institute of Medicine. Washington, DC: National Academies Press, 2003. Available at http://www.nap.edu/openbook.php?record_id = 10681. Ac-cessed November 1, 2010.

24. D’Amour D, Oandasan I. Interprofessionality as the field of interprofessional practice and interprofessional education: an emerging concept. J Interprof Care 2005;19 Suppl 1:8-20.

25. Reeves S, Zwarenstein M, Goldman J, et al. Interprofessional education: effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2008 Jan 23;(1):CD002213.

26. Quan SF. Sleep medicine and graduate medical education – Prospects for the future. J Clin Sleep Med 2009;5:497.

27. Kushida CA. AASM President’s viewpoint: planning for a challenging yet promis-ing future. J Clin Sleep Med 2009;5:301-3.

ACKNOWLEDGMENTSWe would like to express our appreciation for the contributions of the physicians

and patients who participated in this study. We would like to thank Martin Dupuis, M.A., Biagina-Carla Farnesi, M.Sc., Genevieve Myhal, Ph.D., and Kayla Cytryn, R.N., Ph.D., Performance Optimization Associates, AXDEV Group, who were instrumen-tal in carrying out this research. We would also like to express our appreciation of Elaine Turner, Project Coordinator, also from AXDEV Group, and Colleen Connor,

Web/Database Coordinator, Office of Continuing Medical Education of the University of Virginia School of Medicine for their invaluable support. In addition, we would like to acknowledge the American Thoracic Society, the New Jersey Academy of Fam-ily Physicians, Office of Continuing Medical Education of the University of Virginia School of Medicine, and University of Wisconsin Office of Continuing Professional Development for their support and assistance in recruitment for the online surveys.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication September, 2011Submitted in final revised form April, 2012Accepted for publication May, 2012Address correspondence to: Richard J. Castriotta, M.D., Professor and Director, Division of Pulmonary and Sleep medicine, University of Texas Medical School at Houston, 6431 Fannin St., MSB 1.274, Houston, TX 77030; Tel: (713) 500-6823; Fax: (713) 500-6829; E-mail: [email protected]

DISCLOSURE STATEMENTThis study was funded by an unrestricted grant from Cephalon Inc. AXDEV Group

Inc. received two unrestricted educational contracts to carry out this research (2007, 2008). Dr. Castriotta has received research support from Cephalon, Inc. Dr. Malhotra has received a grant from Cephalon, Inc. Dr. Landrigan has received research sup-port from an unrestricted grant from Cephalon, Inc. The other authors have indicated no financial conflicts of interest.


Study Objectives: Sleep disturbances, including obstructive sleep apnea (OSA), commonly limit function and quality of life in people with spondyloarthritis (SpA). Systemic infl ammation has been implicated in the pathophysiology of both OSA and SpA, and suppression of infl ammation with tumor necrosis factor α (TNF) inhibitors may decrease OSA severity. In this study, we compared the frequency of OSA in patients receiving and not receiving TNF-inhibitor therapy.Methods: Data were collected from 63 consecutively screened veterans with SpA. Participant interviews, exami-nations, chart reviews, and referrals to the Salt Lake City Vet-eran Affairs (SLCVA) Sleep Center were used to obtain de-mographic data, comorbidities, SpA features, therapy data,

and sleep study outcomes.Results: OSA occurred in 76% of SpA patients. OSA was less common in patients receiving TNF-inhibitor therapy (57%), compared to patients not receiving TNF-inhibitor therapy (91%) (p = 0.01).Conclusions: OSA is underrecognized in veterans with SpA, and TNF-inhibition was associated with a lower frequency of OSA.Keywords: Obstructive sleep apnea, spondyloarthritis, TNF-inhibitorCitation: Walsh JA; Duffi n KC; Crim J; Clegg DO. Lower fre-quency of obstructive sleep apnea in spondyloarthritis patients taking TNF-inhibitors. J Clin Sleep Med 2012;8(6):643-648.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Normal sleep is a vital component of health and quality of life. Sleep disturbances negatively affect health, with el-

evated risks of neurocognitive dysfunction, coronary artery dis-ease, hypertension, depression, stroke, motor vehicle accidents, and premature death.1 Sleep disturbances and fatigue are promi-nent features of SpA,2,3 and infl ammatory arthritis patients with sleep inadequacies experience more pain and functional limita-tions than patients with healthy sleep patterns.4

The contribution of OSA to sleep disturbances in SpA is not well characterized. Two small uncontrolled studies reported OSA prevalences of 12% and 23% in ankylosing spondylitis (AS) patients, compared to 1% to 9% in the general popula-tion.5,6 Psoriatic arthritis (PsA) has been associated with sleep interference independent of body mass index (BMI),7 and a higher prevalence of OSA has been reported in psoriasis pa-tients compared to controls.8

Various mechanisms have been proposed to explain the high prevalence of OSA in SpA patients including (1) glucocorticoid-induced weight gain, (2) compression of the oropharyngeal airway by the cervical spine, and (3) systemic infl ammation.5,9 Glucocorticoids are selectively used to treat peripheral disease manifestations of SpA. Glucocorticoids cause central obesity,10

and glucocorticoid-induced deposition of fat in the thorax and neck may increase the risk of OSA in SpA patients.5,11

The second hypothesis proposes that SpA patients are at in-creased risk for OSA because of structural changes in the cervical spine. SpA patients frequently develop pathologic boney bridging between vertebrae called syndesmophytes. Syndesmophytes in the anterior cervical spine may contribute to OSA by compressing the oropharyngeal airway.5 This hypothesis is supported by case

Lower Frequency of Obstructive Sleep Apnea in Spondyloarthritis Patients Taking TNF-Inhibitors

Jessica A. Walsh, M.D.1; Kristina Callis Duffi n, M.D.2; Julia Crim, M.D.3; Daniel O. Clegg, M.D.1

1George E. Wahlen Veteran Affairs Medical Center, University of Utah, Division of Rheumatology, Salt Lake City, UT; 2Department of Dermatology, University of Utah, Salt Lake City, UT; 3Department of Radiology, University of Utah, Salt Lake City, UT

reports of improvement in OSA after surgical removal of exces-sively bulky pathologic bone from the anterior cervical spine.12

The hypothesis that systemic infl ammation contributes to OSA risk is particularly intriguing because shared infl ammato-ry pathways contribute to the pathophysiology of both SpA and OSA. OSA is characterized by repetitive airway obstructions causing intermittent hypoxia during sleep. Nuclear factor-κB is activated by intermittent hypoxia and upregulates the transcrip-tion of pro-infl ammatory genes.13 The pro-infl ammatory genes encode multiple infl ammatory cytokines, including TNF-α. El-evated levels of TNF-α have been reported in OSA patients, independent of obesity, and TNF-α levels fall with continuous positive airway pressure (CPAP) therapy for OSA.14,15 TNF-α is a well-recognized mediator of infl ammation in SpA, and inhi-bition of TNF-α improves symptoms of SpA, including fatigue and sleepiness.2,16,17

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Sleep disturbances and fatigue are common in patients with systemic infl ammatory conditions, includ-ing spondyloarthritis. The purposes of this study were to evaluate the frequency of obstructive sleep apnea (OSA) in spondyloarthritis pa-tients and to explore associations between TNF-inhibitor therapy and OSA severity.Study Impact: The high frequency of OSA demonstrated that routine screening for OSA risks should be considered in spondyloarthritis pa-tients. The association between TNF-inhibitor therapy and a lower fre-quency of OSA suggests that further investigation of TNF-inhibition may contribute to our understanding of OSA pathology and the development of novel therapies for OSA.


JA Walsh, KC Duffin, J Crim et al

To explore the hypothesis that suppression of inflammation with TNF-inhibition improves sleep outcomes, Vgontzas et al. treated eight obese OSA patients with etanercept.18 They report-ed improvements in the apnea hypopnea index (AHI) (p < 0.05) and sleepiness (p < 0.05) after three weeks of therapy. Since inflammatory cytokine levels are higher in inflammatory arthri-tis patients than obese patients,19,20 we hypothesize that TNF-in-hibitors may have a greater effect on OSA in SpA patients than obese patients. In this study, we explored the impact of TNF-inhibition on OSA, by comparing OSA frequencies in SpA pa-tients receiving and not receiving TNF-inhibitor therapy.

METHODS

Between July 2008 and July 2010, 63 consecutively en-countered patients enrolled in the Program to Understand the Longterm outcomes of SpondyloARthrits (PULSAR) regis-try from the SLCVA Rheumatology Clinics were screened for this study. Inclusion criteria were participation in PULSAR and willingness to participate in this investigation. According to established practice at the SLCVA, current cigarette smok-ers were excluded because smoking confounds sleep testing and interpretation. Patients were excluded from analyses if they had sleep studies that were unavailable, incomplete, or inconsistent with the SLCVA Sleep Center’s reporting of sleep apnea parameters.

In order to qualify for a sleep study, the SLCVA Sleep Center required patients to have at least one symptom of OSA. SpA patients were evaluated for symptoms of OSA by the principal investigator at the time of clinic visits. The symptom evaluations included questions about snoring, excessive daytime sleepiness, non-restorative sleep, cessation in breathing while sleeping, awakening gasping for air, restlessness at night, difficulty sleep-ing with frequent awakenings, concentration difficulties, fall-ing asleep while driving, changes in mood, morning headaches, and vivid, strange, or threatening dreams. All SpA patients had

≥ 1 symptom of OSA and were offered a referral to the SLCVA Sleep Center for standard-of-care testing and therapy.

The assignment of home monitoring versus laboratory poly-somnography (PSG) was determined by the SLCVA Sleep Cen-ter coordinators and was based on patient comorbidities and the availability of home monitoring devices. Home monitoring tests were conducted for 3 nights with 4-channel NovaSom QSG Di-agnostic devices that measured airflow, oximetry, pulse, and respiratory effort. PSG testing was completed in sleep laborato-ries located in Utah and surrounding states. Apnea events were defined as cessation of airflow ≥ 10 seconds. Hypopnea events were defined as a decrease in airflow ≥ 50% and a decline in oxyhemoglobin saturation of ≥ 4% for ≥ 10 seconds. Epworth Sleepiness Scale (ESS) questionnaires were administered to pa-tients by the SLCVA Sleep Center.

Data including demographics, comorbidities, SpA features, SpA activity, and therapies were obtained from patient inter-views, examinations, and medical records. The Bath Ankylos-ing Spondylitis Disease Activity Index (BASDAI) and Bath Ankylosing Spondylitis Functional Index (BASFI) are validat-ed patient questionnaires that were used to measure SpA dis-ease activity and severity. The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are laboratory inflamma-tory markers commonly used for SpA disease activity assess-ments. BASFI, BASDAI, ESR, and CRP data were included only if they were completed within 2 months of the sleep study date. Treatment information was collected from clinic visit notes and verified via the SLCVA’s pharmacy medication dis-pensing records.

Available cervical spine radiographs were evaluated by a board-certified musculoskeletal radiologist. The scoring system from the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) was applied to quantify radiographic changes in the cervical spine consistent with SpA.21 The number of syndesmo-phytes in the cervical spine was also recorded. The maximum syndesmophyte size was determined by measuring the distance between the most anterior portion of the largest syndesmophyte and the vertical plane of the adjacent anterior vertebral bodies, on the lateral projection of cervical radiographs.

Continuous variables were compared with 2-tailed Student t-tests. Categorical variables were compared with Fisher exact tests. Spearman correlation coefficients were used to compare AHI to SpA features and activity measures, including cervical mSASSS, number of syndesmophytes, maximum syndesmoph-yte size, BASDAI, BASFI, ESR, and CRP.

RESULTS

Sixty-three consecutively encountered SpA patients were screened for participation (Figure 1); 45 patients were included in the analyses. Thirty-six participants were tested with home monitoring devices, and 9 participants received PSG in sleep laboratories. Informed consent was obtained from all partici-pants, and approval was granted by the Institutional Review Board at the University of Utah.

Eighteen patients were excluded: 4 patients declined par-ticipation, and 4 were excluded because of current smoking. Four additional patients with a diagnosis of sleep disordered breathing were excluded because they did not have clinical in-

18 excluded• Declined sleep study n = 4• Current smoking n = 4• AHI records unavailable or

incompatible n = 4• Incomplete sleep test n = 6

1 excluded because therapy data unclear

44 included in therapy analyses

63 consecutive SpA patients screened

45 included in OSA frequency analysis

Home studies, 3 nights n = 36Laboratory PSG n = 9

Figure 1—Study profile

SpA, spondyloarthritis; OSA, obstructive sleep apnea; PSG, polysomnography; AHI, apnea hypopnea index.


OSA in SpA patients with TNF-Inhibitorsdications for retesting, and medical records were unavailable or inconsistent with the SLCVA Sleep Center’s reporting of OSA parameters. Six excluded patients had incomplete home sleep studies and were unwilling or unable to repeat sleep testing. No patient was excluded because of lack of OSA symptoms. One patient was excluded from therapy analyses because it was unclear if a TNF-inhibitor was used when the sleep study was completed. A comparison of patients included and excluded from this investigation demonstrated no differences in demo-graphics, comorbidities, SpA features, SpA disease activity, TNF-inhibitors, or other SpA therapies.

The mean age of participants was 59.7 ± 11.8, and 43 (96%) were male. (table 1) The mean BMI was 29.2 ± 5.2. Thirty-one (69%) participants had hypertension, 4 (9%) had ischemic car-diomyopathy, and 3 (7%) had a history of a stroke. Radiograph-ic measures of cervical spine disease included a mean cervical mSASSS score of 8.5 ± 8.4, a mean number of syndesmophytes of 2.8 ± 2.0, and a maximum syndesmophyte size mean of 2.6 ± 1.4. The mean BASDAI and BASFI scores were 4.5 ± 2.5 and 4.3 ± 2.6, respectively. The mean ESR was 14.6 ± 14.2 mm/h, and the mean CRP was 10.0 ± 10.8 mg/L.

Thirty-four (76%) participants had an AHI ≥ 5. The mean AHI was 17.7 ± 16.4. The AHI was ≥ 5 in 18 (80%) AS patients, 14 (70%) PsA patients, and 2 (100%) enteric arthritis (EA) pa-tients. The mean AHI of participants tested with home monitor-ing devices was not statistically different than the mean AHI of participants tested with laboratory PSG (16.7 ± 17.6 vs. 20.9 ± 13.3, respectively, p = 0.47).

Therapy data were available in 44 of the 45 participants. Twenty-one (48%) were receiving TNF-inhibitor therapy at the time of sleep testing. Non-biologic disease modifying anti-rheumatic drugs (DMARDs) included methotrexate, hy-droxychloroquine, sulfasalazine, and leflunomide. Sixteen (36%) patients received non-biologic DMARDs, and 8 (18%) were on a combination of a TNF-inhibitor and non-biologic DMARD(s). Glucocorticoids were not used by any participant within 12 months prior to sleep testing.

In comparing participants with and without TNF-inhibitor therapy, the demographics, SpA subtype distribution, and co-morbidities were similar (table 2). Radiographic disease of the cervical spine, as measured by the cervical mSASSS, number of syndesmophytes, and maximum syndesmophyte size, was similar in participants receiving and not receiving TNF-inhib-itors. The BASDAI and BASFI disease activity scores were lower in the group receiving TNF-inhibitor therapy than the group without TNF-inhibitor therapy, but the differences were not statistically significant. There were no differences in ESR or CRP between groups.

Participants using TNF-inhibitor therapy had OSA less frequently than participants without TNF-inhibitor therapy (12 [57%] vs. 21 [91%], p = 0.01). Compared to participants taking TNF-inhibitors, participants without TNF-inhibition had more abnormal OSA outcomes, including a higher proportion of participants with AHI ≥ 15, a higher mean AHI, a higher percent time with oxygen saturation < 90%, and a higher mean ESS score, but the differences were not statistically significant. The allocation of home monitoring devices and PSG testing was similar in the TNF-inhibitor and no TNF-inhibitor groups (p = 1.00). Similar proportions of participants with and without

TNF-inhibition were taking non-biologic DMARDs at the time of sleep testing.

Correlation coefficients were calculated between AHI and SpA features and between AHI and SpA activity measures in all participants, regardless of TNF-inhibitor therapy (table 3). There were moderate correlations between AHI and BASDAI (R = 0.49) and between AHI and BASFI (R = 0.47). The ESR, CRP, cervical mSASSS, number of syndesmophytes, and maximum syndesmophyte size did not significantly correlate with AHI.

DISCUSSION

We found that OSA was less frequent in participants re-ceiving TNF-inhibitor therapy, compared to participants not receiving TNF-inhibitor therapy. Additionally, there were non-significant trends toward less severe OSA in the TNF-inhibitor group, with respect to AHI, oxygen saturation, and sleepiness.

Table 1—Characteristics of 45 participantsNo. ± SD or (%) n

DemographicsAge (mean) 59.7 ± 11.8 45Male 43 (96) 45BMI (mean) 29.2 ± 5.2 45

ComorbiditiesHypertension 31 (69) 45Ischemic CM 4 (9) 45Stroke 3 (7) 45

SpA features and activityCervical mSASSS 8.5 ± 8.4 36Number of syn 2.8 ± 2.0 36Max syn size (mm) 2.6 ± 1.4 36BASDAI 4.5 ± 2.5 25BASFI 4.3 ± 2.6 25ESR (mm/h) 14.6 ± 14.2 22CRP (mg/L) 10.0 ± 10.8 20

OSA outcomesAHI ≥ 5, all 34 (76) 45AHI (mean), all 17.7 ± 16.4 45AHI ≥ 5, AS 18 (80) 23AHI ≥ 5, PsA 14 (70) 20AHI ≥ 5, EA 2 (100) 2AHI (mean), home test 16.7 ± 17.6 36AHI (mean), PSG 20.9 ± 13.3 9

TherapyTNF-i 21 (48) 44Non-biologic DMARD 16 (36) 44TNF-i + non-biologic 8 (18) 44Glucocorticoids 0 44

SpA, spondyloarthritis; No., number of participants; SD, standard deviation; n, sample size; BMI, body mass index; CM, cardiomyopathy; mSASSS, modified Stoke Ankylosing Spondylitis Spinal Score; Max, maximum; syn, syndesmophyte; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; OSA, obstructive sleep apnea; AHI, apnea hypopnea index; AS, ankylosing spondylitis; PsA, psoriatic arthritis; EA, enteric arthritis; TNF-i, tumor necrosis factor α inhibitor; DMARD, disease modifying anti-rheumatic drug.


JA Walsh, KC Duffin, J Crim et al

These data suggest that TNF-inhibition may protect against the development of OSA and decrease OSA severity. The improve-ment in AHI and sleepiness in obese patients treated with a TNF-inhibitor reported by Vgontzas et al. also suggested that TNF-inhibition may reduce OSA severity and sleepiness.18 These findings are conceptually supported by the growing body of evidence implicating TNF-α and other inflammatory media-tors in the pathophysiology of OSA.13-15

The mechanistic relationship between inflammation and hypoxia in OSA is unclear. It has been proposed that intermit-

tent hypoxia stimulates inflammatory cytokine production, and resultant inflammation contributes to OSA comorbidities, such as cardiovascular disease.22 This theory suggests that suppression of inflammation may improve comorbidity-relat-ed outcomes for OSA patients, but fails to explain the possi-ble protective effects of TNF-inhibition on OSA development and severity. Research in children and adults with OSA has demonstrated that intranasal steroids improve AHI, presum-ably by diminishing inflammation in hypertrophied lymphoid tissue adjacent to airways.23,24 An alternative hypothesis is that alterations of cholinergic activity with TNF-inhibition may reduce airway collapse during sleep. Acetylcholine in-creases muscle contraction and sympathetic autonomic ac-tivity. It is also involved in the regulation of inflammatory cytokines, including TNF-α.25,26 TNF-inhibition may alter the balance of acetylcholine and inflammatory cytokines. These changes may positively influence airway musculature and re-duce hypoxic events in OSA patients. A pilot study, reporting reductions in AHI after therapy with a cholinesterase inhibi-tor, supports this hypothesis.27

Since systemic inflammation is suspected to contribute to the severity of OSA, we hypothesized that AHI would correlate with measures of SpA disease activity. BASDAI and BASFI scores had moderate correlations with AHI, suggesting that higher burdens of inflammation are positively associated with OSA severity. In contrast, ESR and CRP did not correlate with AHI. This was unexpected, since CRP levels have been demon-strated to correlate with both OSA and SpA disease activity.28 However, other uncontrolled factors may have influenced ESR and CRP levels, including non-biologic DMARD use. Several non-biologic DMARDs were used by participants, and these DMARDS are known to affect ESR and CRP levels.

We also explored the hypothesis that structural changes in the cervical spine may increase the risk of OSA in SpA patients. Our data did not support this premise, with no correlations be-tween AHI and any measure of radiographic disease of the cer-vical spine. We were unable to comment on the relationship between glucocorticoid use and OSA because glucocorticoids were not used by any participant. The lack of support for these alternative hypotheses strengthens our position that suppression

Table 2—Characteristics of 44 SpA patients with and without TNF-I

TNF-i No TNF-iNo. ± SD

or % nNo. ± SD

or % n pDemographics

Age (mean) 58.7 ± 11.7 21 60.6 ± 11.9 23 0.59Male 21 (100) 21 21 (91) 23 0.49BMI (mean) 29.2 ± 5.5 21 29.5 ± 4.8 23 0.84

SpA subtypePsA 12 (57) 21 11 (48) 23 0.56AS 9 (43) 21 11 (48) 23 0.77EA 0 21 1 (4) 23 1.00

ComorbiditiesHypertension 17 (81) 21 14 (61) 23 0.19Ischemic CM 3 (14) 21 1 (4) 23 0.60Stroke 3 (14) 21 0 23 0.10

SpA features and activityCervical mSASSS

8.2 ± 7.3 18 8.9 ± 9.5 18 0.80

Number of syn 2.7 ± 2.0 18 2.8 ± 2.1 18 0.94Max syn size (mm)

2.7 ± 1.5 18 2.3 ± 1.4 18 0.36

BASDAI 3.9 ± 2.5 13 5.1 ± 2.7 11 0.29BASFI 3.6 ± 2.2 13 4.8 ± 2.6 11 0.26ESR (mm/h) 14.1 ± 15.1 13 15.5 ± 13.5 8 0.83CRP (mg/L) 10.0 ± 12.1 11 10.1 ± 9.6 8 0.98

OSAAHI ≥ 5 12 (57) 21 21 (91) 23 0.01AHI ≥ 15 8 (38) 21 12 (52) 23 0.38AHI (mean) 13.7 ± 13.5 21 21.5 ± 18.0 23 0.12% time O2 < 90% 18 ± 28 21 33 ± 32 19 0.11ESS 7.3 ± 3.8 19 10.2 ± 5.2 17 0.07Home testing 17 (80) 21 18 (78) 23 1.00PSG testing 4 (19) 21 5 (22) 23 1.00

TherapyNon-biologic DMARD

9 (43) 21 7 (30) 23 0.53

SpA, spondyloarthritis; TNF-i, tumor necrosis factor α inhibitor; No., number of patients; SD, standard deviation; n, sample size; BMI, body mass index; PsA, psoriatic arthritis; AS, ankylosing spondylitis; EA, enteric arthritis; CM, cardiomyopathy; mSASSS, modified Stoke Ankylosing Spondylitis Spinal Score; Max, maximum; syn, syndesmophyte; BASFI, Bath Ankylosing Spondylitis Functional Index; BASDAI, Bath Ankylosing Spondylitis Disease Activity Instrument; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; OSA, obstructive sleep apnea; PSG, polysomnography; O2, oxygen saturation; ESS, Epworth Sleepiness Scale; DMARD, disease modifying anti-rheumatic drug.

Table 3—Correlations between AHI and SpA features andactivity measures in all participants

Correlation coefficient (R) n p

BASDAI 0.49 25 0.01BASFI 0.47 25 0.02ESR (mm/hr) 0.15 21 0.53CRP (mg/L) 0.11 19 0.65Cervical mSASSS 0.04 36 0.83Number of syn - 0.07 36 0.69Max syn size (mm) 0.15 36 0.36

AHI, apnea hypopnea index; SpA, spondyloarthritis; n, sample size; BASFI, Bath Ankylosing Spondylitis Functional Index; BASDAI, Bath Ankylosing Spondylitis Disease Activity Instrument; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; mSASSS, modified Stoke Ankylosing Spondylitis Spinal Score; Max., maximum; syn, syndesmophyte.


OSA in SpA patients with TNF-Inhibitorsof inflammation contributed to the lower frequency of OSA in participants on TNF-inhibitor therapy.

The markedly elevated frequency of OSA in veterans with SpA suggests that the scope of this medical condition may be greater than previously recognized. The 80% frequency of OSA in the 20 veterans with AS is much higher than the 12% and 23% prevalences reported in AS patients in German and Turk-ish studies.5,6 The older mean age of 60 in this study, compared to 47 and 33 in the respective aforementioned studies, may ac-count for some of the differences in OSA prevalence. Age was established as a risk factor for OSA in a study of 741 men, in which the OSA prevalence was 24.8% in men ≥ 65 years old compared to 7.9% in men 20-44 years old, when OSA was de-fined as AHI ≥ 5, regardless of symptoms.29 Since body weight is the strongest predictor of OSA,1 the higher mean BMI of 29.3 in this study, compared to 26.4 and 24.5 in the German and Turkish studies, also likely contributed to the high OSA fre-quency in this investigation.

There may have been additional characteristics affecting the risk for OSA in this veteran population that differed from the German and Turkish populations. The prevalence of OSA in 4,060,504 U.S. veterans according to ICD-9 codes was 2.9%,30 but these patients likely represent only the most symptomatic veterans. The prevalence of OSA was 27% among 26 randomly selected veteran inpatients, when OSA was defined by > 30 epi-sodes of apnea per night.31 A matched control group of outpa-tient veterans without SpA would be required to quantify the differential risk for OSA with and without SpA. However, the very high frequency of OSA in this population suggests that providers should consider routine screening for OSA risks in SpA patients.

Enhanced awareness of the elevated risk of OSA in SpA will create opportunities to improve patients’ quality of life and potentially survival. In a study evaluating the relative impor-tance of disease features, approximately 50% of AS and PsA patients ranked fatigue among their top 3 priorities for disease improvement.3 OSA is strongly associated with fatigue, and CPAP therapy improves energy levels and quality of life.23,32,33 Importantly, long-term CPAP is also associated with decreased mortality.34,35

Despite the benefits of CPAP, approximately 30% of OSA patients stop CPAP within 10 years, primarily because of in-tolerance.36 The only available nonsurgical alternatives involve oral and nasal appliances, which are generally less effective.37,38 OSA patients would benefit from additional therapeutic op-tions, and improving our understanding of TNF-inhibition in OSA may lead to the development of novel therapies.

This study is limited by the small sample size. An additional limitation involves the use of different types of sleep apnea test-ing. While PSG in a sleep laboratory is the gold standard, porta-ble home monitoring devices are increasingly used to improve access and reduce cost. Comparisons of home monitoring to laboratory PSG have demonstrated strong correlations between the diagnostic results of these testing modalities.39 Additionally, equivalent outcomes have been reported in patients tested with home studies and in-laboratory PSG.40

In conclusion,this investigation demonstrated a high fre-quency of OSA in veterans with SpA and suggests that OSA is underrecognized in this population. SpA patients currently

treated with TNF-inhibitors had OSA less frequently than un-treated patients. Further investigation of TNF-inhibition in OSA may provide mechanistic insights into the pathogenesis of OSA,and TNF-inhibition should be explored as an alternative therapy to CPAP.

REFERENCES1. Punjabi NM. The epidemiology of adult obstructive sleep apnea. Proc Am Thorac

Soc 2008;5:136-43.2. Deodhar A, Braun J, Inman RD, et al. Golimumab reduces sleep disturbance in

patients with active ankylosing spondylitis: results from a randomized, placebo-controlled trial. Arthritis Care Res (Hoboken) 2010;62:1266-71.

3. Heiberg T, Lie E, van der Heijde D, Kvien TK. Sleep problems are of higher prior-ity for improvement for patients with ankylosing spondylitis than for patients with other inflammatory arthropathies. Ann Rheum Dis 2011;70:872-3.

4. Lee YC, Chibnik LB, Lu B, et al. The relationship between disease activity, sleep, psychiatric distress and pain sensitivity in rheumatoid arthritis: a cross-sectional study. Arthritis Res Ther 2009;11:R160.

5. Erb N, Karokis D, Delamere JP, Cushley MJ, Kitas GD. Obstructive sleep apnoea as a cause of fatigue in ankylosing spondylitis. Ann Rheum Dis 2003;62:183-4.

6. Solak O, Fidan F, Dündar U, et al. The prevalence of obstructive sleep ap-noea syndrome in ankylosing spondylitis patients. Rheumatology (Oxford) 2009;48:433-5.

7. Callis Duffin K, Wong B, Horn EJ, Krueger GG. Psoriatic arthritis is a strong predictor of sleep interference in patients with psoriasis. J Am Acad Dermatol 2009;60:604-8.

8. Kim N, Thrash B, Menter A. Comorbidities in psoriasis patients. Semin Cutan Med Surg 2010;29:10-5.

9. Yamamoto J, Okamoto Y, Shibuya E, Nishimura M, Kawakami Y. Obstructive sleep apnea syndrome induced by ossification of the anterior longitudinal liga-ment with ankylosing spondylitis. Nihon Kokyuki Gakkai Zasshi 2000;38:413-6.

10. Peckett AJ, Wright DC, Riddell MC. The effects of glucocorticoids on adipose tissue lipid metabolism. Metabolism 2011;60:1500-10.

11. Isono S. Obesity and obstructive sleep apnoea: mechanisms for increased col-lapsibility of the passive pharyngeal airway. Respirology 2012;17:32-42.

12. Fuerderer S, Eysel-Gosepath K, Schröder U, Delank KS, Eysel PJ. Retro-pha-ryngeal obstruction in association with osteophytes of the cervical spine. Bone Joint Surg Br 2004;86:837-40.

13. Ryan S, Taylor CT, McNicholas WT.Predictors of elevated nuclear factor-kap-paB-dependent genes in obstructive sleep apnea syndrome. Am J Respir Crit Care Med 2006;174:824-30.

14. Garvey JF, Taylor CT, McNicholas WT. Cardiovascular disease in obstructive sleep apnoea syndrome: the role of intermittent hypoxia and inflammation. Eur Respir J 2009;33:1195-205.

15. Kapsimalis F, Basta M, Varouchakis G, Gourgoulianis K, Vgontzas A, Kryger M. Cytokines and pathological sleep. Sleep Med 2008;9:603-14.

16. Herenius MM, Hoving JL, Sluiter JK, et al. Improvement of work ability, quality of life, and fatigue in patients with rheumatoid arthritis treated with adalimumab. J Occup Environ Med 2010;52:618-21.

17. Rudwaleit M, Gooch K, Michel B, et al. Adalimumab improves sleep and sleep quality in patients with active ankylosing spondylitis. J Rheumatol 2011;38:79-86.

18. Vgontzas AN, Zoumakis E, Lin HM, Bixler EO, Trakada G, Chrousos GP. Marked decrease in sleepiness in patients with sleep apnea by etanercept, a tumor ne-crosis factor-alpha antagonist. J Clin Endocrinol Metab 2004;89:4409-13.

19. Edrees AF, Misra SN, Abdou NI. Anti-tumor necrosis factor (TNF) therapy in rheumatoid arthritis: correlation of TNF-alpha serum level with clinical response and benefit from changing dose or frequency of infliximab infusions. Clin Exp Rheumatol 2005;23:469-74.

20. Winkler G, Lakatos P, Salamon F, et al. Elevated serum TNF-alpha level as a link between endothelial dysfunction and insulin resistance in normotensive obese patients. Diabet Med 1999;16:207-11.

21. Braun J, Baraliakos X. Imaging of axial spondyloarthritis including ankylosing spondylitis. Ann Rheum Dis 2011;70 Suppl 1:i97-103.

22. Kent BD, Ryan S, McNicholas WT. Obstructive sleep apnea and inflammation: Relationship to cardiovascular co-morbidity. Respir Physiol Neurobiol 2011 Mar 23.[Epub ahead of print]

23. Kuhle S, Urschitz MS. Anti-inflammatory medications for obstructive sleep ap-nea in children. Cochrane Database Syst Rev 2011:CD007074.

24. Smith I, Lasserson TJ, Wright J. Drug therapy for obstructive sleep apnoea in adults. Cochrane Database Syst Rev 2006 19:CD003002.


JA Walsh, KC Duffin, J Crim et al25. Huston JM, Tracey KJ. The pulse of inflammation: heart rate variability, the cho-

linergic anti-inflammatory pathway and implications for therapy. J Intern Med 2011;269:45-53.

26. Wang X, Yang Z, Xue B, Shi H. Activation of the cholinergic antiinflammatory pathway ameliorates obesity-induced inflammation and insulin resistance. En-docrinology 2011;152:836-46.

27. Hedner J, Kraiczi H, Peker Y, Murphy P. Reduction of sleep-disordered breathing after physostigmine. Am J Respir Crit Care Med 2003;168:1246-51.

28. Mermigkis C, Bouloukaki I, Mermigkis D, et al. CRP evolution patter in CPAP-treated obstructive sleep apnea patients. Does gender play a role? Sleep Breath 2011 Sept 1. [Epub ahead of print]

29. Bixler EO, Vgontzas AN, Ten Have T, Tyson K, Kales A. Effects of age on sleep apnea in men: Prevalence and severity. Am J Respir Crit Care Med 1998;157:144-8.

30. Sharafkhaneh A, Richardson P, Hirshkowitz M. Sleep apnea in a high risk population: a study of Veterans Health Administration beneficiaries. Sleep Med 2004;5:345-50.

31. Kreis P, Kripke DF, Ancoli-Israel S. Sleep apnea: a prospective study. West J Med 1983;139:171-3.

32. Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med 2002;165:1217-39.

33. Sanner BM, Klewer J, Trumm A, Randerath W, Kreuzer I, Zidek W. Long-term treatment with continuous positive airway pressure improves quality of life in obstructive sleep apnoea syndrome. Eur Respir J 2000;16:118-22.

34. Young T, Finn L, Peppard PE, et al. Sleep disordered breathing and mortality: eighteen-year follow-up of the Wisconsin sleep cohort. Sleep 2008;31:1071-8.

35. Jennum P, Kjellberg J. Health, social and economic consequences of sleep-disordered breathing: a controlled national study. Thorax 2011;66:560-6.

36. Kohler M, Smith D, Tippett V, Stradling JR. Predictors of long-term compliance with continuous positive airway pressure. Thorax 2010;65:829-32.

37. Randerath WJ, Verbraecken J, Andreas S, et al. Non-CPAP therapies in obstruc-tive sleep apnoea. Eur Respir J 2011;37:1000-28.

38. Woodson BT. Non-pressure therapies for obstructive sleep apnea: surgery and oral appliances. Respir Care 2010;55:1314-21; discussion 1321.

39. Santos-Silva R, Sartori DE, Truksinas V, et al. Validation of a portable moni-toring system for the diagnosis of obstructive sleep apnea syndrome. Sleep 2009;32:629-36.

40. Collop NA. Portable monitoring for the diagnosis of obstructive sleep apnea. Curr Opin Pulm Med 2008;14:525-9.

ACKNOWLEDGMENTSThis research was conducted at the George E. Wahlen Veteran Affairs Medical

Center in Salt Lake City, Utah. We gratefully acknowledge Jane Elizabeth Bell, APRN, John Shigeoka, M.D., and Sarah Richey, M.D.,for their contributions with sleep study coordination and manuscript preparation.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication September, 2011Submitted in final revised form February, 2012Accepted for publication April, 2012Address correspondence to: Jessica Walsh, Division of Rheumatology, 50 North Medical Drive, Salt Lake City, UT 84109; Tel: (801) 581-4333; Fax: (801) 581-6069; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. The authors have indicated no financial

conflicts of interest.


Background: Both obstructive sleep apnea (OSA) and prolonged QRS duration are associated with hypertension, heart failure, and sudden cardiac death. However, pos-sible links between QRS duration and OSA have not been explored.Methods: Cross-sectional study of 221 patients who under-went polysomnography at our center. Demographics, cardio-vascular risk factors and ECG were collected to explore a rela-tionship between OSA and QRS duration.Results: The apnea-hypopnea index (AHI) was positively correlated with QRS duration (r = 0.141, p = 0.03). Patients were divided into 3 groups: AHI < 5 (61), AHI 5-29 (104), and AHI > 30 (55). The mean QRS duration prolonged signifi cantly as OSA worsened (AHI < 5, 85 ± 9.5; AHI 5-29, 89 ± 11.9; and AHI > 30, 95 ± 19.9 ms, p = 0.001). QRS ≥ 100 ms was

present in 12.7% of patients with severe OSA compared with 0% in the rest of the sample (p < 0.0001). After adjustment for age, race, and cardiovascular risk factors, this association remained signifi cant in women but not in men.Conclusion: QRS duration and OSA were signifi cantly associ-ated. Severity of OSA independently predicted prolonged QRS in women but not men. Nevertheless, prolongation of QRS du-ration in either sex may potentiate arrhythmic risks associated with OSA.Keywords: Obstructive sleep apnea, electrocardiogram, QRS durationCitation: Gupta S; Cepeda-Valery B; Romero-Corral A; Shamsuzzaman A; Somers VK; Pressman GS. Association between QRS duration and obstructive sleep apnea. J Clin Sleep Med 2012;8(6):649-654.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Obstructive sleep apnea (OSA) has been associated with several cardiovascular conditions1,2 including hyperten-

sion,3 congestive heart failure,4 and sudden cardiac death.5 In fact, OSA has been associated with a 60% to 70% increased risk of cardiovascular morbidity and mortality.6

OSA is highly prevalent and underdiagnosed7 in the general population. This problem may be worse in subjects with coro-nary artery disease, as recent studies have shown a 50% to 70% prevalence of OSA in such patients.7,8 This high prevalence suggests that OSA may play a key role in the development of cardiovascular events.

Numerous studies have shown that OSA is associated with changes in cardiac structure, including left ventricular hypertrophy,9-12which can widen the QRS complex on the elec-trocardiogram (ECG).13 QRS duration is an independent predic-tor of sudden cardiac death15,16 and mortality16,17 in conditions commonly associated with OSA, such as hypertension and heart failure. However, the relationship between OSA and QRS duration is unknown.

In this study we hypothesized that OSA severity, measured by the apnea-hypopnea index (AHI), is independently associ-ated with QRS prolongation. If so, this could help explain any increased risk of sudden cardiac death in OSA.

METHODS

We conducted a retrospective cross-sectional study of con-secutive patients who had a clinically indicated polysomno-

Association between QRS Duration and Obstructive Sleep ApneaShuchita Gupta, M.D.2; Beatriz Cepeda-Valery, M.D.2; Abel Romero-Corral, M.D., M.Sc.2,3; Abu Shamsuzzaman, M.D., Ph.D.4;

Virend K. Somers, M.D., Ph.D.3; Gregg S. Pressman, M.D.1

1Cardiovascular Division, Einstein Medical Center, Philadelphia, PA; 2Department of Internal Medicine, Einstein Medical Center, Philadelphia, PA; 3Cardiovascular Division, Mayo Clinic, Rochester, MN;

4Cincinnati Children’s Hospital Medical Center, Cincinnati, OH

gram (PSG) study from January 2007 through July 2009 at Albert Einstein Medical Center, Philadelphia, PA. Inclusion criteria were age over 18 years and availability of a resting 12-lead ECG within the 6 months preceding polysomnography. Patients with known sleep apnea were excluded as were those with classic right or left bundle branch block or a ventricular paced rhythm. Patients with atrial fi brillation were included but other non-sinus rhythms were excluded. The fi nal study popula-tion consisted of 221 subjects. This study was approved by the Albert Einstein Institutional Review Board.

PolysomnogramAll patients were initially evaluated by a certifi ed sleep

medicine certifi ed physician. PSG recording and scoring were performed using the VIASYS SomnoStar Pro System. Electro-encephalogram, electro-oculogram, and electromyogram were monitored for sleep staging. Nasal pressure monitoring, chest

BRIEF SUMMARYCurrent Knowledge/Study Rationale: QRS duration is an indepen-dent predictor of mortality. Obstructive sleep apnea produces structural changes of the heart which can lead to a widened QRS but this potential association has not been adequately explored.Study Impact: QRS duration prolonged signifi cantly as sleep apnea se-verity increased; however, on multivariate analysis the effect remained signifi cant only in women. QRS widening could help explain associa-tions between sleep apnea and cardiovascular events while the effect in women points to possible areas for future research.


S Gupta, B Cepeda-Valery, A Romero-Corral et al

wall movement, abdominal movement, snoring, and oxygen saturation were monitored for respiratory assessment. Electro-cardiogram and tibial electromyogram were monitored for car-diac arrhythmias and nocturnal limb movements, respectively. The AHI, defined as the sum of apneas and hypopneas per hour of sleep, was recorded from the PSG report. OSA severity was defined by AHI according to recommendations from the Sleep Task Force of the American Academy of Sleep Medicine (no OSA when AHI < 5, mild-moderate OSA with AHI 5-30, and severe OSA with AHI > 30).17

ElectrocardiogramThe resting 12-lead ECGs of these patients were accessed

from their electronic medical records. Heart rate, QRS du-ration, PR interval, and corrected QT interval (QTc) were abstracted from the computerized ECG interpretation (Mar-quette 12SL ECG Analysis Program run on the MUSE ECG system). The onset of QRS is detected first because it is the easiest; the slope change is usually very rapid and in great contrast to the other slopes in the median. This is followed by QRS offset. The onsets and offsets are determined by an anal-ysis of simultaneous slopes in all 12 leads. Onsets are defined as the earliest deflection in any lead, and offsets as the latest deflection in any lead. Thus, the QRS duration is measured from the earliest onset in any lead to the latest deflection in any lead. Similarly, the QT interval is measured from the ear-liest detection of depolarization in any lead to the latest detec-tion of repolarization in any lead. The PR interval is measured form the earliest detection of atrial depolarization in any lead to the earliest detection of ventricular depolarization in any lead (the QRS onset). The QT is corrected for heart rate using Bazett’s formula. The values of these intervals were recorded in milliseconds (ms).

Data Collection and Cardiovascular Risk FactorsPSG, ECG, and demographic data including age, gender, and

race were collected from electronic medical records. Cardio-vascular risk factors—namely hypertension, diabetes, coronary artery disease, congestive heart failure, atrial fibrillation, obe-sity (BMI ≥ 30 kg/m2), and smoking status—were noted from the records. Recording of these variables were done blinded to the OSA status or severity.

Statistical MethodsData were summarized by calculating mean ± SD for con-

tinuous variables, and numbers and percentages for categori-cal variables. Due to non-linearity of the variables analyzed we used Spearman correlation coefficients to explore the associa-tion as continuous variables between AHI, mean and minimum oxygen saturation during sleep, with QRS and other ECG inter-vals. We performed one-way ANOVA and independent t-tests to assess differences between cardiovascular risk factors and ECG intervals by presence or absence of OSA and by its sever-ity (AHI < 5; no OSA, AHI 5-29; mild-to-moderate OSA and AHI ≥ 30; severe OSA) and by mean oxygen saturation < 88 % (cut-off used by Medicare for home oxygen) and minimum oxygen saturation < 94% (lowest quartile) during sleep. We then assessed whether severity of OSA was associated with a prolonged QRS (≥ 100 ms). Finally, we performed regression analyses to identify significant univariate predictors for QRS duration (p-value ≤ 0.10). To assess the independent effect of OSA on AHI and mean and minimum oxygen saturation during sleep, we performed multivariate analyses adjusting for age, sex, and cardiovascular risk factors (p-value < 0.05). Due to significant differences in QRS duration between men and wom-en previously reported18,19 and noted in our study, we performed analyses stratified by gender, and therefore no adjustment for gender was needed. All analyses were performed using JMP 8.0 (SAS Institute; Cary, NC) and 2-tailed p values < 0.05 were considered significant in advance.

RESULTS

Baseline characteristics of the 221 participants (64.2% fe-male) are shown in table 1. Characteristics of the subjects and prevalence of cardiovascular risk factors according to the pres-ence or absence of OSA and its severity are shown in table 2. Overall, subjects with OSA tended to be older and male. Sub-jects with mild to moderate OSA had a higher prevalence of congestive heart failure and subjects with severe OSA tended to be more obese and hypertensive. Because there were no signifi-cant differences between the mild and moderate OSA groups, we decided to cluster them in a single group.

QRS and AHI showed a weak but significant and positive correlation (rho = 0.14, p = 0.03). table 3 displays ECG inter-vals by presence or absence of OSA and by its severity. QRS was significantly wider as OSA severity increased (Figure 1). Furthermore, QRS ≥ 100 ms was present in 12.7% of those with severe OSA compared with 0% of patients in the mild to mod-erate OSA group, and 0% in those without OSA, p < 0.0001 (Figure 2). PR and QTc intervals were not significantly corre-lated with AHI (rho = -0.03, p = 0.58 and rho = -0.02, p = 0.75, respectively) and were not significantly different among groups.

Table 1—Baseline characteristics of the subjects included in the study

Variable n = 221 Mean ± SD or Number (%)Age 51.9 ± 13.8Gender (Female) 142 (64.2)Race

Caucasian 29 (13.1)African American 170 (77)Hispanic 19 (8.6)Other (Asian) 3 (1.3)

No Obstructive sleep apnea 61 (27.6)Obstructive sleep apnea 160 (72.3)Mild-moderate 105 (47.5)Severe 55 (24.8)Obese (BMI > 30 kg/m2) 187 (85)Smoking 77 (34.8)Diabetes mellitus type 2 75 (33.9)Hypertension 140 (63.3)Coronary artery disease 34 (15.3)Congestive heart failure 19 (8.6)Atrial fibrillation 8 (3.6)

SD, standard deviation; BMI, body mass index.


OSA and QRS

To examine a possible relationship between nocturnal hy-poxemia and QRS duration, we also looked at mean oxygen saturation and minimal oxygen saturation. Mean QRS dura-tion was 93.4 ± 16.4 ms for the lowest quartile of mean oxygen saturation (< 94%) versus 88.9 ± 14.3 ms for those with mean oxygen saturation ≥ 94% (p = 0.049). We next divided patients into those with minimum oxygen saturation of ≤ 88% (cur-

rent Medicare criteria for home oxygen prescription) and those with minimum oxygen saturation > 88%. Mean QRS duration for patients with nocturnal hypoxemia as defined above was 91.6 ± 16.0 ms compared to 86.1 ± 10.3 ms for those without (p = 0.005). After stratifying by sex, the difference in QRS du-ration remained significant in women (88.6 ± 14.3 ms vs. 83.5 ± 8.3 ms in patients with and without nocturnal hypoxemia), but

0%

2%

4%

6%

8%

10%

12%

14%

Severity of OSA

None Mild to Moderate Severe

Figure 2—QRS > 100 ms by presence or absence of OSA and by its severity

Table 2—Characteristics by presence or absence of obstructive sleep apnea and by its severity

Variable No OSA (n = 61) Mild-moderate OSA (n = 105) Severe OSA (n = 54) p-valueAge 47.3 ± 14.3 54.2 ± 11.5# 52.9 ± 16.1+& 0.0900Sex (female) 50 (81.9) 63 (60)* 29 (52.7)^ 0.0020Race

Caucasian 7 (11.5) 15 (14.3) 7 (12.7)African American 45 (73.7) 79 (75.2) 46 (83.6)Hispanic 8 (13.1) 10 (9.5) 1 (1.8)Other (Asian) 1 (1.64) 1 (0.95) 1 (1.8) 0.51

Obese (BMI > 30 kg/m2) 48 (78.7) 90 (85.7) 49 (90.7)& 0.1800Smoker 23 (37.7) 35 (33.3) 19 (34.5) 0.8400Diabetes mellitus type 2 19 (31.1) 34 (32.4) 22 (40) 0.5400Hypertension 34 (55.7) 67 (63.8) 39 (70.9)& 0.2300Coronary artery disease 8 (13.1) 20 (19) 6 (10.9) 0.3300Congestive heart Failure 2 (3.3) 12 (11.4)# 5 (9.1) 0.1900Atrial fibrillation 1 (1.6) 3 (2.8) 4 (7.2) 0.2200

*p-value ≤ 0.05 for no OSA vs. mild-moderate OSA, #p-value ≤ 0.001 for no OSA vs. mild-moderate OSA, &p-value ≤ 0.07 for no OSA vs. severe OSA, +p-value ≤ 0.05 for moderate vs. severe OSA, ^p-value ≤ 0.001 for no OSA vs. severe OSA. BMI, body mass index.

Table 3—ECG intervals by presence or absence of OSA and by its severity

Variable No OSA (n = 61) Mild-moderate OSA (n = 105) Severe OSA (n = 54) p-valuePR interval 159.01 ± 26.7 163.09 ± 24.9 158.23 ± 27.2 0.4500QRS interval 85.80 ± 9.5 89.19 ± 11.9* 95.20 ± 19.9^ 0.0014QT interval 386.01 ± 39.8 390.30 ± 35.43 390.47 ± 42.4 0.7500

*p-value 0.01 for no OSA vs. mild-moderate OSA, ^p-value 0.0003 for no OSA vs. severe OSA. ECG, electrocardiogram; OSA, obstructive sleep apnea.

0 1 2

150

140

130

120

110

100

90

80

70

60

AHI Groups

QRS

Dura

tion

Figure 1—QRS duration by obstructive sleep apnea severity measured by apnea-hypopnea index

0 = No OSA; 1 = Mild-moderate OSA; 2 = Severe OSA.


S Gupta, B Cepeda-Valery, A Romero-Corral et alnot in men (96.1 ± 17.5 ms vs. 96.4 ± 11.5 ms in patients with and without nocturnal hypoxemia).

Sex, race, hypertension, coronary artery disease, congestive heart failure, and AHI were significant univariate predictors for QRS duration (table 4). After stratifying by sex, women had a shorter QRS duration than men (86.8 ± 12.7 vs. 94.9 ± 15.1 ms), consistent with previous studies.18,19 In multivariate analyses, in men, atrial fibrillation, congestive heart failure, and race remained significant predictors of QRS duration after ad-justments, while in women, atrial fibrillation, congestive heart failure, and AHI remained as significant predictors (table 5). Minimal oxygen saturation was not a significant predictor of QRS duration on multivariate analyses.

DISCUSSION

Our major finding is that QRS duration and OSA are signifi-cantly associated. This was particularly true for patients with severe OSA (AHI > 30), where 12.7% had a QRS ≥ 100 ms compared to 0% of patients without OSA or with only mild to moderate OSA. After multivariate analyses, AHI remained a sig-nificant predictor of prolonged QRS in women but not in men.

Previous StudiesPrevious small studies have looked for a possible association

between QRS duration and OSA. Andreas et al.20 found a mean QRS duration of 96 ± 9 ms in OSA, similar to the 95.20 ± 19.9 ms found in our study. Their control subjects had similar QRS duration (96 ± 10 ms) and the authors concluded that OSA was not associated with QRS prolongation. In our study, subjects without OSA had a more normal QRS duration (85.8 ± 9.5 ms). This difference may relate to the fact that their controls were snorers with an AHI up to 10, while in our study controls had an AHI < 5 as suggested by the American Academy of Sleep

Medicine Task Force.17 Furthermore, their sample size was small (n = 47). In another study, Steiner et al.21 studied 12 pa-tients with congestive heart failure and compared ECG param-eters by presence or absence of OSA. Although they noted no statistically significant changes in QRS duration between OSA groups, the mean QRS was 99 ± 13 ms in controls and 144 ± 43 ms in patients with OSA. This study is limited by small sample size (6 patients in each group) and by the mean AHI of the OSA group being only 17.8; therefore few conclusions can be drawn from these results.

In previous studies, QRS ≥ 120 ms has been independently associated with all-cause mortality15,16 and sudden death.14-16 Though mean QRS duration in our patients with severe OSA was < 120 ms, it is notable that only severe OSA subjects had a QRS ≥ 100 ms while this finding was absent in others. We can speculate that changes in QRS duration occur slowly and move in only one direction (wider). In addition OSA probably wors-ens over time. Therefore our patients might, over time, develop additional QRS widening. Further study is needed to investi-gate this possibility.

One of the interesting findings in this study is that AHI was independently associated with QRS duration in women but not in men. There are several possible ways to explain this. Prior studies have noted shorter QRS duration in women, as we also found. Any increase in QRS duration in response to OSA would produce a greater percentage change in women versus men, making it easier to detect statistically. On the other hand, there may be a true difference between men and women in response of the left ventricle to the afterload burden imposed by OSA. Animal studies have shown differences in gene expression be-tween males and females in response to pressure overload.22 In addition, women appear to have greater wall thickening in response to aortic stenosis than do men.23 Among hyperten-sive patients followed by the LIFE study, women had reduced regression of electrocardiographic hypertrophy after 5 years of therapy versus men.24 Given these findings, it may be that women develop greater left ventricular hypertrophy in response to OSA, leading to greater QRS prolongation. Wall thickness tends to be slightly less in women than men; thus the female heart might experience greater wall stress from the high nega-tive intrathoracic pressure produced by obstructive apneas.

Regarding the relationship between QT duration and OSA, previous studies have reported conflicting findings. Smith et al.25 measured QT and PR interval changes associated with spontaneous and respiratory-related arousals in OSA patients, finding a shortening of the QT during arousal. However, that study included only 20 OSA patients and lacked a control group. In another study, Lue et al.26 reported opposite find-

Table 5—Multivariate predictors for QRS duration stratified by sex

PredictorMales

ß-Estimate (SE) p-valueFemales

ß-Estimate (SE) p-valueRace -6.5369 (3.02) 0.0343 -1.3668 (1.94) 0.4840Congestive heart failure 15.1239 (5.32) 0.0059 8.4250 (4.11) 0.0426Atrial fibrillation 24.7257 (6.64) 0.0004 16.7225 (7.46) 0.0260Apnea-hypopnea index -0.0257 (0.05) 0.6409 0.1055 (0.05) 0.0565

All variables listed in Table 1 were included in the model.

Table 4—Univariate predictors for QRS durationPredictor ß-Estimate (SE) p-value

Gender (Female) -4.0230 (0.96) < 0.0001Race -1.876 (0.97) 0.0600Hypertension -1.7988 (0.98) 0.0700Coronary artery disease -2.8586 (1.31) 0.0308Congestive heart failure -6.1830 (1.65) 0.0002Atrial fibrillation -11.5424 (2.44) < 0.0001Apnea hypopnea index 0.0705 (0.03) 0.0540

All variables in Table 1 were analyzed. Significant predictors were considered if p-value ≤ 0.10.


OSA and QRSings with prolongation of the QT and corrected QT intervals in the OSA group compared to controls. In our study, we did not find an association between OSA and QTc interval. These conflicting results might be explained by the multiple factors that influence QT duration, and the fact that it can change very rapidly.

Mechanistic ConsiderationsProlonged QRS and OSA can each be seen in various car-

diovascular conditions with important clinical implications.1,2 Both have been associated with increased left ventricle mass, hypertension,3 congestive heart failure,4 and ventricular ar-rhythmias.27,28 Several pathophysiologic mechanisms can ex-plain these associations. In OSA, breathing against a closed upper airway generates high negative intrathoracic pressure, which increases left ventricular transmural pressure and af-terload, contributing to left ventricular hypertrophy.29 Further-more, OSA is an accepted causal factor for the development of hypertension, which also contributes to left ventricular hy-pertrophy. Other possible mechanisms by which OSA might contribute to QRS widening include endothelial dysfunc-tion,29 systemic inflammation, sympathetic activation, blood pressure surges, and oxidative stress, all of which contribute to increased arterial stiffness9 and increased afterload stress on the left ventricle.

Strengths and LimitationsOur study has a large sample size in comparison with most

OSA studies. All of our patients underwent polysomnography study, the gold standard test to define this condition; this al-lowed us to assess OSA severity. We performed multivariate analyses to control for many of the clinical factors that could confound our results. Our study has some limitations. Given the cross-sectional nature of the study we cannot determine causal-ity, as we cannot assess if prolonged QRS duration preceded or followed development of OSA. Our sample population was mainly African American, limiting generalizability of these results to other populations. In the multivariate analysis QRS prolongation was independently associated with OSA in wom-en, but not in men, which could be due to overadjustment (ex-ample: hypertension and obesity, conditions commonly found in OSA) or insufficient power, as the majority of our popula-tion sample were women. Lead time bias is likely present in our study, as the onset of OSA in our population is not well established. However, in theory the longer the disease has been present, the greater the impact on cardiovascular disease, in our case the greater the QRS prolongation.

CONCLUSIONS

This is the first study linking increased QRS duration with OSA and may help explain the increased risk of sudden cardi-ac death in both conditions. For the overall sample, increased QRS duration was significantly associated with OSA, partic-ularly in those with severe OSA (AHI > 30). After multivari-ate analyses severe OSA remained a significant independent predictor of prolonged QRS in women but not in men, while congestive heart failure and atrial fibrillation remained inde-pendent predictors in both men and women. Nevertheless,

prolongation of QRS duration in either sex may potentiate arrhythmic risk associated with OSA. Mechanisms underly-ing the gender effects of the QRS-OSA interaction remain to be determined.

REFERENCES1. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence

of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328:1230-5.

2. Malhotra A, White DP. Obstructive sleep apnoea. Lancet 2002;360:237-45.3. Palomaki H, Partinen M, Juvela S, Kaste M. Snoring as a risk factor for sleep-

related brain infarction. Stroke 1989;20:1311-15.4. Shahar E, Whitney CW, Redline S, et al. Sleep-disordered breathing and cardio-

vascular disease: cross-sectional results of the Sleep Heart Health Study. Am J Respir Crit Care Med 2001;163:19-25.

5. Gami AS, Somers VK. Implications of obstructive sleep apnea for atrial fibril-lation and sudden cardiac death. J Cardiovac Electrophysiol 2008;19:997-03.

6. Hung J, Whitford EG, Parsons RW, Hillman DR. Association of sleep apnoea with myocardial infarction in men. Lancet 1990;336:261-4.

7. Lee CH, Khoo SM, Tai BC, et al. Obstructive sleep apnea in patients admitted for acute myocardial infarction. Chest 2009;135:1488-95.

8. Sert Kuiyoshi FH, Garcia-Touchard A, Gami AS, et al. Day-night variation of acute myocardial infarction in obstructive sleep apnea. J Am Coll Cardiol 2008;52:343-6.

9. Drager L, Bortolotto LA, Figueiredo LA, Caldin B, Krieger E, Lorenzi-Filho E. Obstructive sleep apnea, hypertension, and their interaction on arterial stiffness and heart remodeling. Chest 2007;131:1379-86

10. Noda A, Okada T, Yasuma F, Nakashima N, Yokota. Cardiac hypertrophy in ob-structive sleep apnea syndrome. Chest 1995;107:1538-44.

11. Shivalkar B, Van de Heyning C, Kerremans M, Rinkevich D, Verbraecken J, De Backer W. Obstructive sleep apnea syndrome: more insights on structural and functional cardiac alterations, and the effects of treatment with continuous posi-tive airway pressure. J Am Coll Cardiol 2006;47:1433-9.

12. Usui K, Parker JD, Newton GE, Floras JS, Ryan CM, Bradley TD. Left ventricular structural adaptations to obstructive sleep apnea in dilated cardiomyopathy. Am J Respir Crit Care Med 2006;173:1170-5.

13. Dhingra R, Ho Nam, B, Benjamin EJ, et al. Cross-sectional relations of elec-trocardiographic QRS duration to left ventricle dimensions. J Am Coll Cardiol 2005;45:685-9.

14. Olshansky B. Wide QRS, narrow QRS. J Am Coll Cardiol 2005;46:317-9.15. Morin DP, Oikarinen L, Viitasalo M, et al. QRS duration predicts sudden cardiac

death in hypertensive patients undergoing intensive medical therapy: the LIFE study. Eur Heart J 2009;30:2908-14

16. Ott P, Marcus FI. Electrocardiographic markers of sudden death. Cardiol Clin 2006;24:453-69

17. The Report of an American Academy of Sleep Medicine Task Force. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:667-89.

18. Macfarlane PW, McLaughlin SC, Devine B, Yang TF. Effects of age, sex, and race on ECG interval measurements. J Electrocardiol 1994;27:14-9.

19. Okin PM, Roman MJ, Devereux RB, Kligfield P. Gender differences and the electrocardiogram in left ventricular hypertrophy. Hypertension 1995;25:242-9.

20. Andreas S, Von Breska B, Schaumann A, Gonska BD, Kreuzer H. Obstructive sleep apnoea and signal averaged electrocardiogram. Eur Respir J 1995;8:546-50.

21. Steiner S, Schueller PO, Hennersdorf MG, Strauer BE. Obstructive sleep apnea in heart failure patients: evidence for persistent conduction disturbances or sinus node dysfunction. J Physiol Pharmacol 2008;59:669-74.

22. Weinberg EO, Thienelt CD, Katz SE, et al. Gender differences in molecular re-modeling in pressure overload hypertrophy. J Am Coll Cardiol 1999;34:264-73.

23. Piro M, Della Bona R, Abbate A, Biasucci LM, Crea F. Sex-related differences in myocardial remodeling. J Am Coll Cardiol 2010;55:1057-65.

24. Okin P, Gerdts E, Kjeldsen SE, et al. Devereux for the Losartan Intervention for Endpoint Reduction in Hypertension Study Investigators. Gender differences in regression of electrocardiographic left ventricular hypertrophy during antihyper-tensive therapy. Hypertension 2008;52:100-6.

25. Smith JH, Baumert M, Nalivaiko E, McEvoy RD, Catcheside PG. Arousal in ob-structive sleep apnea with ECG RR and QT interval shortening and PR interval lengthening. J Sleep Res 2009;18:188-95.


S Gupta, B Cepeda-Valery, A Romero-Corral et al26. Luo YQ, Yang Y, Yu SF. Influence of obstructive sleep apnea-hypopnea syn-

drome on the Qtc interval. Zhong Nan Da Xue Bao Yi Xue Ban 2004; 29:97-8.27. Guilleminault C, Connolly SJ, Winkle RA. Cardiac arrhythmia and conduction

disturbances during sleep in 400 patients with sleep apnea syndrome. Am J Cardiol 1983;52:490-4.

28. Mehra R, Benjamin EJ, Shahar E, et al. Association of nocturnal arrhythmias with sleep-disordered breathing: The Sleep Heart Health Study. Am J Respir Crit Care Med 2006;173:910-6.

29. Somers VK, White DP, Amin R, et al. Sleep apnea and cardiovascular disease: an American Heart Association/American College of Cardiology Foundation Sci-entific Statement from the American Heart Association Council for High Blood Pressure Research Professional Education Committee, Council on Clinical Cardiology, Stroke Council, and Council on Cardiovascular Nursing. J Am Coll Cardiol 2008;52: 686-717.

30. Shewan LG and Coats AJ. Ethics in the authorship and publishing of scientific articles. Int J Cardiol 2010;144:1-2.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication December, 2011Submitted in final revised form March, 2012Accepted for publication April, 2012Address correspondence to: Shuchita Gupta, M.D., Division of Cardiovascular Medicine, Albert Einstein Medical Center, 3232 Levy Building, 5501 Old York Road, Philadelphia, PA 19141; Tel: (215) 605-6310; Fax: (215) 790-2943

DISCLOSURE STATEMENTThis was not an industry supported study. Dr. Somers serves as a consultant

for Apnex Medical, Inc., Johnson & Johnson, Sova Pharmaceuticals, Respicardia, ResMed, and Neupro. He has received research support from Philips Respironics, Inc. Dr. Pressman has received a research grant from Philips Respironics, Inc. The other authors have indicated no financial conflicts of interest.


Study Objectives: Determining the presence and sever-ity of obstructive sleep apnea (OSA) is based on apnea and hypopnea event rates per hour of sleep. Making this determination presents a diagnostic challenge, given that summary metrics do not consider certain factors that infl u-ence severity, such as body position and the composition of sleep stages.Methods: We retrospectively analyzed 300 consecutive diag-nostic PSGs performed at our center to determine the impact of body position and sleep stage on sleep apnea severity.Results: The median percent of REM sleep was 16% (reduced compared to a normal value of ~25%). The median percent su-pine sleep was 65%. Fewer than half of PSGs contained > 10 min in each of the 4 possible combinations of REM/NREM and supine/non-supine. Half of patients had > 2-fold worsening of the apnea-hypopnea index (AHI) in REM sleep, and 60% had > 2-fold worsening of AHI while supine. Adjusting for body posi-

tion had greater impact on the AHI than adjusting for reduced REM%. Misclassifi cation—specifi cally underestimation of OSA severity—is attributed more commonly to body position (20% to 40%) than to sleep stage (~10%).Conclusions: Supine-dominance and REM-dominance com-monly contribute to AHI underestimation in single-night PSGs. Misclassifi cation of OSA severity can be mitigated in a patient-specifi c manner by appropriate consideration of these vari-ables. The results have implications for the interpretation of single-night measurements in clinical practice, especially with trends toward home testing devices that may not measure body position or sleep stage.Keywords: Supine-dominant, REM-dominant, false-negative, phenotypeCitation: Eiseman NA; Westover MB; Ellenbogen JM; Bianchi MT. The impact of body posture and sleep stages on sleep apnea severity in adults. J Clin Sleep Med 2012;8(6):655-666.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

The diagnosis of obstructive sleep apnea (OSA) and its severity categorization are typically based on the apnea-

hypopnea index (AHI) obtained from a single overnight labo-ratory polysomnogram (PSG). Large studies have shown that OSA is associated with cerebrovascular and cardiovascular morbidity and mortality in proportion to severity.1-3 Accurate assignment of apnea severity is therefore important to establish the diagnosis and to motivate treatment decisions. In addition to patient-specifi c considerations for individual care, accurate assessment of OSA severity is important at the population level for establishing genetic, epidemiological, and medical associa-tions with OSA. Despite the importance of accurate diagnostic assessment, obtaining this information is challenging given that sleep apnea is a complex process with multiple contributing factors, some of which vary over time. Providers may thus be left uncertain about how to interpret the presence or severity of OSA after a single night’s examination of sleep.

Current guidelines suggest offering treatment to patients with AHI values ≥ 5/h (with daytime symptoms or snoring), or > 15/h regardless of symptoms.4,5 The AHI does not capture other details about apnea physiology such as event duration or depth of desaturation but has been accepted as a gateway to diagnosis and treatment. One important question arises with re-spect to the AHI from a single night: what is the likelihood that an observed AHI value < 5/h or < 15/h would have been higher if more REM sleep or more supine sleep had occurred? A single “negative” study may not be suffi cient to rule out OSA, in part

The Impact of Body Posture and Sleep Stages on Sleep Apnea Severity in Adults

Nathaniel A. Eiseman, B.S.1; M. Brandon Westover, M.D., Ph.D.1; Jeffrey M. Ellenbogen, M.D., F.A.A.S.M.1,2; Matt T. Bianchi, M.D., Ph.D.1,2

1Neurology Department, Massachusetts General Hospital, Boston, MA; 2Division of Sleep Medicine, Harvard Medical School, Boston, MA

due to these factors6,7 that infl uence the summary AHI values. This is especially pertinent for those patients with high pre-test probability for OSA.8 The ideal approach to studying variability in any diagnostic test is to obtain repeated measures, but for sleep, this is often not feasible due to cost and inconvenience.

When only a single night of data is available, careful ac-counting for factors that introduce variability may prove critical for accurate patient-specifi c interpretation. Among the potential sources of variability in AHI, the dependence of apnea sever-ity on body position and sleep stage may be evident within a single PSG. For example, decreased REM sleep (as is common in the fi rst night effect)9 or absent REM on a split-night study might lead to relative underestimation of AHI for patients who exhibit REM-dominant OSA. Body position in the laboratory may lead to over- or underestimation, depending on a patient’s home sleep position patterns. Accounting for these factors may improve the clinical phenotyping of OSA patients, and most

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Current practice is to summarize the OSA diagnosis and severity by the summary full-night AHI value. We sought to determine how body position and sleep stage composition im-pact the full-night AHI in 300 diagnostic studies performed in our center.Study Impact: The common fi nding of REM-dominance and supine-dominance of OSA has important implications for diagnostic classifi ca-tions based on a single night PSG. Improved diagnostic phenotyping of OSA patients has important clinical and research implications.


NA Eiseman, MB Westover, JM Ellenbogen et alimportantly, highlight circumstances in which the summary AHI from a single night PSG provides false reassurance in re-gards to disease severity.

We analyzed 300 diagnostic PSGs spanning a range of sum-mary AHI values to quantify the range of stage- and position-dependence of OSA. We use this cohort to answer the following clinically-relevant questions: (1) What is the distribution of po-sition- and stage-dependence of OSA among clinical diagnostic PSGs; and (2) what are the relative contributions of stage- and position-dependence to misclassification of disease severity?

METHODS

A sample of 300 consecutive overnight diagnostic PSGs per-formed in 2011 in our center, for a variety of indications, was analyzed for sleep stage and apnea severity information. Ret-rospective analysis of clinical PSG data was approved by the Partners Human Research Committee. Our pre-specified exclu-sions were age < 18 years; debilitating neurological disease; sleep efficiency < 60%; total sleep time < 4 h; and use of CPAP, oxygen therapy, or dental appliance during the study night. Sleep was scored according to standard criteria of the American Association of Sleep Medicine10 by experienced sleep techni-cians. Because our laboratory implements clinical criteria for conversion of diagnostic studies to split-night studies, the dis-tribution of sleep apnea severity is biased toward lower val-ues in this sample of full-night diagnostic studies, with an AHI value of 30/h representing the 98th percentile in this cohort. Although we did not sub-classify apnea events, central apneas were only a minority of events, prohibiting separate analysis of positional dependence of central apnea11; the median number of central apneas per study was 2, with the 95% percentile value of 20 per study.

Most clinical PSG variables did not pass tests for normality, and thus nonparametric statistical tests were generally used for analysis (Mann-Whitney rank test for 2-group comparison, or Kruskal-Wallis with Dunn post hoc for 3-group comparisons). Similarly, correlation coefficients were Pearson or Spearman rank method as appropriate based on distribution normality. To calculate REM-dependence ratios, defined as the REM AHI di-vided by the NREM AHI, we restricted analysis to patients with ≥ 5 min of REM sleep (n = 285, i.e., 95% of the cohort). To cal-culate supine-dependence ratios in the initial analysis, defined as the supine AHI divided by the non-supine AHI, we restricted analysis to patients who were supine between 5% and 95% of the night (n = 250, i.e., 83% of the cohort). For subsequent mis-classification models, we repeated analysis using 20% to 80% supine to ensure adequate sampling of the supine-dependence ratios. We rounded any AHI values < 1/h up to a value of 1 to avoid artificially inflating dominance ratios due to fractional denominator values. Similar results were obtained when using the RDI (not shown).

We used Mathematica (Wolfram Research, Champaign, IL) to perform the modeling sections. In order to demonstrate the effects of position dependence, we generated contour plots il-lustrating AHI underestimation according to 2 position-related factors. The first is the protection from sleep apnea afforded by non-supine relative to supine body position. For example, a ratio value of 1 means position independence (no protection

while non-supine), while a ratio of zero indicates perfect pro-tection from apnea while non-supine. The second AHI adjust-ment factor in the contour plots is the proportion of the night spent supine, which can vary from 0-100%. Color gradient en-coded the third dimension, AHI, taking into account these two factors varying over their whole range of possible values.

To simulate the effect of body position we made calculations from 4 theoretical cohorts of 1,000 people each based on re-alistic but conservative parameters. We aimed to approximate the distribution of AHI values in the large population Sleep Heart Health Study, with a peak in the normal range followed by a long-tailed distribution extending into the severe range.12 We assumed that the categories of no apnea (AHI < 5/h), mild apnea (AHI 5-15/h), and moderate apnea (AHI 15-30/h) were normally distributed with mean (SD) values of 2.5/h (1.25), 10/h (2.5), and 22.5/h (7.5), respectively. The severe cohort was modeled using a skew-normal (long-tailed) distribution with a mean AHI of 50/h and 95% of the values between 30/h and 90/h. These modeling assumptions allowed us to test the hypothesis that realistic conservative estimates of body position and supine dominance would lead to substantial misclassification.

For the initial modeling, each subject was assigned an AHI value according to these distributions designed to represent the spectrum of OSA severity. By virtue of the small amount of variance built into the distributions, a small fraction of simu-lated patients were by chance assigned an AHI value in a differ-ent class than the “true” class from which they were drawn. At this initial stage, each subject was also assigned a percentage of the night spent in the supine position (although no position dependence was imposed on the OSA values; this was a place-holder for the next step). These percentages were selected from a normal distribution of the ratio of time spent in the supine versus lateral position. The mean and SD of 1.13 (0.16) were approximated based on those found in the literature,13,14 and were similar to the percent time spent in the supine position in our cohort. In the next step, we imposed a 2-fold worsening of AHI while supine, using the baseline AHI value and the portion of the night spent supine assigned previously to each subject. To capture the reality that position dependence can increase or decrease the overall AHI depending on the baseline, those assigned a supine-to-lateral ratio greater than the mean value had an increase in overall AHI when supine dominance was imposed, while those assigned a ratio less than the mean value had a decrease in overall AHI.

RESULTS

We first illustrate representative clinical scenarios one might encounter when interpreting a single night PSG in a patient sus-pected of having OSA, in whom respiratory event frequency could depend on sleep stage or body position (Figure 1A). In the schematized 7-h night, about half of the time is spent supine, including 1 of the 3 REM periods. Event frequency is shown under several realistic conditions: (1) position- and stage-inde-pendent, (2) only REM dominant, (3) only supine dominant, (4) and supine dominant with additive worsening in the supine REM condition. For comparison, the final condition (5th row) shows supine and REM dominant OSA when the full sleep time was spent supine. This schematic allows one to consider


Sleep Apnea Underestimation

the summary event index across the 4 possible combinations of sleep stage (REM versus NREM) and position (supine ver-sus lateral; Figure 1B). Compared to the “worst-case” circum-stance of all-supine sleep, time in the lateral position leads to underestimation of apnea severity. Since some of the variance in overall AHI calculation from a single PSG may be related to supine and/or REM dependence of the OSA severity, observing patients under conditions of greatest apnea vulnerability—typi-cally the combination of supine body position and REM sleep stage—reveals the potential impact of these dependencies.

Sleep Stage and Body Position Distributions in Diagnostic PSGs

We analyzed a sample of consecutive diagnostic PSGs (n = 300) from our center, regardless of reason for referral (most were referred for sleep apnea). The characteristics of this population are given in table 1. The portion of the total sleep time (TST) spent in REM sleep was variable: the median value was 16%, with 22% having < 10% of the TST in REM sleep, and only 13% exhibiting > 25% of the TST in REM sleep (Figure 2A). This reduction in REM sleep may repre-sent the classic fi rst night effect, which also includes excess N1 and decreased sleep effi ciency.9 The median portion of the TST spent in the supine position was 65%. Only 19% of patients spent > 90% of the TST in the supine position (Figure 2B). The chance of observing the supine position while in NREM was greater than while in REM sleep (median of 68% versus 54% was supine, respectively; p < 0.003, Mann-Whitney test). It is unknown for individual patients how their time spent su-pine in the lab relates to their home sleep position patterns. In the lab, patients are encouraged to sleep supine and may even choose to do so because of the sensor montage; thus, the por-

tion supine observed in the lab may be an overestimate com-pared to home patterns.

Only a fraction of the PSGs contained observations suffi cient to characterize all 4 possible combinations of REM, NREM, supine, and lateral conditions (Figure 2c). For example, 43% of PSGs contained ≥ 10 min in each combination, while fewer than 20% contained ≥ 25 min in each combination. Single-night PSG is thus somewhat limited in the chances of providing the requisite patient-specifi c information to accurately determine stage- and position-dependence.

Apnea Severity: Dependence on Sleep Stage and Body Position

Although OSA is often considered a REM-dominant disor-der, there is a range of sleep stage dependence reported in the literature. The distribution of ratios of REM versus NREM AHI in our cohort are given in Figure 3A. Among those patients with ≥ 5 min of REM sleep (n = 285), about half had a REM AHI value ≥ 2-fold larger than the NREM AHI (n = 158); the me-dian ratio in this subset of REM dominant patients was 4.8 (IQ range 2.8-7.4), while the median REM dominance ratio in the whole group was 2.2 (IQ range 1.0-5.4). Among the small por-tion exhibiting a REM to NREM AHI ratio < 0.5 (i.e., > 2-fold NREM dominant; n = 32), the median ratio was 0.33 (IQ range 0.21-0.38). The remaining subjects (n = 95) had < 2-fold differ-ence between REM and NREM AHI values.

We also considered several subgroups to further explore REM dominance patterns. For subjects with at least 20, 40, or 60 min of REM sleep (n = 271, 218, or 158, respectively), the median REM dominance ratio was between 2.4 and 2.6, and the percentages of these subgroups with REM dominance ratios > 2 were between 45% and 60%. These values were similar

WakeREM

NREM

1hr

SupineLateral

AHI values

BA

Indep

R-dom

S-dom

S&R-dom

S&R-dom(all supine)

REMNREM

Supine

Latera

lSupi

ne REM

Total

11 11 11 11 11 11

25 11 13 16 25 14

15 17 21 11 21 16

23 17 24 11 46 18

46 20 27 46 27N/A

Figure 1—Schematic of stage- and position-dependent OSA

(A) Hypnogram from a 7-h PSG, simplifi ed to a single NREM sleep stage. The body position is shown as the dotted line. Tick marks indicate respiratory events. For visual clarity, each tick mark represents 2 apneas. Each row of tick marks is a different potential circumstance: Indep, position- and stage-independence; R-dom, REM dominance alone (2-fold); S-dom, supine dependence alone (2-fold); S&R-dom, supine dominance with worsening in REM only while supine (2-fold for each factor); S&R-dom (all supine), supine and REM dominance (2-fold each) assuming that the patient was supine for the whole night. (B) AHI values from each of the four conditions representing combinations of sleep stage and body position, as well as the total AHI (events/h).


NA Eiseman, MB Westover, JM Ellenbogen et al

to values obtained from group having ≥ 5 min of REM sleep reported above. Among women (n = 136), the REM dominance was greater, with a median REM dominance ratio of 4.2 (IQ range 1-8.3). Male subjects (n = 152) showed a smaller ratio of 1.9 (IQ range 0.6-4.3; p < 0.0001, Mann-Whitney, compared to the ratio in females). There was no significant correlation between age and the REM dominance ratio for the whole group (Pearson r = 0.02), or when females and males were analyzed separately (Pearson r = 0.04 and 0.09).

In addition to sleep stage effects, body position also influ-enced OSA severity. The distribution of supine to non-supine AHI ratios is given in Figure 3B. Among patients with ≥ 5% of TST in either supine or non-supine position (n = 250), 60% showed ≥ 2-fold greater AHI in the supine than the non-supine position (n = 149); the median ratio among these supine-domi-nant patients was 6.5 (IQ range 3.5-10.3), while the median ratio for the whole group was 3.2 (IQ range 1.1-6.9). Only 5 patients had a supine to non-supine AHI ratio < 0.5 (i.e., > 2-fold non-supine dominant OSA). The remaining patients had < 2-fold difference in AHI between supine and non-supine positions (n = 96). Prone sleep was observed in only 15 patients (5%); thus, we could not reliably evaluate the effect of this body posi-tion on apnea severity.

We also considered several subgroups to further explore su-pine dominance patterns. Among subjects with 5% to 95% of the time spent supine (n = 250), males showed higher median supine dominance ratios than females: 4.0 vs 2.2 (p < 0.005, Mann-Whitney). This male preference for supine dominance

persisted if we restricted analysis to those with 20% to 80% of sleep in the supine position (3.6 versus 2.2; p < 0.03, Mann-Whitney). Older age affected supine dominance, as there was a small but significant correlation between age and the supine dominance ratio, with Spearman r = 0.28-0.31 for the groups defined by spending 5% to 95% of sleep supine or 20% to 80% supine. Of note, the male subjects were slightly but significantly older in each of these subgroups, by a mean of approximately 5 years, which may have contributed to the male correlation.

AHI Underestimation Due to Reduced Time in REM SleepWe next considered how the summary (full-night) AHI val-

ue could be underestimated due to the underrepresentation of REM sleep that sometimes occurs due to a first night effect. We extrapolated the AHI by scaling individuals with < 25% REM sleep to a value of 25% of the TST. Those with no REM sleep (n = 8) were conservatively assigned a REM AHI of 30/h, which

Table 1—Population characteristicsAHI < 5 AHI ≥ 5

N 141 159Age 42 (32-53) 55 (45-64)Sex 46% male 57% maleBMI 27.0 (23.9-32.6) 29.6 (26.0-35.0)ESS 7 (4-11) 8 (5-12)TST (min) 393 (362-429) 388 (348-421)Efficiency (%) 90.0 (84.5-95.0) 91.0 (82.0-95.0)N1 (min) 43.0 (27.5-63.8) 51.5 (33-82.5)N1 (%) 10.9 (7.0-17.4) 13.9 (8.3-21.9)N2 (min) 225 (185-259) 208 (175-243)N2 (%) 57.2 (48.2-63.9) 55.0 (47.9-62.1)N3 (min) 52.0 (24.5-86.0) 45.0 (20.5-80.0)N3 (%) 13.8 (7.3-20.9) 12.5 (5.9-20.4)REM (min) 65.5 (44.5-85.3) 58.5 (33-84.5)REM (%) 16.8 (11.6-21.6) 15.4 (9.7-21.6)Latency (min) 4.5 (2-11.8) 3.0 (1-7.5)LPS (min) 12.5 (5.3-22) 9.5 (3.0-20.5)AI (/h) 14.2 (9.4-19.6) 23.5 (16.2-32.0)*PLMI (/h) 2.6 (0.2-13.8) 4.5 (0.9-18.9)O2 min (%) 90 (87-92) 83 (78-94)RDI (/h) 6.6 (3.5-12.1) 22 (14.8-28.2)*AHI (/h) 1.5 (0.6-3.5) 9.8 (7.0-15.2)*

Values are median with interquartile range. TST, total sleep time; ESS, Epworth Sleepiness Scale; LPS, latency to persistent sleep. *p < 0.05 Kruskal-Wallis with Dunn correction.

5 10 15 20 25 300

20

40

60

10

30

50

% o

f PSG

s

0

10

20

30

40

% R

EM

0

20

40

60

80

10

30

50

70

100

90

% S

upin

e

35

25

15

5

A B

CMinimum time spentin all 4 combinations

Figure 2—Distribution of sleep stage and body position in diagnostic PSGs

(A) Box and whiskers plot of the percent of the TST spent in REM sleep, showing the median value (central line), mean value (+), 25-75th percentiles (box edges), and 5% to 95% range (whiskers). (B) Box and whiskers plot of the percent of the TST spent in the supine position. (C) Percentage of PSGs containing minimum amounts of time (X-axis in minutes) in each of the four possible combinations of REM/NREM and supine/non-supine.



was at the 98th percentile of our cohort’s full-night AHI distri-bution. With these adjustments, the median AHI value of the cohort increased only slightly from 5.3/h (IQ range 1.8-10.4) to 6.2/h (IQ range 2.2-12.9), a change that was not significant (p > 0.1, Mann-Whitney; Figure 4B). Among those with a REM dominance ratio of ≥ 2 or ≥ 5, correction for time spent in REM sleep caused reclassification from a normal AHI to an AHI > 5 in 8% and 11% of cases, respectively. Reclassification from an AHI value of < 15 to > 15 occurred in 11% of those with either ≥ 2 or ≥ 5 REM dominance. Thus, while REM can clearly be associated with worsening apnea, the overall impact of this stage dependence on the full-night average AHI can be small. Although the population-level risk of AHI underestima-tion due to decreased time spent in REM sleep was small, indi-vidual patients can clearly be misclassified on this basis.

AHI Underestimation Due to Reduced Time in the Supine Position

Consider a single night PSG during which body position changed—how different might the AHI value be if the patient had slept exclusively supine? To answer this question, we ana-lyzed those patients with 20% to 80% of sleep time spent in the supine position (n = 185), using this more conservative cutoff

to increase confidence in the calculation of the supine domi-nance of AHI values. The supine dominance of this group had a median value of 2.7 (IQ range 1.1-6.7). Extrapolating the AHI values as if the patients were supine the whole night revealed a significant increase in the median AHI, from 5.1/h (IQ range 1.4-9.8) to 7.4/h (IQ range 2.4-15.9) (p < 0.0001, Kruskal-Wal-lis; Figure 4B). The impact of position change was greater than the impact of subnormal REM sleep described above. We also note the skew of the AHI values well into the severe OSA range with this adjustment. Below we will address the apnea severity misclassification problem suggested by this pattern.

Expected Benefit of Enforced Lateral SleepDespite the current limitations in monitoring and enforcing

lateral sleep in the home, it is useful to explore the potential for positional therapy based on the distribution of body positions and the relative protection afforded by non-supine sleep. In the subset of patients with 20% to 80% of sleep spent in the supine position, enforced lateral sleep would be expected to lower the median AHI significantly from 5.1/h (IQ range 1.4 to 9.8) to 1.1 (IQ range 1.0-3.9; Figure 4B). Within this group, 58% (n = 55) of those with observed AHI > 5/h (n = 95) would be predicted to normalize (AHI < 5/h) with enforced lateral sleep; similarly, 57% (n = 27) of those with observed AHI > 10 (n = 47) would be predicted to normalize to values < 10/h with enforced lat-eral sleep. This demonstrates a substantial theoretical benefit afforded by position therapy in patients with mild to moderate OSA as in this cohort.

BL

correc

ted0

10

20

30

5

15

25

AHI

observ

ed

100% su

p

100% lat

0

10

20

30

40

50

60

5

15

25

35

45

55

AHI

A BNS *

*

Figure 4—AHI underestimation related to time spent in REM and supine position

(A) Box and whiskers plots of the percent of the AHI (events/h) in our cohort in baseline conditions (“BL”) compared to that after adjustment for each individual who spent < 25% of TST in REM sleep (“corrected”). Median (central line), mean (+), 25-75th percentiles (box edges), and 5% to 95% range (whiskers) are shown. The groups are not different (Mann-Whitney rank test). (B) Box and whiskers plots of the AHI values related to body position. The baseline condition (“observed”) includes only those patients who spent 20% to 80% of the TST in the supine position. These values are compared to the supine AHI and the lateral AHI. *Significant difference by Kruskal-Wallis test with Dunn post-test.

0 5 10 15 20 250.0

0.1

0.2

0.3

30 40 50

AHI dominance ratio

Frac

tion

0 5 10 15 20 250.0

0.1

0.2

0.3

30 40 50

AHI dominance ratio

Frac

tion

A

B

REM:NREM ratio

Supine:Non-supine ratio

Figure 3—Distribution of stage- and position-dominance: AHI ratios

(A) Frequency histogram showing the relative occurrence (Y-axis) of REM: NREM AHI ratios. Larger values indicate increasing REM-dominant OSA. (B) Frequency histogram showing the relative occurrence of supine: non-supine AHI ratios. Larger values indicate increasing supine-dominant OSA.


NA Eiseman, MB Westover, JM Ellenbogen et alOSA Severity Misclassification Due to Position Dependence

The above results indicated that the within-patient adjust-ment for supine sleep position yields a larger effect in terms of potential underestimation at the population level compared to adjustment for the amount of REM sleep. Relative to the worst-case scenario of 100% supine sleep, our results predict OSA

category misclassifi cation due to position-dependent OSA com-bined with time spent non-supine (Figure 5). Among patients who spent < 50% of TST in the supine position (n = 99) and had an overall AHI < 5/h on their PSG (n = 53 of the 99), 14% were predicted to have AHI > 5 if they had spent the full TST supine (Figure 5B). Among patients with observed AHI < 15/h, 24% to 38% would be reclassifi ed to AHI > 15/h if they spent the full TST supine (depending on whether we use a cutoff for analysis of < 50% [n = 99] or < 25% [n = 32] of the TST spent supine). Among those with observed AHI < 30/h, 16% to 38% would be reclassifi ed as severe (> 30/h) had they spent the full night supine. These results suggest that in our cohort the poten-tial risk for misclassifi cation based on body position was sub-stantial and affected all OSA severity categories; in some cases, those with overall AHI values < 5/h had supine AHI values in the moderate or even severe range.

Visual Guide to Position Related Underestimation of AHI from a Single Night PSG

Next we generated a simple visual framework for con-sidering AHI underestimation given 2 factors obtained on routine laboratory PSG: the portion of the TST spent supine, and the relative protection afforded by non-supine sleep (Figure 6). We conducted simple calculations involving 3 typical patient groups with assumed supine AHI values of 10, 22.5, and 40/h. The percentage of TST spent supine and the relative protection afforded by non-supine sleep were varied to produce a spectrum of observed summary AHI values in each group. For example, a patient with severe OSA based on a supine AHI of 40 could be observed to have an AHI value in the mild or moderate range if lateral position affords at least 2-fold protection and supine position was < 50% of the TST (Figure 2c). These parameters are plausible, having oc-curred in 60 patients in our cohort (of the 90 with 5% to 50% of TST spent in supine position).

A B C1.0

0.8

0.6

0.4

0.2

0.00 20 40 60 80 100

1.0

0.8

0.6

0.4

0.2

0.00 20 40 60 80 100

1.0

0.8

0.6

0.4

0.2

0.00 20 40 60 80 100

Supine AHI = 10 Supine AHI = 22.5 Supine AHI = 40AHI

L:S

ratio

(sev

erity

)

L:S

ratio

(sev

erity

)

L:S

ratio

(sev

erity

)

Time supine (%) Time supine (%) Time supine (%)

40

30

20

10

0

5

55

15

15

30

Figure 6—Simulations of AHI underestimation due to body position

Contour plots illustrate observed AHI values (color gradient scale), when the supine AHI value is 10 (A), 22.5 (B), or 40 (C). In each panel, all possible combinations of the portion of TST spent supine (X-axis) and the relative protection of lateral position compared to supine in terms of AHI (Y-axis; L: S ratio of AHI values) are shown to provide the entire spectrum of possible underestimations in each case. The white contour lines indicate the clinical boundaries for OSA severity categories. In each panel, the observed AHI (color) equals the “true” AHI only when either there is no position dependence (top margin of each panel), or when the time spent supine is 100% (right margin of each panel). Underestimation of the AHI occurs whenever either of these conditions is violated, in proportion to the extent of position dependence of AHI and the proportion of the night spent supine.

0 2010 40 60 8030 50 70 90 100

% of cohort

None Mild Mod SevA

B

Figure 5—OSA category misclassifi cation: risk due to body position

(A) Bar plot of the baseline percentage (X-axis) of the cohort with AHI values in the following disease categories: none (AHI < 5; gray), mild (AHI 5-15; green), moderate (AHI 15-30; orange), and severe (AHI > 30; red).(B) Misclassifi cation for those subjects with either < 25% or < 50% of the TST spent in the supine position (n = 32). The percentage of the cohort that could switch categories based on extrapolation of supine AHI is given.

Category change < 25% supine < 50% supine< 5 to > 5 16% 14%< 5 to > 15 9% 3%< 5 to > 30 3% 1%< 15 to > 15 38% 24%< 15 to > 30 19% 6%< 30 to > 30 38% 16%


Sleep Apnea UnderestimationPhenotypic Misclassification in a Simulated Population

In the prior sections, we focused on the risk of underestimat-ing AHI values from a single PSG. We now explore the issue of misclassification by over- or underestimation, such as might occur in population studies undertaken for OSA genetics or cardiovascular risk associations. Single-night assessments that neglect position dependence could blur phenotype boundaries, which would increase variance, dilute the strength of associa-tions, and necessitate higher numbers of patients to identify re-lationships. We modeled 4 groups of patients with supine AHI values (mean ± SD) of 2.5 ± 1.25/h for normal, 10 ± 2.5/h for mild, and 22.5 ± 7.5/h for moderate apnea. The severe group was modeled by a skewed normal distribution with mean of 50/h, and 95% of values within the range of 30-90/h, to reflect the long tail toward higher AHI values, as seen in large studies such as the Sleep Heart Health Study (SHHS).

Values drawn from these baseline control distributions are shown in Figure 3A. Our intention is to provide a baseline distribution against which to compare the same groups with imposed position-related variance rather than to model actual population observations, which lack nadirs between clinical categories. The overlapping categories in this baseline simula-tion are entirely due to the imposed small degree of stochastic variance in the AHI; these effects are fairly small, with misclas-sification rates of ~5% (not shown).

In this initial model, each subject was also assigned a per-centage of the TST spent supine, but no position dependence was imposed. In Figure 3B, we used this place-holder value to introduce variance into each subject’s AHI value due to 2 factors: portion of the TST spent supine and supine dominance of the OSA severity. Supine versus non-supine time ratios were drawn from a normal distribution of ratios (1.13 ± 0.16), based on prior work.13-16 The supine-dominance of AHI was taken to be 2:1. Thus, each subject had their AHI value reassigned based on position dependence imposed upon their previously assigned portion of the TST spent supine. From this plausible source of variation, the overlap in clinical categories became more pro-nounced (Figure 7B). The OSA category misclassification is summarized in a confusion matrix (Figure 7c). The portion of patients remaining in their original categorical phenotype ranged from 64% to 93%, with 7% to 19% being adjusted to either one category lower or higher than their original category.

DISCUSSION

Reliance on a single night of sleep data continues to repre-sent a challenge to OSA diagnosis, decision making, and re-search phenotyping because a single night of sleep often does not provide sufficient examination of all combinations of sleep stages and body positions. This study highlights several impor-tant points relevant to interpretation of the single-night sleep study with regard to OSA classification: (1) we found a wide variation in the stage- and position dependence of OSA during routine diagnostic PSG testing; (2) we observed all four com-binations of stage (REM and NREM) and position (supine and non-supine) in only a fraction of PSGs; (3) supine dominance had a greater impact than REM dominance regarding the risk of underestimating AHI; (4) the risk of misclassification is sub-stantial, as indicated by models driven by even conservative

parameter estimates. Considering position dependence and REM dependence of the AHI provides a framework to guide patient-specific PSG interpretation, although certain limita-tions clearly remain that are inherent to the single-night assess-ment of OSA.

What Is the “True” AHI?The question may arise when patients reporting a lateral

sleep position preference in the home are requested to sleep

Histogram of AHI values obtained from a simulated cohort of n = 1,000 subjects in each of 4 categories: AHI < 5 (purple), AHI 6-15 (pink), AHI 16-30 (gold), and AHI > 31 (green). See methods for the distribution assumptions in this and subsequent panels. Histogram of AHI values from the same cohort as in panel A after imposing position dependence to the AHI for each individual. Note the change in X-axis range. Confusion matrix showing misclassification of individuals according to the distribution in panel B. Each row is the “true” AHI assigned to each group. Each column is the OSA category assigned given stochastic variation as well as position dependence. Gray shading indicates the correctly classified members of each group. The percentage of individuals misclassified is also given based on the standard categories, as well as a single cutoff value of AHI = 15.

A

B

C

Control conditions

Position dependence

# of s

ubjec

ts# o

f sub

jects

500

400

300

200

100

500

400

300

200

100

0 20 40 60 80 100

20 40 60 80 100 120 140

Figure 7—OSA severity categories misclassification

None Mild Mod Sev correct lower higher AHI 15

2.5 933 67 0 0 93% – 7%94%

10.0 115 765 120 0 77% 12% 12%

22.5 14 177 636 173 64% 19% 17%89%

50.0 6 15 140 839 84% 16% –


NA Eiseman, MB Westover, JM Ellenbogen et alsupine during their laboratory PSG. Such a patient might argue that their study’s AHI does not accurately represent their AHI at home, particularly if their supine AHI is more severe than other body positions, as is often the case. The true AHI, they would argue, should be that obtained from their more typical sleep po-sition (lateral, in this example). However, it is difficult to know about (or control) body position in sleep. Moreover, if one has severe apnea while supine, one might not want to risk the po-tential health consequences of OSA occurring during (perhaps unknown) time spent in the supine position. The true AHI in this clinical situation would be the weighted average of supine and lateral values according to a distribution of time spent in different positions over time. Since there is currently no fea-sible way to routinely measure this, the conservative approach would be to strongly weight the supine AHI in decision making.

From a research standpoint, we suggest that two types of misclassification are most deserving of attention, and which is more relevant would depend upon the goals of the research. From a sleep apnea genetics standpoint, patients should argu-ably be classified using metrics that capture AHI according to stage and position, as these may carry important phenotypic value. Genetic associations of OSA severity should not, in con-trast, depend on stochastic nightly variance in body position or REM% (which presumably are not genetic). On the other hand, in large population studies aiming to draw epidemiologi-cal links between OSA severity and cardiovascular events, the “true” AHI value should arguably take into account night-to-night variation (of any cause) occurring within individual pa-tients, because presumably the actual AHI value is contributing to the risk. In other words, the distribution of AHI values over time (from variance in position or other factors) is critical in linking OSA with clinical outcomes. To illustrate that the true AHI is context dependent, consider a theoretical patient with strongly supine-dominant OSA who sleeps exclusively in the lateral position in the home. Such a patient could be classified (appropriately) as mild or even normal in an epidemiological study of cardiovascular outcomes, while the same patient could also be classified (appropriately) as severe OSA in a genetic study of OSA risk.

Sleep Stage Dependence of OSASeveral reports have characterized the stage dependence

of OSA, and approximately half of patients with OSA were REM dominant.13-14,17-19 In our study, the distribution of stage dependence was quite large and non-normally distributed (Figure 3). Some individuals had quite large REM: NREM AHI ratios: 27% of subjects had a ratio > 5. The worsening of OSA severity in REM sleep is likely to be related to the accompanying atonia resulting in more prominent airway col-lapse as well as greater desaturation.20 Despite the fact that REM dominance can be quite substantial, the impact of sub-normal REM sleep times on the overall AHI values was small, due to the fact that REM occupies a minority of sleep time even under normal conditions.

Among patients with severe AHI values (e.g., 30-60/h or higher) reported in the literature, NREM dominance was the more common pattern.13,14,17,21 We could not assess this pattern in our cohort, which was comprised of mainly normal, mild, or moderate OSA patients. However, the reported NREM

dominance trend for severe OSA is not surprising, since 75% or more of the TST typically consists of NREM sleep. Thus, REM-related events are at a 3:1 “disadvantage” on average, compared to NREM-related events, in terms of contribution to the total night AHI (Supplemental Figure S1). For example, to achieve an overall AHI of 75, with a 3:1 REM-to-NREM AHI ratio, the REM AHI would have to be the nearly impossible rate of 150/h (and the NREM rate would be 50/h).

Despite the relatively lower impact of reduced REM% on the full-night AHI in our cohort, a substantial minority of pa-tients (~10%) exhibited underestimation sufficient to cause clinically important misclassification (using AHI cutoffs of 5 or 15). Patient-specific mitigation of this risk could be reasonably achieved by reassessing borderline cases as if the REM% were a normal (for example, 25%) portion of the TST. The finding of greater REM dominance in females than males suggests that sex should be considered in the clinical interpretation of over-all AHI values when subnormal REM% times are encountered. And while the focus of this paper is on the AHI which is itself a summary metric, it should be noted that respiratory pauses can be most severe in REM sleep, by virtue of event duration and oxygen desaturation, but these features are not captured by the AHI.22,23

Position Dependence of OSAThe basis for supine-related worsening of OSA severity

likely relates to gravity-driven collapse of the tongue and pha-ryngeal soft tissue to occlude the airway,15 and in this way may share similarities with REM-related worsening. In our cohort, the distribution of position dependence was quite broad and non-normally distributed (Figure 3), with large ratios > 5 seen in 36% of the subjects. Despite the likely contribution of body position change to variance in OSA severity, only some reports observed this correlation (table 2).7,24-26

For patients intolerant of CPAP who also have sufficient supine dominance, positional therapy may be considered as a conservative measure.27,28 However, positional therapies such as anti-snore shirts are of uncertain utility and long-term adher-ence is typically suboptimal.29,30 The lack of available body po-sition monitors contributes to this uncertainty. One recent study aimed to improve the efficacy and compliance with a novel vest-like device to enforce lateral sleep, which also contained an actigraphy monitor to track compliance.31 That study showed promising results in a small population (16 patients) of posi-tional patients, and suggests that improved positional devices may overcome some of the previously reported limitations. In our cohort, the expected benefit of enforced lateral sleep was striking, with > 60% of patients with an AHI > 5/h predicted to normalize (to < 5/h) given this intervention. Enforcing su-pine sleep during testing has the advantage of decreasing the potential risk of underestimating the AHI. However, it carries the liability that position dependence remains hidden, and thus one may not be able to establish the extent to which lateral sleep position might be protective. It is important to note that few of the approved home PSG devices report body position, which our data suggests is critical for interpretation of the AHI value.32

Interestingly, in comparison to REM dominance, the role of supine dominance appears to be a larger contributor to variance in AHI. This is not unexpected: the body position has a plau-



Table 2—Literature regarding night-to-night variability in OSAReference N PSG Loc Retest interval; population CommentsDean 199248 241 2 L 18 month retest interval

Repeat tests on initial negative studies if suspicion for OSA

~1.8-fold difference in supine % (but NS)3-fold more REM on 2nd night9 had OSA on retesting (7 of these severe)

Redline 1991*49 29 2 H OSA defined as RDI > 101-300 day retest interval

No first night effect seen in RDIHigh correlation (> 0.9) between studies

Meyer 19936 11 2 L 8 week retest intervalOSA defined as AHI > 5

Half were only positive on second nightOnly stage N1 percentage was predictor

Aarab 200925 15 4 L 1-9 week retest intervalKnown OSA (AHI > 5)

Group mean AHI unchanged on retestingHowever, large variation (60-80% of the mean)Body position may play a role

LeBon 20007 243 2 L Consecutive night retestingSuspected OSA referrals

High variability, highly skewed AHI values15-25% were misclassifiedMany patients met OSA criteria only on retestRole of stage and position implicated

Stepnowsky 200450 1091 3 H Consecutive night retestingSuspected OSA referrals

Consistent AHI across nights15% misclassification

Levendowski 200926 37 2 L 6 month retest intervalMild to moderate OSA

~0.6 r-value (AHI test-retest)Bias 8 AHI points worse on retestSupine correlation ~0.4 (% test-retest)

Levendowski 200926 37 2 H 6 month retest intervalMild to moderate OSA

~0.7 r-value (AHI test-retest)No bias on retesting in terms of AHISupine correlation ~0.7 (% test-retest)

Levendowski 200951 20 2 L > 1 month retest intervalSuspected OSA referrals

~0.4 r-value (AHI test-retest)Bias to AHI increase by 7 on retest25% increased AHI by > 20/hr~0.7 supine (test-retest %)

Bliwise 199152 71 2 L Consecutive night retestingElderly population

18% had nightly variability of > 10/hr AHIVariability greater for high AHI subjects12-25% crossed AHI cutoffs of 5 or 10Nasal obstruction predicted variability

Lord 199153 15 4 H Consecutive night retesting(repeated at 4-6 months)Elderly populationOSA not suspected

OSA defined by RDI > 150.6 Kappa for RDI across nightsClassification changed in 15% vs first night

Ahmadi 200954 193 2 L Consecutive night retestingSleep Clinic referrals

21% differed by > 5/hUsing RDI 15 cutoff: 20% were misclassified

Chediak 199655 37 2 L Consecutive night retestingMen undergoing evaluation for impotence

50% false negative for N1 (AHI 5 cutoff)AHI > 10 difference in 32% of subjectsNo architecture or supine determinants< 30% supine on either night

Mosko 198856 46 3 L Consecutive night retestingElderly population

OSA defined as AHI > 543% were misclassified vs first night

Gouveris 201057 130 2 L Consecutive night retestingKnown OSA patients

Mean AHI was unchangedUsing AHI > 10 cutoff, 6% were misclassified

Aber 198958 14 2 L Consecutive night retestingElderly men

Mean AHI was unchangedUsing AHI > 5 cutoff, 35% were misclassified

Fietze 200459 35 7 H Consecutive night retestingAll had ODI 5-30 at baseline

Using ODI > 15 cutoff, 14% were misclassifiedODI variance not related to stage, position, BMI

Bittencourt 200124 20 4 L Consecutive night retestingOSA based on AHI > 10

Mean AHI stable across nights50% changed classification65% had AHI vary by > 10/hBody position influenced variance (as did stage)

Quan 200241 91 2 H < 4 month retest intervalSubset of the SHHS

No bias using 3 or 4% desaturationIntra-class correlation coefficient ~0.812-21% misclassified using 5, 10, 15 cutoffREM% varied (position not measured)

Davidson 200360 44 2-3 H Consecutive night retestingKnown OSA

Group mean AHI not different on re-testing7-11% misclassified for AHI (15 or 30 cutoff)The AHI SD on retesting was ~10/h

N, number of subjects studied; #, number of PSGs recorded per subject; Loc, location of PSG; H, home study; L, laboratory study; ODI, oxygen desaturation index; NS, not significant; SD, standard deviation; *home device validation also performed.


NA Eiseman, MB Westover, JM Ellenbogen et alsible range of time spent supine from 0-100% of the TST, while the plausible range of time spent in REM sleep is more restrict-ed, to about 10% to 30% of the TST (Supplemental Figure S1). This distinction may not hold for split-night studies, in which the AHI during 2-3 hours of baseline sleep might be more sen-sitive to the presence or absence of a REM period (although the AHI calculated from “simulated” split nights derived from full diagnostic PSGs had high reported correlation with the full night AHI in previous work33).

Finally, supine dominance may be a useful marker of more than just apnea severity. Chervin et al. reported that the supine AHI correlated better with sleepiness than the overall AHI in a large retrospective study.34 From a treatment standpoint, posi-tion dependence might serve as a predictor for the efficacy of weight loss or dental appliance interventions,28 although BMI has been reported as an inverse predictor of position depen-dence.35,36 Neither a recent meta-analysis nor the practice pa-rameter regarding surgical treatments for OSA lists position dependence as a predictor of efficacy.37,38 The degree of supine dependence was influenced by sex and age in our cohort, sug-gesting these factors should be considered in the evaluation of position-dependence in future work. It is also important to rec-ognize that the time spent supine during a PSG may not be an accurate reflection of the time spent supine during sleep in the home. Due to the routine practice of encouraging supine sleep during clinical studies, the portion supine may in some cases be an overestimate compared to home sleep patterns. Despite this uncertainty, as discussed herein, a conservative approach would be to determine how severe apnea could be while supine, to avoid under-diagnosis.

Night-to-Night Variabilitytable 2 illustrates the spectrum of reported variability

across two or more nights of testing. The literature varies in terms of population, number of PSG assessments, time be-tween tests, and location of testing. In general, these studies found similar mean AHI values across study nights and good test-retest correlations. However, it is important to note that, despite the reported stability of group means, the individual variability was actually quite large. In order for stable group mean values and high individual misclassifications to both occur, there must be a relative balance among those with in-creasing, decreasing, or unchanged event rates between the first and subsequent nights of assessment. This serves as a re-minder that variability can lead to under- or overestimation of OSA severity for any given individual, even if group means appear stable in test-retest studies. It is also worth noting that apnea classification issues extend beyond the issues covered in this report and those surrounding night to night variabil-ity—for example, differences in apnea/hypopnea definitions can alter the apnea indices.39

For reasons of cost and convenience, it is common for large clinical studies to involve single night PSG to establish the diag-nosis and severity classification of OSA.40 Although test-retest reliability was reportedly good when home PSG was repeated in a small subsample of the SHHS (n = 91),41 characterizing the phenotype of OSA patients is not straightforward due to heterogeneity of the disease as well as the manner in which it is clinically characterized in various studies.42,43

Clinical ImplicationsAs in many aspects of diagnostic medicine, for practical rea-

sons one must often assign cutoff values to continuous vari-ables, such as the AHI, even when the population distribution does not have clear statistical demarcations to guide these cut-offs. Categorical OSA clinical classifications have facilitated links to sleepiness44 and to cardiovascular risk.2,3 Yet it should be emphasized that considerations regarding how accurate one must be in determining the AHI depends in part on the context. Whether a sleepy patient has an AHI of 20 or 80 per hour on their diagnostic PSG will not change the fact that CPAP is the gold standard treatment. However, patients meeting criteria for OSA while supine may escape diagnosis because only a portion of the PSG was spent in the supine position, which reduced the summary AHI value.

One major limitation of the single-night study of sleep, re-garding the ability to adjust for stage- and position-dependence, is that one may not observe all four combinations of REM/NREM and supine/non-supine sleep. For patients who spend the entire PSG supine and have a normal amount of REM sleep, the apnea severity is unlikely to be underestimated. However, even under these conditions that insulate against underestima-tion, some information is lost, as one is still left with uncer-tainty as to relative protection afforded by lateral sleep (should a position approach be attempted). For patients with position dependence, the contour plots of Figure 5 can be used as a guide to estimating potential underestimation. For those with an unexpectedly negative single night PSG, considering the pre-test probability of OSA should drive selective decisions to repeat testing.8

Underestimating the AHI has implications for individual pa-tients in whom the diagnosis and treatment decisions are made based on a single study. Repeat testing may be useful in the short term to confirm unexpected results, or over longer time frames to assess worsening.45,46 It is also worth mentioning that while the diagnosis and severity categorization formally rests upon the AHI, this metric does not take into account other po-tentially important features such as length of event, associated cardiovascular response, depth of desaturation, or fragmenta-tion of EEG metrics of sleep.

The main limitation of our study is that we were limited to single-night PSG, as is customary in the clinical laborato-ry setting. Ideally, repeated nights of observation would pro-vide important information as to the impact of differences in REM sleep and body position, assuming that these parameters changed from night to night. Such a study would allow impor-tant validation to the predictions made herein. In addition, we note that our selected population of mainly mild-to-moderate apnea is more likely to show position dependence compared to patients with severe apnea. Thus, the extent to which the posi-tion-related classification issues discussed here may generalize to more severe sleep disordered breathing remains uncertain. By extension, our finding that position is more relevant than sleep stage may not generalize to populations with more severe AHI values, in which position dependence is less apparent.35,47 These patterns however emphasize the potential role of posi-tional therapy in those with mild or moderate sleep apnea.

In summary, supine dominance and REM dominance are important risk factors for underestimating the AHI that can be


Sleep Apnea Underestimationpotentially recognized and accounted for in a patient-specific manner from a single night’s study of sleep. Although home monitoring devices may facilitate repeat testing to address cer-tain contributors to AHI variance, many of these do not mea-sure sleep stage or body position,32 thus limiting their value. When position- and/or stage-dependence is evident in a single night PSG, the conditional AHI values should be considered regarding the potential for OSA misclassification, in particular the risk of underestimation of severity or missing the diagnosis altogether. We suggest that routine PSG interpretation should include not only the average AHI and RDI values, but also the details of respiratory event frequency by position and sleep stage. This would enable clinicians to extrapolate the positional data across the spectrum of full-supine sleep (presumed to be the worst case scenario) to the full-lateral sleep position for the purposes of positional therapy considerations. In addition, it would facilitate incorporation of stage-dependent data (perhaps in combination with position data) in the case of underrepresen-tation of REM sleep in particular. Routine reporting at this level will provide improved patient-specific sleep apnea phenotyping and thus improved management.

REFERENCES1. Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM, Mohsenin V.

Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med 2005;353:2034-41.

2. Eastwood PR, Malhotra A, Palmer LJ, et al. Obstructive sleep apnoea: from pathogenesis to treatment: Current controversies and future directions. Respi-rology 2010;15:587-95.

3. Budhiraja R, Budhiraja P, Quan SF. Sleep-disordered breathing and cardiovas-cular disorders. Respir Care 2010;55:1322-32; discussion 30-2.

4. Epstein LJ, Kristo D, Strollo PJ Jr., et al. Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med 2009;5:263-76.

5. American Academy of Sleep Medicine. International Classification of sleep dis-orders: diagnostic and coding manual. 2nd ed. Westchester, IL: American Acad-emy of Sleep Medicine, 2005.

6. Meyer TJ, Eveloff SE, Kline LR, Millman RP. One negative polysomnogram does not exclude obstructive sleep apnea. Chest 1993;103:756-60.

7. Le Bon O, Hoffmann G, Tecco J, et al. Mild to moderate sleep respiratory events: one negative night may not be enough. Chest 2000;118:353-9.

8. Bianchi MT. Screening for obstructive sleep apnea: Bayes weighs in. The Open Sleep Journal 2009;2:56-9.

9. Toussaint M, Luthringer R, Schaltenbrand N, et al. Changes in EEG power den-sity during sleep laboratory adaptation. Sleep 1997;20:1201-7.

10. Silber MH, Ancoli-Israel S, Bonnet MH, et al. The visual scoring of sleep in adults. J Clin Sleep Med 2007;3:121-31.

11. Oksenberg A, Arons E, Snir D, Radwan H, Soroker N. Cheyne-Stokes respiration during sleep: a possible effect of body position. Med Sci Monit 2002;8:CS61-5.

12. Eiseman NA, Westover MB, Mietus JE, Thomas RJ, Bianchi MT. Classifica-tion algorithms for predicting sleepiness and sleep apnea severity. J Sleep Res 2012;21:101-12.

13. Haba-Rubio J, Janssens JP, Rochat T, Sforza E. Rapid eye movement-related disordered breathing: clinical and polysomnographic features. Chest 2005;128:3350-7.

14. Liu Y, Su C, Liu R, et al. NREM-AHI greater than REM-AHI versus REM-AHI greater than NREM-AHI in patients with obstructive sleep apnea: clinical and polysomnographic features. Sleep Breath 2011;15:463-70.

15. Oksenberg A, Silverberg DS. The effect of body posture on sleep-related breath-ing disorders: facts and therapeutic implications. Sleep Med Rev 1998;2:139-62.

16. Cartwright RD. Effect of sleep position on sleep apnea severity. Sleep 1984;7:110-4.

17. Siddiqui F, Walters AS, Goldstein D, Lahey M, Desai H. Half of patients with obstructive sleep apnea have a higher NREM AHI than REM AHI. Sleep Med 2006;7:281-5.

18. Loadsman JA, Wilcox I. Is obstructive sleep apnoea a rapid eye movement-predominant phenomenon? Br J Anaesth 2000;85:354-8.

19. Oksenberg A, Arons E, Nasser K, Vander T, Radwan H. REM-related obstructive sleep apnea: the effect of body position. J Clin Sleep Med 2010;6:343-8.

20. Malhotra A, White DP. Obstructive sleep apnoea. Lancet 2002;360:237-45.21. Muraki M, Kitaguchi S, Ichihashi H, et al. Apnoea-hypopnoea index during rapid

eye movement and non-rapid eye movement sleep in obstructive sleep apnoea. J Int Med Res 2008;36:906-13.

22. Oksenberg A, Khamaysi I, Silverberg DS. Apnoea characteristics across the night in severe obstructive sleep apnoea: influence of body posture. Eur Respir J 2001;18:340-6.

23. Oksenberg A, Khamaysi I, Silverberg DS, Tarasiuk A. Association of body posi-tion with severity of apneic events in patients with severe nonpositional obstruc-tive sleep apnea. Chest 2000;118:1018-24.

24. Bittencourt LR, Suchecki D, Tufik S, et al. The variability of the apnoea-hypop-noea index. J Sleep Res 2001;10:245-51.

25. Aarab G, Lobbezoo F, Hamburger HL, Naeije M. Variability in the apnea-hypop-nea index and its consequences for diagnosis and therapy evaluation. Respira-tion 2009;77:32-7.

26. Levendowski D, Steward D, Woodson BT, Olmstead R, Popovic D, Westbrook P. The impact of obstructive sleep apnea variability measured in-lab versus in-home on sample size calculations. Int Arch Med 2009;2:2.

27. van Mannen JP, Richard W, van Kesteren ER, et al. Evaluation of a new simple treatment for positional sleep apnoea patients. J Sleep Res 2012;21:322-9.

28. Randerath WJ, Verbraecken J, Andreas S, et al. Non-CPAP therapies in obstruc-tive sleep apnoea. Eur Respir J 2011;37:1000-28.

29. Oksenberg A, Silverberg D, Offenbach D, Arons E. Positional therapy for ob-structive sleep apnea patients: A 6-month follow-up study. Laryngoscope 2006;116:1995-2000.

30. Bignold JJ, Deans-Costi G, Goldsworthy MR, et al. Poor long-term patient com-pliance with the tennis ball technique for treating positional obstructive sleep apnea. J Clin Sleep Med 2009;5:428-30.

31. Heinzer RC, Pellaton C, Rey V, et al. Positional therapy for obstructive sleep ap-nea: An objective measurement of patients’ usage and efficacy at home. Sleep Med 2012;13:425-8.

32. Collop NA, Tracy SL, Kapur V, et al. Obstructive sleep apnea devices for out-of-center (OOC) testing: technology evaluation. J Clin Sleep Med 2011;7:531-48.

33. Khawaja IS, Olson EJ, van der Walt C, et al. Diagnostic accuracy of split-night polysomnograms. J Clin Sleep Med 2010;6:357-62.

34. Chervin RD, Aldrich MS. Characteristics of apneas and hypopneas during sleep and relation to excessive daytime sleepiness. Sleep 1998;21:799-806.

35. Mo JH, Lee CH, Rhee CS, Yoon IY, Kim JW. Positional dependency in Asian patients with obstructive sleep apnea and its implication for hypertension. Arch Otolaryngol Head Neck Surg 2011;137:786-90.

36. Oksenberg A, Silverberg DS, Arons E, Radwan H. Positional vs nonpositional obstructive sleep apnea patients: anthropomorphic, nocturnal polysomnograph-ic, and multiple sleep latency test data. Chest 1997;112:629-39.

37. Aurora RN, Casey KR, Kristo D, et al. Practice parameters for the surgical modifications of the upper airway for obstructive sleep apnea in adults. Sleep 2010;33:1408-13.

38. Caples SM, Rowley JA, Prinsell JR, et al. Surgical modifications of the upper airway for obstructive sleep apnea in adults: a systematic review and meta-analysis. Sleep 2010;33:1396-407.

39. Ruehland WR, Rochford PD, O’Donoghue FJ, Pierce RJ, Singh P, Thornton AT. The new AASM criteria for scoring hypopneas: impact on the apnea hypopnea index. Sleep 2009;32:150-7.

40. Quan SF, Howard BV, Iber C, et al. The Sleep Heart Health Study: design, ratio-nale, and methods. Sleep 1997;20:1077-85.

41. Quan SF, Griswold ME, Iber C, et al. Short-term variability of respiration and sleep during unattended nonlaboratory polysomnography--the Sleep Heart Health Study. Sleep 2002;25:843-9.

42. Riha RL, Gislasson T, Diefenbach K. The phenotype and genotype of adult ob-structive sleep apnoea/hypopnoea syndrome. Eur Respir J 2009;33:646-55.

43. Varvarigou V, Dahabreh IJ, Malhotra A, Kales SN. A review of genetic associa-tion studies of obstructive sleep apnea: field synopsis and meta-analysis. Sleep 2011;34:1461-8.

44. Gottlieb DJ, Whitney CW, Bonekat WH, et al. Relation of sleepiness to respira-tory disturbance index: the Sleep Heart Health Study. Am J Respir Crit Care Med 1999;159:502-7.

45. Pendlebury ST, Pepin JL, Veale D, Levy P. Natural evolution of moderate sleep apnoea syndrome: significant progression over a mean of 17 months. Thorax 1997;52:872-8.

46. Berger G, Berger R, Oksenberg A. Progression of snoring and obstructive sleep apnoea: the role of increasing weight and time. Eur Respir J 2009;33:338-45.


NA Eiseman, MB Westover, JM Ellenbogen et al47. Oksenberg A, Dynia A, Nasser K, Gadoth N. Obstructive sleep apnoea in adults:

body postures and weight changes interactions. J Sleep Res 2011 [Epub ahead of print]

48. Dean RJ, Chaudhary BA. Negative polysomnogram in patients with obstructive sleep apnea syndrome. Chest 1992;101:105-8.

49. Redline S, Tosteson T, Boucher MA, Millman RP. Measurement of sleep-related breathing disturbances in epidemiologic studies. Assessment of the validity and reproducibility of a portable monitoring device. Chest 1991;100:1281-6.

50. Stepnowsky CJ Jr., Orr WC, Davidson TM. Nightly variability of sleep-disordered breathing measured over 3 nights. Otolaryngol Head Neck Surg 2004;131:837-43.

51. Levendowski DJ, Zack N, Rao S, et al. Assessment of the test-retest reliability of laboratory polysomnography. Sleep Breath 2009;13:163-7.

52. Bliwise DL, Benkert RE, Ingham RH. Factors associated with nightly variability in sleep-disordered breathing in the elderly. Chest 1991;100:973-6.

53. Lord S, Sawyer B, O’Connell D, et al. Night-to-night variability of disturbed breathing during sleep in an elderly community sample. Sleep 1991;14:252-8.

54. Ahmadi N, Shapiro GK, Chung SA, Shapiro CM. Clinical diagnosis of sleep ap-nea based on single night of polysomnography vs. two nights of polysomnogra-phy. Sleep Breath 2009;13:221-6.

55. Chediak AD, Acevedo-Crespo JC, Seiden DJ, Kim HH, Kiel MH. Nightly vari-ability in the indices of sleep-disordered breathing in men being evaluated for impotence with consecutive night polysomnograms. Sleep 1996;19:589-92.

56. Mosko SS, Dickel MJ, Ashurst J. Night-to-night variability in sleep apnea and sleep-related periodic leg movements in the elderly. Sleep 1988;11:340-8.

57. Gouveris H, Selivanova O, Bausmer U, Goepel B, Mann W. First-night-effect on polysomnographic respiratory sleep parameters in patients with sleep-disordered breathing and upper airway pathology. Eur Arch Otorhinolaryngol 2010;267:1449-53.

58. Aber WR, Block AJ, Hellard DW, Webb WB. Consistency of respiratory measure-ments from night to night during the sleep of elderly men. Chest 1989;96:747-51.

59. Fietze I, Dingli K, Diefenbach K, et al. Night-to-night variation of the oxygen desaturation index in sleep apnoea syndrome. Eur Respir J 2004;24:987-93.

60. Davidson TM, Gehrman P, Ferreyra H. Lack of night-to-night variability of sleep-disordered breathing measured during home monitoring. Ear Nose Throat J 2003;82:135-8.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication January, 2012Submitted in final revised form May, 2012Accepted for publication May, 2012Address correspondence to: Dr. Matt T. Bianchi, Wang 7 Neurology, Massachusetts General Hospital, 55 Fruit Street, Boston, MA 02114; Tel: (617) 724-7426; Fax: (617) 724-6513; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. Dr Bianchi and Dr Ellenbogen receive

funding from the Department of Neurology, Massachusetts General Hospital. Dr Bian-chi receives funding from a Young Clinician Award from the Center for Integration of Medicine and Innovative Technology. Dr Bianchi has a patent pending on a home sleep monitoring device. The other authors have indicated no financial conflicts of interest.


Introduction: Obstructive sleep apnea (OSA) is a common co-morbid condition in patients with posttraumatic stress disorder (PTSD); insuffi ciently treated OSA may adversely impact out-comes. Sleep fragmentation and insomnia are common in PTSD and may impair CPAP adherence. We sought to determine the impact of combat-related PTSD on CPAP adherence in soldiers.Methods: Retrospective case-control study. Objective mea-sures of CPAP use were compared between OSA patients with and without PTSD. Groups were matched for age, BMI, and apnea-hypopnea index (AHI).Results: We included 90 patients (45 Control, 45 PTSD). Among the cohort, mean age was 39.9 ± 11.2, mean BMI 27.9 ± 8.0, mean ESS 13.6 ± 5.7, and mean AHI 28.2 ± 22.4. There was a trend towards a higher rate of comorbid insomnia among patients with PTSD (25.8% vs. 11.1%, p = 0.10). PTSD was as-sociated with signifi cantly less use of CPAP. Specifi cally, CPAP was used on 61.4% ± 22.2% of nights in PTSD patients com-

pared with 76.8% ± 16.4% in patients without PTSD (p = 0.001). Mean nightly use of CPAP was 3.4 ± 1.2 h in the PTSD group compared with 4.7 ± 2.2 h among controls (p < 0.001). Regular use of CPAP (> 4 h per night for > 70% of nights) was also lower among PTSD patients (25.2% vs. 58.3%, p = 0.01).Conclusion: Among soldiers with OSA, comorbid PTSD was associated with signifi cantly decreased CPAP adherence. Given the potential for adverse clinical outcomes, resolution of poor sleep quality should be prioritized in the treatment of PTSD and potential barriers to CPAP adherence should be overcome in patients with comorbid OSA.Keywords: Posttraumatic stress order, sleep apnea, obstruc-tive sleep apnea, CPAP, CPAP compliance, CPAP adherence, sleep disordered breathing, deployment related sleep disordersCitation: Collen JF; Lettieri CJ; Hoffman M. The impact of post-traumatic stress disorder on CPAP adherence in patients with obstructive sleep apnea. J Clin Sleep Med 2012;8(6):667-672.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Insomnia, nightmares, and sleep fragmentation are hallmark features of posttraumatic stress disorder (PTSD). Poor sleep

quality is commonly reported in patients with PTSD and can po-tentiate symptoms of anxiety and depression.1 In addition, sleep avoidance, sleep terrors, nocturnal anxiety attacks, dream enact-ing behavior, and periodic limb movement disorder (PLMD) are also more frequent in these individuals. Concomitant sleep disor-ders have been shown to independently worsen outcomes. Com-pared to patients without sleep complaints, PTSD patients with sleep disorders experience an increased severity of depression,2suicidality,2 psychiatric distress,3 poor quality of life and function-ing,3 poorer perceived physical health,4 and substance abuse.5-7

Obstructive sleep apnea (OSA) is common among patients with underlying psychiatric conditions. The prevalence of OSA is higher among patients with PTSD than the general popula-tion.8-13 In patients with comorbid OSA and PTSD, adequate treatment of sleep disordered breathing has been shown to improve anxiety, depression, and other PTSD-related symp-toms.13,14 Similarly, unrecognized or insuffi ciently treated OSA is associated with worse outcomes.9,11,15 Recognizing and treat-ing OSA in patients with PTSD is crucial in optimizing the therapeutic response.

While continuous positive airway pressure (CPAP) therapy is the most effective treatment of OSA, acceptance and ad-herence of CPAP therapy are problematic in this population. CPAP adherence has been observed to be reduced in patients with comorbid psychiatric disease.16 Prior studies have docu-

The Impact of Posttraumatic Stress Disorder on CPAP Adherence in Patients with Obstructive Sleep Apnea

Jacob F. Collen, M.D.; Christopher J. Lettieri, M.D., F.A.A.S.M.; Monica Hoffman, M.D.Pulmonary, Critical Care and Sleep Medicine, Walter Reed National Military Medical Center, Bethesda, MD

mented lower CPAP adherence in patients with psychological disease and demonstrated relationships between low therapeu-tic adherence and anxiety and depression.11,17 Misperception of symptoms, overlapping symptoms of depression, and atypical presentations of OSA may limit acceptance of the diagnosis and need for treatment.9 Poor adherence with therapy is more common among PTSD patients in particular. Insomnia, sleep fragmentation, and recurrent awakenings common to PTSD may limit adherence with CPAP.9 Poor sleep quality may cre-ate additional barriers to CPAP therapy and further compromise therapeutic adherence. Similarly, nightmares, mask discomfort, air hunger, and claustrophobia are correlated with poor CPAP adherence among PTSD patients.18

A recent study in an older veterans population demonstrated that CPAP adherence was signifi cantly reduced in veterans with

BRIEF SUMMARYCurrent Knowledge/Study Rationale: There has been a dramatic rise in diagnoses of PTSD in US military combat veterans over the past de-cade. PTSD often coexists with both sleep-disordered breathing and in-somnia, potentially worsening clinical outcomes. While CPAP is an effec-tive therapy for obstructive sleep apnea, compliance is often worsened in patients with PTSD.Study Impact: In a cohort of young combat veterans with PTSD, sleep-disordered breathing signifi cantly worsened adherence with CPAP. Given the particular vulnerability of this population to worsened clinical outcomes, measures that improve CPAP adherence should be a priority.


JF Collen, CJ Lettieri and M Hoffmanconcomitant PTSD.18 The authors found that excessive sleepi-ness was predictive of improved use of CPAP, while nightmares correlated with poorer CPAP adherence. Given the increased prevalence of OSA among patients with PTSD and the adverse impact of untreated OSA on clinical outcomes, understanding how PTSD affects CPAP adherence is critical.

With the rising prevalence of PTSD among soldiers return-ing from combat deployments in support of Operations Iraqi Freedom (OIF) and Enduring Freedom (OEF), we sought to determine the impact of combat-related PTSD on CPAP adher-ence in patients with underlying OSA.

METHODS

Study DesignWe conducted an observational, case-controlled study to assess

the effects that PTSD had on objective measures of CPAP adher-ence among patients with concomitant OSA. The protocol was approved by our institution’s Department of Clinical Investiga-tion (Scientific Review Committee, Human Use Committee and Institutional Review Board: Exempt Protocol IRB#352703-1). No external funding was utilized to complete this study.

PatientsWe included consecutive adult patients with both PTSD

and OSA evaluated at the Walter Reed Army Medical Center Sleep Disorders Center between January and October 2009. We did not include soldiers with traumatic brain injury, as we felt this could potentially confound our results. Otherwise, no records were excluded from this analysis. For direct compari-son, we included an equal number of patients without PTSD who were diagnosed with OSA and initiated CPAP therapy during the same time period. Control subjects were matched for age, gender, body mass index (BMI), and apnea-hypopnea index (AHI).

All included patients initiated CPAP therapy for the treat-ment of obstructive sleep apnea syndrome. OSA was diagnosed by level I polysomnography in accordance with American Academy of Sleep Medicine (AASM) criteria in all patients.19 All polysomnographic studies were interpreted by the study in-vestigators in accordance with established AASM criteria.20

Soldiers with combat-related PTSD were evaluated at our hospital after returning from deployment. All patients under-went a structured clinical interview by a doctoral-level behav-ioral health provider and were found to meet diagnostic criteria for PTSD, in accordance with both the Diagnostic and Statis-tical Manual of Mental Disorders, Fourth Edition, Text Revi-sion (DSM-IV-TR) criteria and a standardized military PTSD screening tool (the PTSD Checklist-Military Version, PCL-M Scoring Criteria).21,22 The PTSD Checklist-Military Version in-cludes 17 questions (1-5 points each) addressing the 17 DSM-IV criteria for PTSD. A cutoff score of 50 points was used to assess for the presence of PTSD (17-85 points possible).21 This is a self-report questionnaire, and the results were interpreted by the administering provider.

Patients were categorized as having insomnia if they had a subjective sleep latency ≥ 30 min during the majority of nights and/or subjective sleep fragmentation (nocturnal awakening)

associated with daytime impairment not better explained by sleep disordered breathing, pain, or other more likely identifi-able factors, in accordance with DSM-IV criteria for the diag-nosis of insomnia.22

To prevent confounding, all patients received the same CPAP device (Respironics System One Auto, Phillips-Respironics, Murrysville, PA), and the same clinical evaluations, follow-up assessments, and education regarding OSA and proper use of CPAP. All patients underwent formal mask fitting prior to initi-ating CPAP therapy.

Data MeasurementsData used in this analysis were obtained from the initial sleep

consultation, follow-up evaluations, and polysomnographic stud-ies. For each patient we collected demographic, clinical, poly-somnographic, and objectively measured CPAP adherence data. Demographic variables included age, gender, and BMI. Subjec-tive sleepiness was assessed using the Epworth Sleepiness Scale (ESS)23 and the fatigue analog scale. Clinical data included comor-bid insomnia and chronic use of sedating or psychoactive medica-tions (≥ 3 months). Polysomnographic data included the AHI and SpO2 nadir observed during a diagnostic polysomnogram.

CPAP use was objectively measured in all patients during their initial follow-up evaluation 4-6 weeks after initiating ther-apy using a downloadable monitoring smart-card (Respironics Encore Anywhere). Specifically, we recorded the percentage of nights CPAP was used, the mean hours of CPAP use per night for all nights, and the mean hours of CPAP use per night during nights used. We also measured the rate of regular use of CPAP between groups, which we defined as CPAP use > 4 h/night on > 70% of nights.24

EndpointsThe primary endpoint was the difference in the absolute use

of CPAP between the 2 groups. The impact of chronically used sedating medications on CPAP adherence and the rates of regu-lar use of CPAP served as secondary endpoints.

Statistical AnalysisData are presented as the mean ± one standard deviation.

Comparisons between categorical variables were performed using the χ2 test, and continuous variables were assessed us-ing independent samples t-tests. P values < 0.05 were assumed to represent statistical significance. Data were analyzed using PASW 17.0 (SPSS Inc, Chicago, IL).

RESULTS

We included 90 patients with newly diagnosed OSA who ini-tiated CPAP therapy (45 with combat-related PTSD and 45 con-trols). Among the cohort, the mean age was 37.7 ± 10.3 years, and the majority of patients (84.8%) were men. The mean BMI was 27.1 ± 6.6 kg/m2 and the mean ESS was 13.8 ± 4.2. The majority of patients had moderate to severe OSA, with a mean AHI of 29.3 ± 16.1 events/h among the entire cohort.

As expected, patients with PTSD tended to have less sub-jective sleepiness and fatigue. In addition, there were twice as many patients with insomnia in the group with PTSD (25.8% vs. 11.1%, p = 0.10); however, given the overall small cohort


Impact of PTSD on CPAP Adherence in OSA

size, this difference did not reach statistical significance. The comparison groups were otherwise similar at baseline (table 1).

PTSD was associated with significantly less CPAP use (table 2). Specifically, CPAP was used on 61.4% ± 22.2% of nights in patients with PTSD compared with 76.8% ± 16.4% in patients without PTSD (p = 0.01). Similarly, the mean nightly use of CPAP during nights used was only 3.4 ± 1.2 h/night in the PTSD group, compared with 4.7 ± 2.2 h/night among controls (p < 0.001). Dur-ing all nights, patients with PTSD used CPAP for only 2.5 ± 1.8 h/night versus 4.2 ± 2.1 h/night in the control group (p < 0.001). Regular use of CPAP was also significantly less common among patients with PTSD and was observed in 25.2%, compared with 58.3% among patients without PTSD (p = 0.01).

The majority (82.9%) of patients with PTSD were chroni-cally using sedating medications (≥ 3 months) for the treatment of their PTSD and/or comorbid insomnia. Greater adherence with CPAP was observed among patients with PTSD who were chronically using sedatives. Among patients chronically using sedating agents, CPAP was used for 70.4% ± 19.8% of nights compared with only 46.2% ± 24.4% in patients not prescribed sedating medications (p = 0.009). Similarly, the mean nightly use of CPAP during nights used was 4.1 ± 1.9 h versus 2.6 ± 1.2 h (p = 0.006) in patients using versus not using sedative med-ications, respectively. Regular use of CPAP was observed in 34.6% of patients using sedatives and only 9.1% of patients not using sedating agents (p = 0.09). Improvements in CPAP adher-ence did not differ between patients using non-benzodiazepine sedative hypnotics, benzodiazepines, or atypical antipsychotics. Furthermore, the AHI and SpO2 nadir did not differ between pa-tients using and not using sedative agents, and these measures did not differ between the different classes of agents. However, compared to patients using non-benzodiazepines, patients using benzodiazepines or atypical antipsychotics had greater BMIs (26.3 ± 12.9 versus 29.9 ± 4.4 kg/m2, p = 0.007) and more sub-jective sleepiness as measured by the ESS (11.1 ± 6.1 versus 14.3 ± 5.7, p < 0.001).

DISCUSSION

We found that adherence with CPAP was significantly low-er among soldiers with combat-related PTSD than controls. CPAP was used on fewer nights and for approximately one

hour less per night when it was used. In addition, regular use of CPAP was observed in half as many patients with PTSD as the control group.

Concomitant insomnia tended to be more common in pa-tients with PTSD, and while this difference did not reach statis-tical significance, it may have contributed to the discrepancy in CPAP use. Similar to prior reports in patients with OSA without PTSD, comorbid insomnia can create a barrier to CPAP thera-py, as patients may experience greater difficulties initiating and maintaining sleep.25-28 Initiation insomnia, sleep fragmentation, and nightmares, common in PTSD, are all potential barriers to CPAP adherence. In our cohort, we found that patients with PTSD who were chronically using sedating medications had greater use of CPAP than patients not using these agents. This may reflect improvements in sleep initiation and reductions in insomnia, which has been shown to enhance therapeutic adher-ence with CPAP among non-PTSD patients.29 While CPAP was used more often and for longer periods among PTSD patients using sedating medications, adherence was still lower than that observed in the control group, suggesting that the diminished use of CPAP among patients with PTSD is likely multifactorial.

Poor therapeutic adherence with other disease processes has been observed in prior reports of PTSD patients. Several stud-ies have shown that medication adherence to HAART therapy is decreased in HIV patients with comorbid PTSD and depres-sion.30,31 A recent study by Lockwood et al. demonstrated that adherence with antidepressant medication was poor among PTSD patients discharged from a residential treatment program during long term follow-up.32 In addition, prior reports have found an increased likelihood of missed appointments, under-use of medications, abuse of prescription psychoactive agents, and the propensity of self-adjust medical treatments among pa-tients with PTSD.33-37

OSA is more common among patients with underlying psy-chiatric disorders. In a sample of 118,105 military veterans, the prevalence of psychiatric disorders was significantly greater in patients with OSA than those without.12 Similarly, OSA is sig-nificantly more common among patients with PTSD and has been identified in 11.9% to 90%.3,8,9,12,14,38,39 In a sample of 78 individuals seeking treatment for posttraumatic sleep distur-bances after being evacuated from a fire, 95% of those tested (50% of the subjects) experienced diminished airflow during sleep suggestive of sleep disordered breathing (SDB).8 Among 44 consecutive crime victims with PTSD reporting nightmares and insomnia, 91% had SDB.9

Untreated OSA appears to be associated with worse out-comes among patients with PTSD.14,15 Likewise, CPAP therapy has been shown to improve symptoms of depression among

Table 2—CPAP adherenceControlN = 45

PTSDN = 45 p

Percentage of nights used 76.8 ± 16.4 61.4 ± 22.2 0.001Hours/night, all nights 4.2 ± 2.1 2.5 ± 1.8 < 0.001Hours/night, nights used 4.7 ± 2.2 3.4 ± 1.2 < 0.001Regular use of CPAP (%) 58.3 25.2 0.01

Table 1—Baseline characteristicsControlN = 45

PTSDN = 45 p

Age (years) 38.6 ± 9.2 37.0 ± 11.2 0.28Male (%) 84.4 84.9 0.80BMI (kg/m2) 27.8 ± 4.4 26.9 ± 9.2 0.66Epworth Sleepiness Scale 14.7 ± 4.8 12.5 ± 6.0 0.09Fatigue 7.4 ± 0.5 5.3 ± 1.0 < 0.001AHI (events/h) 31.6 ± 24.4 28.2 ± 19.3 0.18SpO2 nadir (%) 83.3 ± 5.2 84.2 ± 4.8 0.20Insomnia (%) 11.1 25.8 0.10Chronic use of sedating agents (%) 13.3 82.9 < 0.001

BMI, body mass index; AHI, apnea hypopnea index.


JF Collen, CJ Lettieri and M Hoffmanpatients with concomitant PTSD.8,9,11-14 In a case report by Youakim et al., CPAP therapy resulted in dramatic improve-ments in sleep apnea control, daytime sleepiness, and night-mares in one veteran with PTSD.13 In a retrospective review of 15 patients with PTSD and SDB, patients who were adherent with CPAP reported a 75% improvement in PTSD symptoms compared with a 43% worsening in PTSD symptoms among patients who were non-adherent.14 Although the literature ad-dressing the impact of CPAP adherence on PTSD is limited, it suggests improved outcomes.

Sleep disturbances are common among soldiers returning from combat deployments. A recent multi-service survey of US military members found that approximately 30% reported difficulty sleeping, and that military members who had expe-rienced combat were 52% to 74% more likely to report dif-ficulty sleeping than noncombatants.40 Among 200 veterans of OIF/OEF evaluated in a Veterans Administration Polytrauma Outpatient Clinic (pain, TBI, PTSD), 93.5% reported difficul-ty sleeping.41 In addition, the overall prevalence of sleep dis-orders among service members has also markedly increased over the past decade. For example, 19,631 service members were diagnosed with insomnia in 2009, compared with 1,013 in 2000. Similarly, the diagnosis of OSA increased from 3,563 to 20,435 among service members over this same time pe-riod.42 The incidence of PTSD among combat veterans is also significantly higher than the general population, and the prevalence among US Service members has increased dramat-ically since the start of combat operations in Iraq and Afghani-stan.43 Nearly 90,000 Service members have been diagnosed with PTSD over the past decade, with the majority occurring in the Army (67%), predominantly among deployed military members.44 Given the increasing prevalence of both PTSD and sleep disturbances among US military Service members, understanding the interplay between these disorders is needed to improve outcomes.42,44,45

Our findings are similar to a recent study by El-Solh et al., who assessed CPAP use in a cohort of older veterans with PTSD and multiple comorbidities.18 Despite differences in study populations, they also found that CPAP use was signifi-cantly decreased in patients with PTSD and OSA compared to a control group without PTSD. This suggests that PTSD, and not the underlying comorbid conditions, can have a profound impact of adherence.

Our study has several limitations. As a retrospective case-control study, we were unable to assess the presence of in-somnia or PTSD using standard psychiatric interview systems (Clinician-administered PTSD scale, Duke sleep inventory, and Structured Clinical Interview for DSM-IV among oth-ers).46-49 Similarly, we were unable to calculate an insomnia severity score. We were only able to correlate CPAP use with the presence of insomnia and not the severity of insomnia, and as such cannot determine if a linear or dose-dependent cor-relation exists. While insomnia was more common among pa-tients with PTSD and likely affected outcomes in this group, this was difficult to assess without being able to correlate in-somnia severity with CPAP adherence. We did not examine other medical or behavioral treatments of PTSD that may have altered the use of CPAP and influenced our results. However, the purpose of this study was to assess the impact that PTSD

had on CPAP use, and clinical applicability would depend on both the effects of the disease process and its treatment. Fur-thermore, our comparison group provided an effective means to assess the impact of PTSD and its treatment on CPAP use. Another limitation is that our population was comprised of predominantly younger male military personnel with combat-related PTSD. As such, our results may not be generalizable to other populations. However, PTSD has been shown to simi-larly reduce adherence with other medical therapies, and in-sufficiently treated OSA has been shown to worsen outcomes in patients with psychiatric disorders. As such, we believe our findings are clinically relevant to all PTSD populations. The majority of our patients were habitually using sedating medi-cations; prior studies have shown that non-benzodiazepine sedative-hypnotics may improve CPAP adherence.29,50,51 This likely diminished the observed differences in CPAP use, as we found that patients using sedating agents had increased use of CPAP than patients not using sedatives. Despite this, PTSD still impaired CPAP adherence regardless of concomitant sed-ative use. Finally, while we only measured CPAP use during the first month of therapy and did not assess the long-term im-pact of PTSD on adherence, it has been previously established that long-term CPAP adherence is predicted by use during the first weeks of therapy.52-57

Similar to the treatment of other medical conditions, we found that patients with PTSD had significantly less use of CPAP than matched controls. This may reflect the impact of comorbid insomnia, habitual use of psychoactive medi-cations, and nightmares, which are common among patients with PTSD.9,10,15,39,58 Insomnia in particular has been linked to sleep disordered breathing, with worsened sleep quality, anxiety, depression, and CPAP adherence.27,28,59-61 Given the potential for poorer clinical outcomes in this already com-promised population, identification and early interventions targeting common barriers to CPAP use should be prioritized. A multi-modal approach should be used to improve CPAP adherence.62 Strategies may include telephonic follow-up, group education, early interventions (non-benzodiazepine sedative hypnotic to improve CPAP titration and early ad-herence,29,51 close follow-up to improve issues such as poor mask fit and difficulties with equipment), and goal-setting, to improve a patient’s sense of self-efficacy in their ability to be adherent with therapy.63 Furthermore, there should be a sense of urgency among providers caring for patients with PTSD and sleep disorders, especially combat veterans, as poor sleep may increase the risk for suicidality in this population.64-67 PTSD should be considered a significant barrier to CPAP ac-ceptance and adherence and should prompt a greater focus to enhance CPAP use.

ABBREVIATIONS

AHI, apnea-hypopnea indexCPAP, continuous positive airway pressureESS, Epworth Sleepiness ScaleOSA, obstructive sleep apneaPTSD, posttraumatic stress disorderOIF, Operation Iraqi FreedomOEF, Operation Enduring Freedom


Impact of PTSD on CPAP Adherence in OSA

REFERENCES1. American Psychiatric Association. Diagnostic and statistical manual of mental

disorders (4th ed.). Washington, DC: American Psychiatric Association, 1994.2. Krakow B, Artar A, Warner TD, et al. Sleep disorder, depression, and suicidality

in female sexual assault survivors. Crisis 2000;21:163-70.3. Krakow B, Melendrez D, Johnston L, et al. Sleep-disordered breathing, psychi-

atric distress, and quality of life impairment in sexual assault survivors. J Nerv Ment Dis 2002;190:442-52.

4. Clum GA, Nishith P, Resick PA. Trauma-related sleep disturbance and self-reported physical health symptoms in treatment-seeking female rape victims. J Nerv Ment Dis 2001;189:618-22.

5. Saladin ME, Brady KT, Dansky BS, et al. Understanding comorbidity between PTSD and substance use disorders: two preliminary investigations. Addict Be-hav 1995; 20:643-55.

6. Germain A, Buysse D, Nofzinger E. Sleep-specific mechanisms underlying post-traumatic stress disorder: Integrative review and neurobiological hypotheses. Sleep Med Rev 2008;12:185-95.

7. Nishith P, Resick PA, Mueser KT. Sleep difficulties and alcohol use motives in female rape victims with posttraumatic stress disorder. J Trauma Stress 2001;14:469-79.

8. Krakow B, Haynes PL, Warner TD, et al. Nightmares, insomnia, and sleep-disordered breathing in fire evacuees seeking treatment for posttraumatic sleep disturbance. J Trauma Stress 2004;17:257-68.

9. Krakow B, Melendrez D, Pedersen B, et al. Complex insomnia: insomnia and sleep-disordered breathing in a consecutive series of crime victims with night-mares and PTSD. Biol Psychiatry 2001;49:948-53.

10. Krakow B, Melendrez D, Warner TD, et al. Signs and symptoms of sleep-disor-dered breathing in trauma survivors: a matched comparison with classic sleep apnea patients. J Nerv Ment Dis 2006;194:433-9.

11. Sateia MJ. Update on sleep and psychiatric disorders. Chest 2009;135:1370-9.12. Sharafkhaneh A, Giray N, Richardson P, et al. Association of psychiatric disor-

ders and sleep apnea in a large cohort. Sleep 2005;28:1405-11.13. Youakim JM, Doghramji K, Schutte SL. Posttraumatic stress disorder and ob-

structive sleep apnea syndrome. Psychosomatics 1998;39:168-71.14. Krakow B, Lowry C, Germain A, et al. A retrospective study on improvements in

nightmares and post-traumatic stress disorder following treatment for comorbid sleep-disordered breathing. J Psychosom Res 2000;49:291-8.

15. Lamarche LJ, De Koninck J. Sleep disturbance in adults with posttraumatic stress disorder: a review. J Clin Psychiatry 2007;68:1257-70.

16. Means MK, Ulmer CS, Edinger JD. Ethnic differences in continuous positive airway pressure (CPAP) adherence in veterans with and without psychiatric dis-orders. Behav Sleep Med 2010;8:260-73.

17. Kjelsberg FN, Ruud EA, Stavem K. Predictors of symptoms of anxiety and de-pression in obstructive sleep apnea. Sleep Med 2005;6:341-6.

18. El-Solh AA, Ayyar L, Akinnusi M, et al. Positive airway pressure adherence in veterans with posttraumatic stress disorder. Sleep 2010;33:1495-500.

19. American Academy of Sleep Medicine: International classification of sleep dis-orders: diagnostic and coding manual. 2nd ed. Westchester, IL, 2005: 51-56.

20. Kushida CA, Littner MR, Morgenthaler T, et al. Practice parameters for the indi-cations for polysomnography and related procedures: an update for 2005. Sleep 2005;28:499-521.

21. Weathers FW, Litz BT, Herman DS, Huska JA, Keane TM. The PTSD Checklist (PCL): reliability, validity, and diagnostic utility. Paper presented at the Annual Meeting of International Society for Traumatic Stress Studies, San Antonio, TX, October 1993.

22. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: American Psychiatric Associa-tion, 2000.

23. Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540-5.

24. Kribbs NB, Pack AI, Kline LR, et al. Objective measurement of patterns of na-sal CPAP use by patients with obstructive sleep apnea. Am Rev Respir Dis 1993;147:887-95.

25. Engleman HM, Asgari-Jirhandeh N, McLeod AL, et al. Self-reported use of CPAP and benefits of CPAP therapy: a patient survey. Chest 1996;109:1470-6.

26. Haynes PL. The role of behavioral sleep medicine in the assessment and treat-ment of sleep disordered breathing. Clin Psychol Rev 2005;25:673-705.

27. Lavie P. Insomnia and sleep-disordered breathing. Sleep Med 2007; 8 Suppl 4:S21-25.

28. Wickwire EM, Smith MT, Birnbaum S, et al. Sleep maintenance insomnia complaints predict poor CPAP adherence: A clinical case series. Sleep Med 2010;11:772-6.

29. Lettieri CJ, Shah AA, Holley AB, et al. Effects of a short course of eszopiclone on continuous positive airway pressure adherence: a randomized trial. Ann Intern Med 2009;151:696-702.

30. Keuroghlian AS, Kamen CS, Neri E, et al. Trauma, dissociation, and antiretroviral adherence among persons living with HIV/AIDS. J Psychiatr Res 2011;45:942-8.

31. Safren SA, Gershuny BS, Hendriksen E. Symptoms of posttraumatic stress and death anxiety in persons with HIV and medication adherence difficulties. AIDS Patient Care STDS 2003;17:657-64.

32. Lockwood A, Steinke DT, Botts SR. Medication adherence and its effect on re-lapse among patients discharged from a Veterans Affairs posttraumatic stress disorder treatment program. Ann Pharmacother 2009;43:1227-32.

33. Chilcoat HD, Breslau N. Posttraumatic stress disorder and drug disorders: test-ing causal pathways. Arch Gen Psychiatry 1998;55:913-7.

34. Shemesh E, Lurie S, Stuber ML, et al. A pilot study of posttraumatic stress and nonadherence in pediatric liver transplant recipients. Pediatrics 2000;105:E29.

35. Shemesh E, Rudnick A, Kaluski E, et al. A prospective study of posttraumatic stress symptoms and nonadherence in survivors of a myocardial infarction (MI). Gen Hosp Psychiatry 2001;23:215-22.

36. Sparr LF, Moffitt MC, Ward MF. Missed psychiatric appointments: who returns and who stays away. Am J Psychiatry 1993; 150:801-805.

37. Spoont M, Sayer N, Nelson DB. PTSD and treatment adherence: the role of health beliefs. J Nerv Ment Dis 2005;193:515-22.

38. Krakow B, Melendrez D, Ferreira E, et al. Prevalence of insomnia symptoms in patients with sleep-disordered breathing. Chest 2001;120:1923-9.

39. Spoormaker VI, Montgomery P. Disturbed sleep in post-traumatic stress disor-der: secondary symptom or core feature? Sleep Med Rev 2008;12:169-84.

40. Seelig AD, Jacobson IG, Smith B, et al. Sleep patterns before, during, and after deployment to Iraq and Afghanistan. Sleep 2010;33:1615-22.

41. Lew HL, Pogoda TK, Hsu PT, et al. Impact of the “polytrauma clinical triad” on sleep disturbance in a department of veterans affairs outpatient rehabilitation setting. Am J Phys Med Rehabil 2010;89:437-45.

42. MSMR: A Publication of the Armed Services Health Surveillance Center. In: Brundage J, ed. MSMR 2010:1-28.

43. Jelinek P. Wartime PTSD cases jumped roughly 50 pct. in 2007. http://www.huffingtonpost.com/2008/05/27/wartime-ptsd-cases-jumped_n_103779.html. Huffington Post, 2008.

44. Fischer H. U.S. Military Statistics: Operation New Dawn, Operation Iraqi Free-dom, and Operation Enduring Freedom. Congressional Research Service: CRS Report for Congress, 2010.

45. Kennedy K. Sleep apnea cases soar in past decade. Army Times 2010.46. Blake DD, Weathers FW, Nagy LM, et al. The development of a clinician-admin-

istered PTSD scale. J Traumatic Stress 1995;8:75-90.47. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured clinical interview for

DSM-IV-TR Axis I Disorders, research version, patient edition (SCID-I/P). New York: Biometrics Research, New York State Psychiatric Institute, Nov 2002.

48. Edinger JD, Fins AI, Glenn DM, et al. Insomnia and the eye of the beholder: are there clinical markers of objective sleep disturbances among adults with and without insomnia complaints? J Consult Clin Psychol 2000;68:586-93.

49. Schramm E, Hohagen F, Grasshoff U, et al. Test-retest reliability and validity of the Structured Interview for Sleep Disorders According to DSM-III-R. Am J Psychiatry 1993;150:867-72.

50. Collen J, Lettieri C, Kelly W, et al. Clinical and polysomnographic predic-tors of short-term continuous positive airway pressure compliance. Chest 2009;135:704-9.

51. Lettieri CJ, Collen JF, Eliasson AH, et al. Sedative use during continuous posi-tive airway pressure titration improves subsequent compliance: a randomized, double-blind, placebo-controlled trial. Chest 2009;136:1263-8.

52. Stepnowsky CJ Jr., Marler MR, Ancoli-Israel S. Determinants of nasal CPAP compliance. Sleep Med 2002;3:239-47.

53. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc 2008;5:173-8.

54. Aloia MS, Arnedt JT, Stanchina M, et al. How early in treatment is PAP adher-ence established? Revisiting night-to-night variability. Behav Sleep Med 2007; 5:229-240.

55. Budhiraja R, Parthasarathy S, Drake CL, et al. Early CPAP use identifies subse-quent adherence to CPAP therapy. Sleep 2007;30:320-4.

56. Lewis KE, Seale L, Bartle IE, et al. Early predictors of CPAP use for the treat-ment of obstructive sleep apnea. Sleep 2004;27:134-8.

57. Pepin JL, Krieger J, Rodenstein D, et al. Effective compliance during the first 3 months of continuous positive airway pressure. A European prospective study of 121 patients. Am J Respir Crit Care Med 1999;160:1124-9.

58. Neylan TC, Marmar CR, Metzler TJ, et al. Sleep disturbances in the Vietnam generation: findings from a nationally representative sample of male Vietnam veterans. Am J Psychiatry 1998;155:929-33.


JF Collen, CJ Lettieri and M Hoffman59. Beneto A, Gomez-Siurana E, Rubio-Sanchez P. Comorbidity between sleep ap-

nea and insomnia. Sleep Med Rev 2009;13:287-93.60. Wickwire EM, Collop NA. Insomnia and sleep-related breathing disorders. Chest

2010;137:1449-1463.61. Pieh C, Bach M, Popp R, et al. Insomnia symptoms influence CPAP compliance.

Sleep Breath 2012.62. Sawyer AM, Gooneratne NS, Marcus CL, et al. A systematic review of CPAP

adherence across age groups: clinical and empiric insights for developing CPAP adherence interventions. Sleep Med Rev 2011;15:343-56.

63. Aloia MS. Understanding the problem of poor CPAP adherence. Sleep Med Rev 2011;15:341-2.

64. Agargun MY, Kara H, Solmaz M. Sleep disturbances and suicidal behavior in patients with major depression. J Clin Psychiatry 1997;58:249-51.

65. Agargun MY, Kara H, Solmaz M. Subjective sleep quality and suicidality in pa-tients with major depression. J Psychiatr Res 1997;31:377-81.

66. Bernert RA, Joiner TE Jr., Cukrowicz KC, et al. Suicidality and sleep disturbanc-es. Sleep 2005;28:1135-41.

67. Ribeiro JD, Pease JL, Gutierrez PM, et al. Sleep problems outperform de-pression and hopelessness as cross-sectional and longitudinal predictors of suicidal ideation and behavior in young adults in the military. J Affect Disord 2012;136:743-50.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication January, 2012Submitted in final revised form April, 2012Accepted for publication May, 2012Address correspondence to: Jacob Collen, M.D., Pulmonary, Critical Care and Sleep Medicine, Walter Reed National Military Medical Center, 8901 Rockville Pike, Bethesda, MD 20889; Tel: (301) 295-4191; E-mail: [email protected]

DISCLOSURE STATEMENTThe views expressed in this paper are those of the authors and do not reflect

the official policy of the Department of the Army, Department of Defense, or the US Government. This was not an industry supported study. The authors have indicated no financial conflicts of interest.


in children with Arnold-Chiari malformations, even in the ab-sence of vocal cord paralysis or dysfunction.6,7 Woodson dem-onstrated increased retropalatal collapse during the expiratory phase in adults with mild obstructive sleep apnea and snoring.8

Further understanding and defi nition of these expiratory related events are required, both in children with evidence of sleep dis-ordered breathing and in normal children.

Here we describe the respiratory, cardiac, and sleep-state characteristics of two types of prolonged sleep-related respirato-ry pauses during exhalation: post-sigh central apnea (PSCA) and prolonged expiratory apnea (PEA). Both events are suggestive of airfl ow limitation during exhalation. An augmented breath, which is a sigh, precedes both PEA and PSCA. Both events ful-fi ll criteria for apnea according to standard classifi cations. These events may be inappropriately categorized as pathologic events when in fact they appear, in this study, to be benign. We hypoth-

Study Objectives: We describe the respiratory, cardiac, and sleep-related characteristics of two types of sleep-related re-spiratory pauses in children that can fulfi ll current criteria of pathological apnea, but often seem to be benign: prolonged expiratory apnea (PEA) and post-sigh central apnea (PSCA).Methods: All outpatient comprehensive overnight polysom-nography completed on children without signifi cant underly-ing medical conditions completed during an 18-month period were retrospectively reviewed for the presence of augmented breaths followed by a respiratory pause. Events were identifi ed as a PEA or PSCA based on characteristic features. Physi-ologic parameters associated with the respiratory events were recorded and compared.Results: Fifty-seven (29 PEA and 28 PEA) events were identi-fi ed in 17 patients (8.5 ± 3.5 years old). Median durations of PEA and PSCA were not signifi cantly different. For both PEA and PSCA, average heart rate (HR) during the augmented

breath before the respiratory pause differed from lowest instan-taneous HR during the fi rst half of the pause. When compared to each other, the lowest instantaneous HR recorded in the fi rst half of PEA was lower than that for PSCA (63.9 [59.41–68.3] vs 66.75 [61.7–80.75]) beats per min, p = 0.03. No PEA or PSCA event was associated with an oxygen desaturation more than 3% from baseline.Conclusion: PEA and PSCA have stereotypic HR changes and resemble pathologic apneas but appear to be benign. Clinical signifi cance of PEA and PSCA is yet to be determined. Consistent recognition of the events is required, given their fre-quency of occurrence and potential for misclassifi cation.Keywords: Sleep disorders, sleep apnea syndromes, pediat-rics, apnea, polysomnography, exhalationCitation: Haupt ME; Goodman DM; Sheldon SH. Sleep re-lated expiratory obstructive apnea in children. J Clin Sleep Med 2012;8(6):673-679.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Prolonged apnea during infancy and childhood may be life-threatening. Normal full-term infants and children, how-

ever, exhibit some respiratory pauses during sleep that are not pathological.1 Differentiation of normal physiological respira-tory pauses and pathological apnea can be diffi cult.

Apnea is a transient cessation in respiration and can be clas-sifi ed as obstructive, central, or mixed. Obstructive apnea in children is conventionally defi ned as a > 90% decrease in air-fl ow that lasts ≥ 90% of the duration of two normal breaths, as determined from the baseline breathing pattern, and has con-tinued or increased respiratory effort during the period of de-creased airfl ow.2 Central apnea occurs when inspiratory effort is absent during the entire period of airfl ow cessation. Current scoring rules require the event to last 20 seconds or to have at least two missed breaths and be associated with an electrocorti-cal arousal, an awakening, or ≥ 3% desaturation. Mixed apnea occurs if there is absent inspiratory effort during the initial por-tion of the respiratory pause, followed by resumption of inspi-ratory effort before the end of the period of decreased airfl ow. Clearly, the focus of the defi nition of these events is related to the inspiratory phase oscillation of respiration.

Data do exist suggesting that signifi cant respiratory events may occur only during the expiratory phase of the respiratory cycle in sleep. Southhall fi rst presented the notion of prolonged expiratory apnea in children in 1985, demonstrating a relation-ship between upper airway obstruction and severe hypoxemia.3

Recent models of airfl ow limitation during the expiratory phase have been developed.4,5 Expiratory apneas are also described

Sleep Related Expiratory Obstructive Apnea in ChildrenMark E. Haupt, M.D.1; Denise M. Goodman, M.D., M.S.2; Stephen H. Sheldon, D.O., F.A.A.S.M.3

1Division of Pulmonary Medicine, 2Division of Critical Care Medicine, 3Division of Sleep Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Northwestern University Feinberg School of Medicine, Chicago, IL

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Respiration during sleep in children is complex and differentiation of normal physiologic respiratory pauses and pathologic apneas can be diffi cult. The purpose of this study is to describe the respiratory, cardiac and sleep related characteristics of two types of sleep related respiratory pauses in children that can fulfi ll criteria for pathological apnea that, in this study, appear to be benign.Study Impact: We provide detailed description of two types of expiratory pauses in children: Prolonged Expiratory Apnea and Post-Sign Central Apnea. Consistent recognition and accurate scoring of such events is necessary given their frequency and potential for misclassifi cation.


ME Haupt, DM Goodman and SH Sheldonesize that there are changes in autonomic and mechanically me-diated physiologic parameters that differentiate the two events. Both events demonstrate a significant decrease in heart rate. The decrease in heart rate is exaggerated in PEA in response to oc-clusion of the upper airway during exhalation.

METHODS

This retrospective study evaluated the characteristics of two types of prolonged sleep related respiratory pauses that occur in children. We reviewed all archived outpatient polysomnograms completed on patients aged 5 to 14 years old for routine clini-cal care between January 2010 and April 2011. All children un-derwent a comprehensive history and physical examination by a board certified pediatric sleep medicine physician or under-went a comprehensive review of the history and physical exam according to accepted protocol prior to polysomnography. A standardized validated comprehensive sleep history question-naire (Children’s Memorial Pediatric Sleep Questionnaire) was also reviewed by the pediatric sleep medicine specialist. All data were obtained before the polysomnogram. Patients were excluded from analysis if their polysomnogram demonstrated severe obstructive sleep disordered breathing, had develop-mental delay with static encephalopathy, required supplemental oxygen at baseline, or had significant upper airway abnormali-ties (Pierre-Robin Sequence, history of tracheostomy, history of laryngotracheal reconstruction, etc.), or if they were unable to complete the study. Studies were not included if there was insufficient data recording for analysis. Studies that were com-pleted after adenotonsillectomy were not included. Studies that met the above inclusion criteria were then screened for the pres-ence of augmented breaths associated with respiratory pauses, described in detail below. Seventeen patients, 5 to 14 years old, had events and complete recording data available for data anal-ysis. Fifty-seven events were included for analysis. This study was approved by the Institutional Review Board of Children’s Memorial Hospital.

Comprehensive video-polysomnography was performed on all patients. Standard clinical consent was obtained prior to each polysomnogram. All polysomnograms were obtained on a 32-channel Cadwell Easy III PSG machine (Cadwell Labo-ratories Inc, Kennewick, WA, USA). Identical techniques and recording montage were used for all patients and consisted of: electroencephalogram ([EEG]: 8 channels), electro-oculogra-phy ([EOG]: consisting of left outer canthus and right outer canthus), chin muscle electromyography (EMG), modified lead II electrocardiogram (ECG), and left and right anterior tibialis EMG. Respiratory effort was monitored with inductive plethys-mography with separate chest and abdominal belts. Intercostal EMG recording was obtained by placement of an electrode in the 5th intercostal space in the anterior axillary line. Airflow was monitored by nasal pressure transduction, nasal thermistry, and continuous monitoring of the capnography wave form (at the nose and mouth) with breath-to-breath analysis and 10-sec av-erage (Smiths BCI Capnocheck Plus, Smiths Medical, St. Paul, MN, USA). A 3-sec sampling delay occurs with use of a split-lumen cannula. Such a delay was incorporated into scoring of events. Continuous pulse oximetry was recorded for the dura-tion of each study (Cadwell Laboratories Inc, Kennewick, WA,

USA, Nonin 6000ci Sensor, Nonin, Plymouth, MN, USA). All studies included digital video recording. Prior to each study, calibration and graphic representation of inspiration and expi-ration were validated using polysomnographic methods modi-fied to verify polarity of graphic representation. Inspiration was identified polysomnographically by documenting negative voltage change (producing an upward movement) from an iso-electric baseline for airflow, chest effort, and abdominal effort during prolonged voluntary inspiration. Expiration was identi-fied by documenting positive voltage change (producing down-ward movement) from an isoelectric baseline for airflow, chest effort, and abdominal effort during prolonged voluntary expira-tion. All records were scored for sleep staging by an accredited clinical polysomnographer according to the accepted criteria of the AASM 2007 Scoring Manual.2

All polysomnograms that met initial inclusion criteria were screened for the presence of augmented breaths. Augmented breaths were defined as an inspiratory/expiratory cycle, (as measured by polarity and voltage change on both chest and ab-dominal channels) at least 50% greater than the average cyclic voltage variation for the preceding 4 breaths. Only augmented breaths followed by a respiratory pause ≥ 6 sec were included for analysis. Augmented breaths were accompanied by brief (< 3 sec) electrocortical EEG arousals. Those augmented breaths associated with major body movements were excluded from analysis. The respiratory events were not included if they had previously been scored as a central, obstructive, or mixed ap-nea, even if they met the criteria described above. Only events in NREM sleep were included.

All augmented breaths that occurred during sleep were tagged, printed without patient identifiers, and provided to one of the authors (SS) who was blinded on the type of event for classification as a PEA or PSCA. Capnography was used to differentiate the two events. A dual-lumen capnography can-nula was used to record changes in CO2 at the nose and mouth. Use of such a device permits accurate capnography record-ing irrespective of the route of inhalation and exhalation. PEA demonstrates a significantly prolonged alveolar plateau in the capnography wave form after the augmented breath. PSCA may demonstrate a prolonged alveolar plateau, but is proportional to the change in tidal volume with the augmented breath, although not as prolonged as seen in PEA. Additionally, classification was based on the polarity of voltage change on the airflow channel for the first breath following the respiratory pause. Events were classified according to whether the first post-apnea breath was inspiratory or expiratory. When the first breath after the augmented breath and respiratory pause was inspiratory, the event was classified as a PSCA. When the first breath after the augmented breath and associated respiratory pause was expira-tory, the event was classified as a PEA. There are some cases in which each event did not hold true to this rule, demonstrating the importance of the role of capnography in identifying these events. Representative images are included and discussed in the results section. Scoring of the events was done in an encrypted manner so as not to bias data analysis.

Cardiac and respiratory variables were calculated and coded by a research fellow, who was blinded to the categorization of each event. Average heart rate and respiratory rate were deter-mined from the average heart rate and respiratory rate for 10 sec


Expiratory Apnea in Childrenbefore the apnea and 10 sec after the tagged events. R-R inter-vals on ECG were measured from the onset of the augmented breath (beginning at zero baseline crossing) and throughout the apnea with the use of integrated tools within the polysomno-gram software. Apnea length and oxygen saturation recorded during and after the events were noted.

All data were entered into a commercially available data base computer program and analyzed using the R system for statistical analysis version 2.12.2 (R Development Core Team). Descriptive analysis was completed on the patient character-istics, results of the polysomnograms, and for the cardiopul-monary variables. Paired comparisons between pre-event vital signs and those during the event were compared using the sign test. Comparisons between the two types of events were per-formed using the Mann-Whitney U test. The study was pow-ered to detect a 15% difference between groups with a power of 90% and a type I error of 0.05. A 15% difference in heart rate was determined to be clinically significant based on recently published normative data.9 Results are reported as medians and interquartile ranges unless otherwise specified.

RESULTS

Seventeen patients, 9 girls and 8 boys, were included in the analysis, with an average age of 8.5 ± 3.5 years. Indications for referral for polysomnography are listed in table 1. Sleep architecture and continuity variables are presented in table 2. Fifty-seven augmented breaths followed by apneas ≥ 6 sec were identified. Comparative data for each event, PEA (n = 28) and PSCA (n = 29), may be found in table 3. Representative ex-amples of a PEA and PSCA are demonstrated in Figures 1–4. During the augmented breath, all 57 events were associated

with an electrocortical arousal of 3–4 seconds. The duration of the PEA and PSCA were not significantly different, lasting 10.3 (9.14–12.59) seconds and 9.55 (8.4–11.8) seconds, respectively (p = 0.26).

Both PEA and PSCA demonstrated a significant difference between the average heart rate during the augmented breath prior to the pause and the lowest instantaneous heart rate dur-ing the first half of the pause. In the PEA group, the heart rate dropped from 81.8 (76.4–96.4) beats per minute to 63.9 (59.41–68.3) beats per minute (p < 0.001). In the PSCA group, the heart rate dropped from 86.85 (73.6–101.45) beats per min-ute to 66.75 (61.7–80.75) beats per minute (p < 0.001). There was no difference between the lowest heart rate in the first half and second half of the event within each group. A significant difference was noted between the lowest instantaneous heart rate in the second half of each event to the average heart rate after each event (p < 0.001). In summary, the heart rate dem-onstrates a significant decrease from baseline during the initial portion of the respiratory event that is maintained until the end of the apnea, when the heart rate returns to baseline.

When the events were compared to each other, there was statistically significant evidence to discriminate between the

Table 3—Summary data for PEA and PSCA (median and interquartile range)PEA PSCA p-Value Mann Whitney U+

Event duration (s) 10.3 (9.14–12.59) 9.55 (8.4–11.8) 0.26HR prior (bpm) 81.8 (76.4–96.4) 86.85 (73.6–101.45) 0.678Lowest 1st half HR (bpm) 63.9 (59.41–68.3)* 66.75 (61.7–80.75)* 0.03Lowest 2nd half HR (bpm) 66.08 (60.0–68.5) 66.08 (58.68–78.63) 0.38HR after (bpm) 82.4 (76.2–89.9)‡ 87.2 (73.05–97.47)‡ 0.79

*(p < 0.001) for HR comparison prior to and during first half for PEA and PSCA, respectively. ‡(p < 0.001) for HR comparison during second half of the event and after the respiratory event for PEA and PSCA, respectively. +Comparison between PEA and PSCA.

Table 1—Summary of referral characteristicsReferral Characteristic Number of Patients

Snoring 14Restless sleep 8Tonsillar hypertrophy 6Behavioral disturbance 4Sleepiness 2Apnea 1Mouth breathing 2Limb movements in sleep 2Other (allergic rhinitis, asthma, teeth grinding, sleep talking, enuresis, obesity)

5

Table 2—Summary of polysomnography resultsMedian Interquartile Range

Total sleep time(min) 367.5 329.5–403Sleep latency (min) 26 17.5–43.5REM latency (min) 147.3 114.2–171.4Sleep efficiency (%) 77 72–89Stage N1% 3 2–3Stage N2% 48 44–50Stage N3% 33 29–36Stage R% 15 11–20Arousal index 16 14–21AHI 2 1–3RDI 2 1–6REM index 2 0–6Hypopneas 5 2–10Obstructive apnea 0 0–1Mixed apneas 0 –Average SpO2 (%) 97 96–98Low SpO2 (%) 92 90–94Average EtCO2 42.4 40.6–44Max EtCO2 50 49–51% Time EtCO2 > 50 0 0–0.2


ME Haupt, DM Goodman and SH Sheldon

two episodes. Specifically, the lowest instantaneous heart rate recorded in the first half of each event was statistical-ly significantly lower in the PEA group when compared to the PSCA group (63.9 [59.41–68.3] vs 66.75 [61.7–80.75] beats per minutes, p = 0.03). However, the percentage de-crease in heart rate is statistically different. The heart rate in

PEA decreases a median 23.8% (IQR: 17.2% to 32.2%). The heart rate in PSCA decreases a median 18.1% (IQR: 14.1% to 22.8%) (Mann-Whitney U Test p = 0.03). No significant difference was seen comparing the lowest instantaneous heart rate in the second half of the PEA group and PSCA group (p = 0.38). There was no statistical significance between the

A B C

Alveolar Plateau

D

End Tidal CO2

Electrocortical Arousal

Capnography Wave Form

Figure 2—Prolonged expiratory apnea, 60-sec epoch

The image on the left includes the capnography tracing, while the one on the right does not. Augmented breath (A); respiratory pause (B); first post-pause breath (C) in the direction of exhalation (positive polarity); lowest instantaneous heart rate during the respiratory pause (D). Note the prolonged alveolar plateau phase in the capnopgraphy tracing. Note there is about a 3- to 5-sec sampling delay in the capnography tracing, as illustrated by the horizontal line below the tracing.

A B C

Alveolar Plateau

D

End Tidal CO2



Figure 1—Prolonged expiratory apnea, 30-sec epoch

The image on the left includes the capnography tracing, while the one on the right does not. Augmented breath (A); respiratory pause (B); first post-pause breath (C) in the direction of exhalation (positive polarity); lowest instantaneous heart rate during the respiratory pause (D). Note the prolonged alveolar plateau phase in the capnopgraphy tracing. Note there is about a 3- to 5-sec sampling delay in the capnography tracing, as illustrated by the horizontal line below the tracing.


Expiratory Apnea in Children

heart rate prior to or after each event, suggesting that the dif-ference in heart rates was not due to a baseline difference in heart rate. There was no significant difference in age between the two groups (p = 0.96).

Interestingly, during split screen video/audio recording, PEA was associated with continuous expiratory stridulous vocaliza-tions, audible expiratory braking, and/or visible forced expira-tory effort.

DISCUSSION

Although brief central apneas during active/REM sleep are common and are considered normal physiological occurrences, prolonged central apneas (especially during quiet/NREM sleep) may be classified as pathological if lasting longer than 20 sec-onds, or if associated with oxygen desaturations or bradycardia. Typically, three types of apnea are described in infants and chil-

A B C



Alveolar Plateau

D

Figure 3—Post-sigh central apnea, 30-sec epoch

The image on the left includes the capnography tracing, while the one on the right does not. Note the augmented breath (A) and respiratory pause (B). The first post-pause breath (C) is in the direction of inspiration (negative polarity); lowest instantaneous heart rate during the respiratory pause (D). Note there is about a 3- to 5-sec sampling delay in the capnography tracing, as illustrated by the horizontal line below the tracing.

A B C

Alveolar Plateau

D



Figure 4—Post-sigh central apnea, 60-sec epoch

The image on the left includes the capnography tracing, while the one on the right does not. Note the augmented breath (A) and respiratory pause (B). The first post-pause breath (C) is in the direction of inspiration (negative polarity); lowest instantaneous heart rate during the respiratory pause (D). Note there is about a 3- to 5-sec sampling delay in the capnography tracing, as illustrated by the horizontal line below the tracing.


ME Haupt, DM Goodman and SH Sheldondren: obstructive, central and mixed. These events are based on physiologic changes associated with the inspiratory portion of the respiratory cycles. However, significant respiratory distur-bances occur during exhalation as well. In 1982, Onal observed prolongation of the expiratory phase of the respiratory cycle dur-ing NREM sleep central apneas.10 Southall described prolonged forced expiration against obstruction of the upper airway result-ing in cyanotic spells in infants.3 Vocal cord adduction and glot-tic closure have been identified in children who have suffered apparent life-threatening events (ALTEs) and during the central pause associated with periodic breathing.11 Perez-Padilla and Issa describe the characteristics of sighs during sleep in adults, presenting a similar description to PSCA.12,13 Eiselt describes the role of sighs in the respiration of preterm and term infants, specifically focusing on lung volume recruitment and associ-ated heart rate changes.14 These studies demonstrate that respi-ratory events during expiration and sighs occur during sleep. There is little literature specifically describing these events and their role in the spectrum of upper airway resistance syndromes and sleep disordered breathing in children. PSCA and PEA are unique expiratory related respiratory events and no literature describes them.

We described two types of respiratory events that may ap-pear similar to pathological central apnea in infants and chil-dren when limited technology is used, but are associated with continued expiratory phase of the respiratory cycle, as dem-onstrated by the prolonged alveolar plateau noted during cap-nography. When reviewed without the capnography tracing, all PSCA and PEA included in this analysis could be scored as central apnea based on 2007 AASM criteria. The prolonged al-veolar plateau noted in the PEA, however, is suggestive of con-tinued airflow during expiration, despite what appears to be an apneic event based on lack of abdominal and chest movement. Capnography can identify variations in respiratory patterns in children, and we feel this is a key component of defining these events.15 Both PSCA and PEA follow an augmented inspiration and are of sufficient duration to meet the current definition of pathological central apnea in children (PSCA range 6.54 to 14.1 seconds, PEA 6.8 to 15.2 seconds). PEA was associated with audible expiratory braking or expiratory stridor during the im-mediate post-inspiratory phase. Statistically significant changes in heart rate were identified when PSCA events were compared with PEA events. Lowest instantaneous heart rate occurred dur-ing a PEA and reached a nadir of 36 beats per minute in one case. A nadir to this degree would typically occur within the range of bradycardia considered pathological on pneumogra-phy and/or event recordings, but did not appear to be associ-ated with any untoward events during these studies. In a central apnea, the heart rate nadir occurs in the second half of the event and results in an arousal. In PSCA and PEA, the heart rate nadir occurs in the first half of the event. PSCA and PEA are preceded by an arousal.

PSCA and PEA differ from pathological central apnea in several important ways. First, PSCA and PEA are preceded by an augmented breath and brief (< 3 second) electrocortical arousal noted on EEG. Pathological central apneas are typi-cally not preceded by an augmented breath or arousal. Instanta-neous heart rate reached its nadir immediately after augmented breaths and gradually returned to baseline during the last half

of the respiratory pause during a PEA. Similar changes in heart are seen during inspiratory Valsalva maneuver.16 The pattern of heart rate changes in PSCA and PEA is similar. However, it is clear that the heart rate decreases further in PEA. This may re-flect narrowing of the upper airway during exhalation resulting in a Valsalva-like maneuver in PEA events that is not seen in PSCA events. Pathological central (as well as obstructive) ap-nea is typically associated with a gradual deceleration in heart rate, with the nadir of instantaneous heart rate occurring during the last third of the respiratory pause, immediately prior to the arousal. The bradycardia is a result of a vagal response to de-creased cardiac preload in the setting of absent respiratory mo-tion of the chest. Finally, despite the prolonged nature of these respiratory pauses, they appeared, in our sample, to be associ-ated with minimal physiological consequences. Oxygen satura-tion rarely fell more than 3% from baseline (data not shown). Pathological apneas are often associated with gradual oxygen desaturation as the respiratory pause continues.

Interestingly, PEA events were often associated with con-tinuous expiratory vocalizations and/or visible forced expira-tory effort. Given the combination of expiratory vocalization and a prolonged expiratory pause preceded by an augmented breath, one would be correct to make a possible association be-tween the events we describe and catathrenia. Catathrenia is defined as a chronic sleep disorder characterized by expiratory groaning during sleep. According to the International Classi-fication of Sleep Disorders, 2nd edition (ICSD-2), catathrenia is identified during polysomnography as a deep inhalation fol-lowed by a prolonged exhalation during which moaning noises are produced and may last between 2 and 49 seconds, with an increased frequency during the late REM cycles.17 Recent studies have demonstrated that catathrenia, or catathrenia-like symptoms, can be successfully treated with continuous positive airway pressure (CPAP) in adult patients.18 This suggests that catathrenia results from upper airway resistance and occlusion during exhalation. The origin of the sounds during catathrenia is different from inspiratory snoring, again suggesting a differ-ent mechanism, namely upper airway resistance.19 Catathrenia vocalizations occur during REM sleep and tend to occur in clusters. All of the events we demonstrate here occurred during NREM sleep and were isolated occurrences without evidence of a temporal pattern.

Clinical significance of PSCA and PEA is unknown and may be an age-related phenomenon in the pediatric population, re-sulting in more severe consequences in infants but not in older children. However, these events merit attention, as they may re-sult in incorrect diagnoses and inappropriate evaluations based on misinterpretation or misclassification of the event.

Our study has limitations, specifically the true incidence of expiratory apnea events in children is not known. Selection bias is present in this study, given that the polysomnograms used for analysis were done for children with suspected sleep disordered breathing, not for otherwise normal children. We attempted to limit this effect by selecting studies from patients who did not demonstrate significant sleep disordered breathing or major upper airway abnormalities. Determination of age dependence and incidence of abnormal expiratory events in children analy-sis of polysomnograms in normal children for these specific events is needed.


Expiratory Apnea in ChildrenIn summary, we describe two unique expiratory related respi-

ratory events that occur during sleep in children. PEA and PSCA demonstrate stereotypic heart rate changes and may resemble pathologic apneas. Both PEA and PSCA can be identified by an augmented breath followed by a prolonged respiratory pause. The heart rate significantly decreases from baseline during both events, more so in PEA, before returning to baseline. The long-term sequelae and clinical significance of these events is yet to be identified and understood. Future prospective studies are required to determine the frequency of these events in children with and without sleep disordered breathing. Given the similar appearance of these events to pathologic apneas and the poten-tial for unnecessary evaluation, addition of PEA and PSCA to polysomnographic scoring manuals should be considered.

ABBREVIATIONS

PEA, prolonged expiratory apneaPSCA, post-sigh central apneaCPAP, continuous positive airway pressureEEG, electroencephalogramEOG, electro-oculographyEMG, electromyographyECG, electrocardiogram

REFERENCES1. Haddad GG, Farber JP. Developmental neurobiology of breathing. New York:

Dekker; 1991.2. Iber C, Ancoli-Israel S, Chesson A, Quan SF for the American Academy of Sleep

Medicine. The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications, 1st ed.: Westchester, IL: Ameri-can Academy of Sleep Medicine, 2007.

3. Southall DP, Talbert DG, Johnson P, et al. Prolonged expiratory apnoea: a dis-order resulting in episodes of severe arterial hypoxaemia in infants and young children. Lancet 1985;2:571-7.

4. Series F, Marc I. Accuracy of breath-by-breath analysis of flow-volume loop in identifying sleep-induced flow-limited breathing cycles in sleep apnoea-hypop-noea syndrome. Clin Sci (Lond) 1995;88:707-12.

5. Stanescu D, Kostianev S, Sanna A, Liistro G, Veriter C. Expiratory flow limitation during sleep in heavy snorers and obstructive sleep apnoea patients. Eur Respir J 1996;9:2116-21.

6. Cochrane DD, Adderley R, White CP, Norman M, Steinbok P. Apnea in patients with myelomeningocele. Pediatr Neurosurg 1990;16:232-9.

7. Stevenson KL. Chiari Type II malformation: past, present, and future. Neurosurg Focus 2004;16:E5.

8. Woodson BT. Expiratory pharyngeal airway obstruction during sleep: a multiple element model. Laryngoscope 2003;113:1450-9.

9. Flemming S, Thompson M, Stevens R. Normal ranges of heart rate and respira-tory rate in children from birth to 18 years of age: a systematic review of obser-vational studies. Lancet 2011;377:1011-8.

10. Onal E, Lopata M, O’Connor T. Pathogenesis of apneas in hypersomnia-sleep apnea syndrome. Am Rev Respir Dis 1982;125:167-74.

11. Ruggins NR, Milner AD. Site of upper airway obstruction in infants following an acute life-threatening event. Pediatrics 1993;91:595-601.

12. Perez-Padilla R, West P, Kryger M. Sighs during sleep in adult humans. Sleep 1983;6:234-43.

13. Issa FG, Porostocky S. Effect of sleep on changes in breathing pattern accom-panying sigh breaths. Respir Phsyiol 1993:175-87.

14. Eiselt M, Curzi-Dascalova L, Leffler C, et al. Sigh-related heart rate changes dur-ing sleep in premature and full-term newborns. Neuropediatrics 1992;23:286-91.

15. Tirosh E, Bilker A, Bader D, Cohen A. Capnography in spontaneously breathing preterm and term infants. Clin Physiol 2001;21:150-4.

16. Looga R. The Valsalva manoeuvre--cardiovascular effects and performance technique: a critical review. Respir Physiol Neurobiol 2005;147:39-49.

17. American Academy of Sleep Medicine. The international classification of sleep disorders: diagnostic and coding manual. 2nd ed. Westchester, IL: American Academy of Sleep Medicine; 2005.

18. Abbasi AA, Morgenthaler TI, Slocumb NL, Tippmann-Peikert M, Olson EJ, Ra-mar K. Nocturnal moaning and groaning-catathrenia or nocturnal vocalizations. Sleep Breath 2012;16:367-73.

19. Iriarte J, Fernandez S, Fernandez-Arrechea N, et al. Sound analysis of catathre-nia: a vocal expiratory sound. Sleep Breath 2011;15:229-35.

ACKNOWLEDGMENTSWork for this study was performed at Children’s Memorial Hospital, Chicago IL.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication November, 2011Submitted in final revised form April, 2012Accepted for publication April, 2012Address correspondence to: Mark Haupt, M.D., Ann & Robert H. Lurie Children’s Hospital of Chicago, Division of Pulmonary Medicine, Box #43, 225 East Chicago Ave., Chicago, IL 60610; Tel: (312) 227-6260; Fax: (312) 227-9419; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. The author has indicated no financial



Introduction: Though relaxation training is recommended for insomnia, national patterns of use remain unknown. Similarly, rates of complementary and alternative medicine (CAM) use by adults with insomnia are not well established. We sought to elucidate the patterns and reasons for use of relaxation tech-niques and CAM use by adults with insomnia symptoms.Methods: We used the 2007 National Health Interview Survey (n = 23,358) to estimate prevalence of use among adults by self-reported insomnia symptom status. Among adults reporting insomnia symptoms (n = 4,415), we examined reasons for use and disclosure to medical professionals. We employed logistic regression to determine the adjusted associations between re-laxation techniques use, CAM use, and insomnia symptoms.Results: Among adults with insomnia symptoms, 23% used relaxation techniques and 45% used CAM annually. After ad-justment, adults with insomnia symptoms had higher likelihood of using relaxation techniques (aOR 1.48, 95% CI 1.32, 1.66) and CAM (aOR 1.29, 95% CI 1.15, 1.44) compared with adults without insomnia. Deep breathing exercise was the most com-

monly used relaxation technique. Fewer than 2% of adults with insomnia used CAM specifi cally for insomnia. Only 26% of adults with insomnia symptoms disclosed their relaxation techniques use to medical professionals. Being male, lower educational and physical activity levels, income < $20,000, living in South, and hypertension were associated with lower likelihood of relaxation techniques use among adults with in-somnia symptoms.Conclusion: While adults with insomnia symptoms commonly use relaxation techniques and CAM, few are using for their insomnia. Facilitating discussions about relaxation techniques may foster targeted use for insomnia.Keywords: Insomnia, relaxation techniques, complementary therapies, behavioral therapies, United States, epidemiology, NHISCitation: Bertisch SM; Wells RE; Smith MT; McCarthy EP. Use of relaxation techniques and complementary and alternative medicine by American adults with insomnia symptoms: results from a national survey. J Clin Sleep Med 2012;8(6):681-691.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Approximately 10% to 15% of adults suffer from chronic insomnia.1 Both central and autonomic nervous system

hyperarousal have been strongly implicated in the pathophysi-ology of insomnia,2,3 and, as such, relaxation techniques have been explored as treatment for insomnia for several decades.4-6

Relaxation training comprises a diverse group of practices (e.g., progressive muscle relaxation, diaphragmatic breathing) to reduce “somatic tension” or “intrusive thoughts.”7 When employed during pre-sleep periods, these therapies are thought to counteract the cognitive and physiologic mechanisms that initiate and perpetuate insomnia, and are recommended as “standard” treatment for insomnia by the American Academy of Sleep Medicine (AASM).7 However, little is known about the prevalence, patterns, and reasons for use of relaxation tech-niques by adults with insomnia.

While conventional pharmacologic and behavioral treat-ments can be effective for insomnia, many patients with insom-nia report seeking alternative therapies for sleep complaints.8

The National Institutes of Health defi nes complementary and

Use of Relaxation Techniques and Complementary and Alternative Medicine by American Adults with Insomnia

Symptoms: Results from a National SurveySuzanne M. Bertisch, M.D., M.P.H.1; Rebecca Erwin Wells, M.D., M.P.H.2; Michael T. Smith, Ph.D.3; Ellen P. McCarthy, Ph.D., M.P.H.4

1Divisions of General Medicine and Primary Care, and Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Brookline, MA: 2Department of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; 3Department of Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Behavioral Sleep Medicine Program, Johns Hopkins Bayview Medical Center, Baltimore, MD; 4Division of General Medicine

and Primary Care, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Brookline, MA

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Although relaxation techniques (RT) and complementary and alternative medicine (CAM) approaches are commonly used in the United States, and practice guidelines recom-mend RT as a “standard treatment” for insomnia, neither RT nor CAM usage among adults with insomnia is well understood. We therefore in-vestigated national usage patterns of both RT and CAM among adults with insomnia symptoms.Study Impact: We found that adults with insomnia symptoms have high prevalence of use of RT and CAM, though specifi c use for insomnia was uncommon. Further investigations to determine barriers to use of relax-ation techniques for insomnia are warranted.

alternative medicine (CAM) as a group of diverse medical and health care systems, practices, and products that are not gener-ally practiced by allopathic physicians and allied health profes-sionals.9 One study estimated 1.6 million Americans used CAM to treat insomnia in 2002, with biologically based therapies (e.g., herbs) and mind-body therapies (e.g., meditation), report-


SM Bertisch, RE Wells, MT Smith et aled as the most commonly used CAM therapies for insomnia.8 Other studies have also demonstrated that CAM is used com-monly by adults with chronic medical conditions linked with insomnia.10,11 There are also emerging data to suggest that some CAM therapies, including acupressure,12 yoga,13 and Tai Chi14-16 may treat insomnia. However, the extent to which adults with insomnia symptoms use CAM is not well established.

In this context, we sought to determine the prevalence of relaxation techniques and overall CAM use, identify demo-graphic and health-related correlates of relaxation techniques and CAM use, and to quantify reasons for use as well as the ex-tent to which patients disclose their use to conventional medical professionals.

METHODS

Data SourceThe National Health Interview Survey (NHIS) is an in-per-

son household survey of the civilian, noninstitutionalized US population, conducted by the Census Bureau for the National Center for Health Statistics. Households are selected for face-to-face interviews in English and/or Spanish using a multistage area probability sampling design described elsewhere.17 Infor-mation on sociodemographic characteristics, health status, in-surance status, and access to and use of healthcare services is collected on all household members. One adult, ≥ 18 years, is then randomly selected from each family to provide additional detailed information on height, weight, common medical condi-tions, and healthcare utilization. In 2007, these sampled adults were also administered the Alternative Medicine Supplement, which queried respondents about use of relaxation techniques and more than 20 CAM therapies. Respondents were asked, “During the past 12 months, have you used (specific therapy)?” In 2007, 23,393 sampled adults completed the Adult Question-naire, representing a 67.8% response rate.18

Insomnia SymptomsInsomnia symptom status was assessed based upon partici-

pants’ response (yes/no) to the question, “During the past 12 months, have you regularly had insomnia or trouble sleeping?” We excluded 35 adults who responded “don’t know” or refused to answer. Our final sample size included 23,358 adult respondents.

Outcomes of Interest

Relaxation TechniquesWe defined relaxation techniques by use of at least one of

the following practices within the past 12 months: deep breath-ing exercises, progressive muscle relaxation, biofeedback, and guided imagery. We selected these techniques to be consistent with the recommendations from the AASM.7

Complementary and Alternative MedicineWe also examined recent use of 4 broad CAM categories,

including alternative mind-body medicine (meditation, hypno-sis, qigong, Tai Chi, yoga, stress management groups); manipu-lative practices (chiropractic/osteopathic medicine, massage); other CAM practices (acupuncture, Ayurveda, chelation, ho-

meopathy, energy healing/reiki, movement therapies, natu-ropathy, traditional healers), and natural products (nonvitamin, nonmineral supplements).9 Our definition did not include use of prayer, special diets (e.g., Atkins), or support groups, since these therapies are commonly excluded from epidemiologic studies of CAM use.10,11 Respondents were asked about their use of each CAM therapy during the past 12 months, except for use of natural products, which was assessed only within the previous 3 months.

We further explored use of individual relaxation techniques (deep breathing exercises, progressive muscle relaxation, guid-ed imagery) and alternative therapies (meditation, yoga, Tai Chi, chiropractic, massage, homeopathy, acupuncture) that had a prevalence of use ≥ 1% in this sample. Among natural prod-ucts, we explored use of supplements commonly promoted for treatment of insomnia, including melatonin and valerian.19

Reasons for Use and Disclosure to Conventional Medical Professionals

To assess reasons for using relaxation techniques, sample adults were first asked to report the relaxation technique that they practiced most commonly. They were then asked to re-spond yes/no to the following 7 questions assessing reasons for use in the past 12 months: (1) to improve or enhance energy; (2) for general wellness/general disease prevention; (3) to improve/enhance immune function; (4) because conventional medical treatments did not help; (5) because conventional medical treat-ments were too expensive; (6) it was recommended by a con-ventional medical professional; and (7) it was recommended by family, friends, or coworkers. These 7 reasons for use were also asked for each of the other individual CAM therapies.

Next, we further examined whether adults reported using relaxation techniques and CAM specifically to treat their in-somnia. For each therapy used within the past year, respondents were asked, “Did you use (specific therapy) for a specific health problem or condition?” Those responding “yes” were then que-ried, “For what health problem did you use (specific therapy)?” Thus, all respondents who reported having insomnia symptoms within the past year and used a specific relaxation or CAM ther-apy for a health problem or condition were asked whether they had used that therapy to treat their insomnia.

Finally, respondents were also asked whether they disclosed their use to medical providers. Respondents were asked, “Dur-ing the past 12 months, did you let any of these conventional medical professionals know about your use of [each therapy]?” Conventional medical professionals included medical doctor (including specialists), doctor of osteopathy, nurse practitioner/physician assistant, psychiatrist, dentist, psychologist/social worker, and pharmacist.

Covariates of InterestWe considered several factors previously found to be associ-

ated with use of relaxation techniques and CAM, and those we believed a priori to be potential correlates.20,21 As described be-low, these factors play different roles (predictors, confounders) in different analyses. These factors included sociodemographic characteristics, healthcare access, illness burden, and health habits. Sociodemographic characteristics included age, sex, race/ethnicity, marital status, educational attainment, imputed


Relaxation Technique Use by Adults with Insomniafamily income provided by NHIS, and region of residence. We defined health care access using several proxies, including type of insurance (uninsured, Medicare, Medicaid, private), delayed care because of worries about cost (yes/no), and delayed care because couldn’t afford it (yes/no). We used indicators to cap-ture respondents’ illness burden, including perceived health sta-tus (excellent, very good, good, fair, poor), body mass index, number of emergency room visits in the past year (none, 1, > 1), having mobility impairment (no impairment, ≥ 1 limitation), and self-reported history of chronic medical conditions. We considered 34 medical conditions available in NHIS, includ-ing a history of conditions related to cardiovascular, respiratory, gastrointestinal, endocrine, neurologic, rheumatologic, gyneco-logic, and genitourinary systems; cancer; pain syndrome; and neuropsychiatric symptoms. Indicators of health habits were smoking status (current, former, never), alcohol intake (abstain, infrequent/light, moderate, heavy consumption),22 and physical activity level (low, moderate, high).23

Statistical AnalysesWe used bivariable analyses to compare adults with insom-

nia symptoms to those without these symptoms. We estimated the age-sex adjusted prevalence of use of relaxation techniques and CAM therapies. We further compared reasons for use and disclosure to conventional medical professionals among adults with and without insomnia symptoms.

Multivariable Methods to Identify Medical Conditions as Potential Confounders

Since few previous studies have examined the relation-ship between relaxation techniques use and chronic medical conditions, we first developed a multivariable logistic regres-sion model to identify medical conditions associated with use of relaxation techniques in the past 12 months in our sample. Conditions with a p-value < 0.25 on bivariable analyses were considered for inclusion in the multivariable model. We used a stepwise backward elimination process that adjusted for so-ciodemographic characteristics, insurance status, and health habits. Conditions with a Wald statistic p-value ≤ 0.05 were retained in our final model. We repeated this procedure using CAM as an outcome to assess for medical conditions associated with CAM use.

Multivariable Methods to Assess the Associations between Insomnia Symptoms and Relaxation Technique Use and CAM Use

Next, we used multivariable logistic regression to assess whether insomnia symptoms are independently associated with the use of relaxation techniques, after adjusting for sociodemo-graphic characteristics, health care access, illness burden (see Appendix for list of medical conditions), and health habits. Factors considered a priori to be important (i.e., imputed fam-ily income, insurance status) and factors with a Wald statistic p-value ≤ 0.05 were retained in our final model. We assessed confounding by adding each nonsignificant factor back into the model and assessed changes in the estimated β-coefficient for insomnia symptoms (our main factor of interest). Factors that changed this β-coefficient by ≥ 10% were considered con-founders and retained in our final model. We also employed this strategy to build a separate model examining the associa-

tions between insomnia symptoms status and CAM use (see Appendix for list of medical conditions).

Multivariable Methods to Identify Predictors of Relaxation Technique and CAM Use

We employed stepwise backward-elimination modeling pro-cedures to identify independent correlates of relaxation tech-niques use in adults with insomnia symptoms adjusting for sociodemographic characteristics, health care access, medi-cal conditions associated with relaxation techniques use, and health habits. Factors with a Wald statistic p-value ≤ 0.05 were retained in our final model. Lastly, we repeated this modeling procedure to identify correlates of CAM use among adults with insomnia symptoms. To facilitate comparisons of predictors between the 2 outcomes, our final models included a common set of predictors that included factors which were significantly associated with either outcome. We present this common set of predictors in tables 3 and 4 and note changes.

All analyses were performed using SAS-callable SUDAAN version 10.0 (Research Triangle Park, NC) to account for the complex sampling design, and results were weighted to re-flect US national estimates. The Beth Israel Deaconess Medi-cal Center Institutional Review Board determined this study to be exempt.

RESULTS

Sample CharacteristicsOverall, 4,415 (18.1%) of adults self-reported regularly

experiencing insomnia/difficulty sleeping within the past 12 months. table 1 presents the characteristics of adults by insom-nia symptom status. Compared with adults without insomnia symptoms, those with insomnia symptoms were more often older, female, had lower educational and income levels, and had poorer health habits (i.e., smoke, have heavy alcohol con-sumption, and low physical activity levels).

Use of Relaxation Techniquestable 2 lists the age-sex adjusted prevalence of relaxation

techniques and CAM use by insomnia symptom status. More than 1 in 5 respondents with insomnia (22.9%) used some form of relaxation technique annually, compared with 11.2% of adults without insomnia symptoms (p < 0.0001). After adjust-ment, adults with insomnia symptoms were more likely to use relaxation techniques (adjusted odds ratio [OR] 1.42, 95% con-fidence interval [CI] 1.25, 1.61). Among relaxation techniques, deep breathing exercise was the most commonly used practice by adults with insomnia symptoms, with lower prevalence of use of progressive muscle relaxation and guided imagery.

Reasons for Use and Disclosure to Conventional Medical Professionals

Figure 1 depicts reasons for relaxation techniques use by insomnia symptom status. Respondents both with and without insomnia most frequently cited using relaxation techniques for general wellness/disease prevention. Compared with adults without insomnia, adults with insomnia symptoms were more likely using relaxation techniques to treat a specific medical


SM Bertisch, RE Wells, MT Smith et al

Table 1—Characteristics of adults by insomnia symptom status (n = 23,358)

Without Insomnia (n = 18,943 )n (%)*

With Insomnia (n = 4,415 )n (%)*

Prevalence of Insomnia Symptoms

%*Chi-square

p-valueEstimated Population Size† N = 182,379,237 N = 40,426,178

Age (years) < 0.000118-29 3,982 (23.6) 558 (15.3) 12.530-39 3,645 (18.5) 668 (15.0) 15.240-49 3,537 (19.7) 883 (20.5) 18.750-59 3,004 (16.3) 896 (20.6) 21.960-69 2,328 (11.5) 652 (13.8) 21.070-79 1,496 (6.5) 449 (8.9) 23.3> 80 951 (3.9) 309 (5.9) 25.5

Sex < 0.0001Male 8,736 (50.0) 1,616 (40.2) 15.1Female 10,207 (50.0) 2,799 (59.8) 21.0

Education < 0.001< High School Graduate 3,320 (15.2) 901 (17.6) 20.5High School Graduate 5,270 (28.7) 1,245 (29.9) 18.8> High School Graduate 10,134 (56.1) 2,237 (52.5) 17.2

Race/Ethnicity < 0.0001Non-Hispanic White 11,050 (67.7) 2,840 (73.8) 19.5Non-Hispanic Black 2,976 (11.8) 641 (10.1) 16.0Hispanic 3,523 (13.9) 671 (11.1) 15.0Asian 1,067 (5.0) 140 (2.6) 10.3Other 327 (1.7) 123 (2.4) 23.6

Imputed Income < 0.0001< $20,000 4,285 (15.8) 1,436 (24.0) 25.2$20,000-$34,999 3,580 (16.5) 923 (20.0) 21.2$35,000-$65,000 5,119 (28.0) 1,038 (25.5) 16.8> $65,000 5,959 (39.6) 1,018 (30.5) 14.6

Insurance < 0.0001Uninsured 3,313 (16.8) 728 (16.4) 17.7Medicare 3,688 (16.1) 1,296 (26.3) 26.6Medicaid 1,045 (4.5) 418 (8.0) 28.0Private 9,504 (55.2) 1,663 (42.0) 14.4Other 1,333 (7.4) 298 (7.3) 18.1

Region < 0.001Northeast 3,217 (17.3) 702 (16.2) 17.1Midwest 4,253 (24.4) 953 (22.7) 17.0South 6,992 (36.3) 1,710 (38.3) 19.0West 4,481 (22.0) 1,050 (22.9) 18.7

Marital Status < 0.0001Married 10,032 (63.7) 1,955 (56.7) 16.5Widowed 1,676 (5.7) 582 (9.0) 26.0Divorced or separated 2,780 (9.6) 999 (16.3) 27.4Never Married 4,341 (21.0) 868 (18.0) 16.0

Smoking Status < 0.0001Current 3,238 (18.1) 1,130 (27.4) 25.2Former 3,771 (20.4) 1,151 (26.4) 22.3Never 11,601 (61.5) 2,075 (46.2) 14.3

n = sample size; N = Population estimate. †NHIS complex sampling scheme allows for estimates of U.S. population. *%, estimate weighted to reflect population. ‡Alcohol intake: abstainer (< 12 drinks in lifetime), rare (< 1 drink/month in past year), light (≤ 3 drinks/week), moderate (> 3 and ≤ 7 drinks/week for women, > 3 and ≤ 14 drinks/week for men), or heavy (> 7 drinks/week for women and > 14 drinks/week for men). §Physical activity levels: Vigorous = vigorous activity 2 times/week or moderate activity 4 times/week; moderate = vigorous activity 1 time/week or moderate activity 1–3 times/week, sedentary = no vigorous or moderate activity/week.

Table 1 continues on the following page


Relaxation Technique Use by Adults with Insomnia

condition, because a conventional medical professional recom-mended it, because conventional medical treatments did not help, and because medical treatments were too expensive. Few-er than 20% of all adults (19.1%) disclosed use of relaxation techniques to conventional medical professionals; adults with insomnia symptoms disclosed use of relaxation techniques to their providers more often than those without insomnia (26.3% vs. 15.7%; p < 0.0001). Although 29.9% (n = 291) of adults with insomnia reported using relaxation techniques for a specif-ic medical condition, fewer than 30 adults in our sample subse-quently responded using a relaxation technique specifically for insomnia. This small sample size precluded analysis of popula-tion estimates for use of relaxation techniques for specific treat-ment of insomnia.

Correlates of Use of Relaxation Techniques among Adults with Insomnia Symptoms

Among adults with insomnia symptoms, factors significant-ly associated with relaxation techniques use included moder-ate alcohol consumption, having mania/psychosis, excessive daytime sleepiness, anxiety, dental pain, neck pain, jaw pain, and asthma (table 3). Factors significantly associated with a lower likelihood of relaxation technique use included male

sex, lower educational attainment, family income < $20,000, low physical activity level, living in the South, and having hypertension. Of note, although jaw pain was a significant independent predictor of relaxation techniques, when factors independently associated with overall CAM use were added to the final model for comparison, jaw pain was no longer sig-nificant as shown in table 3. The most likely explanation for the shift to non-significance observed in these factors is high correlations between variables.

Use of Complementary and Alternative Medicine (CAM)Forty-five percent (45.2%) of adults with insomnia symptoms

reported using ≥ 1 CAM therapy in the past year, compared with 30.9% of adults without insomnia (p < 0.0001; table 2). After adjustment, adults with insomnia symptoms were more likely to use overall CAM (aOR 1.29, 95% CI [1.15, 1.44]). Adults with insomnia symptoms also used most categories of CAM (i.e., al-ternative mind-body medicine, natural products, manipulative practices, and other CAM practices) more than those without insomnia. Among individual CAM therapies, natural products were the most commonly used by adults with insomnia, followed by manipulative practices. The sleep-related natural products, melatonin and valerian, were used more frequently by adults

Without Insomnia (n = 18,943 )n (%)*

With Insomnia (n = 4,415 )n (%)*

Prevalence of Insomnia Symptoms

%*Chi-square

p-valueAlcohol Intake‡ < 0.0001

Abstainer 7,463 (38.7) 1,759 (38.7) 18.3Light 7,366 (41.6) 1,776 (42.7) 18.7Moderate 2,491 (14.8) 498 (12.3) 15.7Heavy 843 (4.9) 263 (6.2) 22.3

Physical Activity Level§ < 0.0001Low 7,940 (39.9) 2,104 (47.0) 21.0Moderate 2,879 (16.4) 692 (15.7) 17.6High 7,755 (43.7) 1,562 (37.3) 16.0

Neuropsychiatric SymptomsMania/psychosis 65 (0.3) 91 (2.3) 64.0 < 0.0001Frequently anxious 1,083 (5.7) 1,514 (34.1) 57.7 < 0.0001Excessive daytime sleepiness 933 (4.9) 1,482 (34.2) 60.6 < 0.0001Attention deficit disorder 339 (2.3) 217 (5.7) 35.9 < 0.0001

Pain SyndromesDental pain/gum disease 2,077 (11.3) 1,172 (25.4) 33.3 < 0.0001Neck pain 1,784 (9.4) 1,319 (29.5) 41.1 < 0.0001Jaw/ear pain 493 (2.7) 465 (10.2) 45.5 < 0.0001Low back pain 3,858 (20.3) 2,205 (49.3) 35.0 < 0.0001

Other Medical ConditionsHypertension 4,894 (24.0) 1,942 (41.2) 27.6 < 0.0001Asthma 1,742 (9.3) 812 (18.5) 30.1 < 0.0001Heartburn 3,244 (17.5) 1,798 (41.1) 34.3 < 0.0001Musculoskeletal strain 904 (5.1) 605 (14.7) 39.1 < 0.0001

n = sample size; N = Population estimate. †NHIS complex sampling scheme allows for estimates of U.S. population. *%, estimate weighted to reflect population. ‡Alcohol intake: abstainer (< 12 drinks in lifetime), rare (< 1 drink/month in past year), light (≤ 3 drinks/week), moderate (> 3 and ≤ 7 drinks/week for women, > 3 and ≤ 14 drinks/week for men), or heavy (> 7 drinks/week for women and > 14 drinks/week for men). §Physical activity levels: Vigorous = vigorous activity 2 times/week or moderate activity 4 times/week; moderate = vigorous activity 1 time/week or moderate activity 1–3 times/week, sedentary = no vigorous or moderate activity/week.

Table 1 (continued)—Characteristics of adults by insomnia symptom status (n = 23,358)


SM Bertisch, RE Wells, MT Smith et al

with insomnia; however, overall use of these natural products was very low. Small sample sizes (n < 30) precluded analysis of individual natural products for treatment of insomnia.

Reasons for CAM Use and Disclosure to Conventional Medical Professionals

Similar to use of relaxation techniques, both adults with and without insomnia symptoms most commonly cited high preva-

lence of CAM use for general wellness/disease prevention (Figure 2). Adults with insomnia symptoms more often cited CAM use to treat a specific medical condition, because a health care professional recommended it, because conventional medi-cal treatments did not help, and because medical treatments were too expensive, compared with adults without insomnia. Adults with insomnia also disclosed their CAM use to conven-tional medical professionals more frequently than adults with-

Table 2—Age-sex adjusted prevalence of relaxation technique and CAM Use by insomnia status, within past 12 months

TherapyWith Insomnia Symptoms,

(n = 4,415)Without Insomnia Symptoms,

(n = 18,943)Chi-square

p-valueRelaxation Techniques 997 (22.9) 2,082 (11.2) < 0.0001

Deep breathing exercises 939 (21.5) 1,977 (10.6) < 0.0001Progressive muscle relaxation 210 (5.2) 434 (2.4) < 0.0001Guided imagery 152 (3.6) 329 (1.8) < 0.0001

Overall CAM 1,936 (45.2) 5,615 (30.9) < 0.001

Alternative Mind-Body Medicine 879 (20.7) 2,256 (12.0) < 0.0001Meditation 678 (16.0) 1,476 (7.8) < 0.0001Yoga 289 (6.9) 1,053 (5.8) 0.07Tai chi 60 (1.4) 207 (1.0) 0.056

Natural Products* 1,059 (24.4) 2,920 (16.2) < 0.0001Melatonin 64 (1.5) 75 (0.4) < 0.0001Valerian 43 (1.0) 46 (0.3) < 0.0001

Manipulative Practices 766 (18.2) 2,374 (13.2) < 0.0001Chiropractic 464 (11.0) 1,389 (7.6) < 0.0001Massage 467 (11.0) 1,363 (7.5) < 0.0001

Other CAM Practices 270 (6.0) 620 (3.2) < 0.0001Homeopathy 121 (2.7) 174 (1.2) < 0.0001Acupuncture 104 (2.5) 240 (1.2) < 0.0001

Values are n (weighted populations%); CAM, Complementary and Alternative Medicine. *Natural products assessed within past three months. Not presented because overall prevalence < 1% or n < 50: Alexander Technique, Ayurveda, Biofeedback, Chelation, Energy Healing Therapy, Feldenkrais, Hypnosis, Naturopathy, Qigong, Stress Management Group, Curandero, Espiritista, Hierbero or Yerbera, Shaman, Native American Healer, Sobador.

Prev

alenc

e (es

timat

ed %

)

With Insomnia Symptoms (n = 997)Without Insomnia (n = 2,082)

General

wellness

/

diseas

e prev

ention

Improve

or enha

nce en

ergy

Treat s

pecific

medical

condi

tion

Friend, f

amily,

or co-work

er

recom

mend

Health c

are pr

ovider

recom

mend

Improve

or enha

nce im

mune

functio

n

Convent

ional t

reatment

s

did no

t help

Convent

ional t

reatment

s

too ex

pensiv

e

70

60

50

40

30

20

10

0

p = 0.26

p = 0.49 p < 0.001

p = 0.18p < 0.001

p < 0.001p < 0.001

p = 0.45

Figure 1—Reasons for relaxation technique use by insomnia symptom status

Prev

alenc

e (es

timat

ed %

)

With Insomnia Symptoms (n = 1,936)Without Insomnia (n = 5,615)

General

wellness

/

diseas

e prev

ention

Improve

or enha

nce en

ergy

Treat s

pecific

medical

condi

tion

Friend, f

amily,

or co-work

er

recom

mend

Health c

are pr

ovider

recom

mend

Improve

or enha

nce im

mune

functio

n

Convent

ional t

reatment

s

did no

t help

Convent

ional t

reatment

s

too ex

pensiv

e

70

60

50

40

30

20

10

0

p = 0.45

p = 0.18

p < 0.001

p = 0.39

p < 0.001

p < 0.001

p < 0.001

p = 0.13

Figure 2—Reasons for CAM use by insomnia symptom status



Table 3—Independent correlates of relaxation techniques use among adults with insomnia symptoms (n = 4,232)*Factors Adjusted Odds Ratio† 95% Confidence Interval

SociodemographicsSex

Male 0.54 0.44, 0.67Female 1.00 (reference)

RaceNon-Hispanic White 1.00 (reference)Non-Hispanic Black 0.65 0.51, 0.82Hispanic 0.64 0.48, 0.84Asian 1.04 0.64, 1.69Other 1.42 0.75, 2.66

Education< High School Graduate 0.44 0.31, 0.63High School Graduate 0.67 0.53, 0.85> High School Graduate 1.00 (reference)

Income< $20,000 0.66 0.49, 0.89$20,000-$34,999 0.77 0.57, 1.05$35,000-$65,000 0.90 0.67, 1.20> $65,000 1.00 (reference)

RegionNortheast 1.24 0.93, 1.65Midwest 1.61 1.27, 2.04South 1.00 (reference)West 1.48 1.15, 1.90

Health Habits

Physical Activity LevelLow 0.51 0.41, 0.64Moderate 1.08 0.84, 1.39High 1.00 (reference)

AlcoholAbstainer 1.00 (reference)Light 1.21 0.98, 1.52Moderate 1.46 1.08, 1.98Heavy 0.85 0.56, 1.28

Smoking StatusCurrent 0.77 0.61, 0.99Former 1.24 1.03, 1.50Never 1.00 (reference)

Medical ConditionsHypertension, history 0.78 0.63, 0.96Asthma, history 1.45 1.17, 1.81Heartburn, past 12 months 1.02 0.83, 1.24Dental pain, past 12 months 1.22 1.01, 1.49Neck pain, past 3 months 1.38 1.11, 1.71Jaw pain, past 3 months 1.32 0.98, 1.78Low back pain, past 3 months 1.11 0.91, 1.35Mania/psychosis, history 2.68 1.39, 5.14Anxiety, past 12 months 1.71 1.42, 2.19Excessive daytime sleepiness, past 12 months 1.28 1.06, 1.55Attention Deficit Disorder, history 1.27 0.79, 2.06Musculoskeletal strain, past 12 months 1.27 0.94, 1.73

Bold denotes significant predictors. *Observations with complete data. †Adjusted for sociodemographic factors, health care access, and health habits.


SM Bertisch, RE Wells, MT Smith et alout insomnia (47.4% vs. 40.7%, p < 0.0001). While 54.2% of adults (n = 1,040) with insomnia reported using a CAM therapy for a specific health problem or condition, the reported condi-tions were usually not insomnia. Only 1.8% (n = 82) reported using a CAM therapy for specific treatment of their insomnia.

Correlates of CAM Use among Adults with Insomnia Symptoms

Among adults with insomnia symptoms, factors significantly associated with CAM use included a former smoking history, having attention deficit disorder, heartburn, severe musculo-skeletal sprains or strains, dental pain, neck pain, or low back pain during the past 3 months (table 4). Factors significantly associated with a lower likelihood of relaxation techniques use included male sex, Hispanic or Non-Hispanic black race, lower educational attainment, family income < $35,000, low physical activity level, living in the South, currently smoking, and hav-ing hypertension. While musculoskeletal strain was a signifi-cant independent predictor of CAM use, it no longer achieved statistical significance once factors associated with relaxation use were included in the model.

DISCUSSION

We found that greater than 1 in 5 adults with insomnia symp-toms used relaxation techniques, and 45% of adults with insom-nia symptoms used at least one CAM therapy in the past year. Although adults with insomnia symptoms had higher likelihood of use of both relaxation techniques and CAM compared with adults without insomnia, fewer than 2% of adults with insom-nia symptoms said that they used CAM to treat their insomnia. Hence, although adults with insomnia are more likely to use relaxation techniques and CAM and have higher use for spe-cific medical conditions, their use is not driven by treatment of insomnia symptoms. General wellness/disease prevention was the most frequently cited reason for use of both relaxation tech-niques and CAM. Only 26% of adults with insomnia symptoms reported that they had told their conventional providers about their use of relaxation techniques, while 47% had disclosed their use of CAM. Among adults with insomnia symptoms, fac-tors associated with lower likelihood of relaxation techniques use were being male, having lower educational attainment, low physical activity, family income < $20,000, living in the South, and having a history of hypertension.

To our knowledge, this is the first study to exclusively exam-ine the national prevalence of relaxation techniques by adults with insomnia symptoms in the United States. Relaxation train-ing has been included in the AASM practice parameters as a standard therapy in the treatment of chronic insomnia since 2006.7 While we found that relaxation techniques are com-monly used by adults with insomnia symptoms, we also found that very few patients are using relaxation techniques to spe-cifically treat their insomnia. It is unclear why use of relaxation training for insomnia is uncommon, but may relate to factors at the practitioner and systems levels.24 For example, there is a limited supply of behavioral sleep medicine practitioners and relatively few sleep medicine clinics have fully integrated be-havioral sleep medicine services.25 There may also be lingering questions regarding the efficacy of relaxation techniques given

their generally weaker effect sizes compared with stimulus con-trol and multi-component CBT-I therapies for insomnia,23,24 and further challenges related to current reimbursement models for behavioral therapies.25

Additionally we found that nearly 75% of adults with in-somnia symptoms using relaxation techniques did not disclose their use to a conventional medical professional. Patients’ non-disclosure of relaxation techniques use may represent a missed opportunity for providers to discuss patients’ interest in and ex-perience with relaxation techniques, particularly in relation to insomnia. However, it is unclear why patients do not tell their providers about their use of relaxation techniques. Given the potential under use of relaxation techniques for this prevalent condition, more research is needed to understand the barriers to implementation of targeted relaxation therapies for patients with insomnia.

Alternative mind-body therapies such as meditation and Tai Chi, represent a diverse group of meditative practices, thought to evoke similar physiologic pathways to traditional relaxation training,26 and have been purported as treatments for insomnia.27 Nascent research in this area has been promising, though results vary by individual therapy. For example, despite popularized claims of the therapeutic benefit for meditation for sleep, data from randomized trials evaluating mindfulness meditation for sleep disturbances are mixed.28-33 Conversely, three recent ran-domized controlled studies evaluating the benefit of Tai Chi for sleep have demonstrated improvement in subjective sleep pa-rameters (i.e., Pittsburgh Sleep Quality Index [PSQI]).14-16 How-ever, larger randomized controlled trials with objective outcome measures are needed to improve our understanding of the ef-ficacy of alternative mind-body therapies, with particular atten-tion given to important methodological issues that are related to behavioral interventions, such as difficulty with blinding.

While we found that nearly one in four adults with insom-nia symptoms had used natural products within the past three months, we were surprised to find low prevalence of use of va-lerian and melatonin, which are natural products promoted spe-cifically for their hypnotic effects. It is possible that the observed low use of these sleep-related supplements may reflect the recent findings of two large randomized controlled trials, that failed to demonstrate improvement in subjective sleep parameters among patients with insomnia taking valerian compared with place-bo.34,35 Similarly, recent reviews on melatonin for treatment of primary insomnia failed to show robust improvements, particu-larly when accounting for circadian rhythm disorders.36 Despite low prevalence of use, it remains important to query patients about their use of natural products, since potential interactions exist between some natural products (e.g., valerian, melatonin) and commonly prescribed hypnotics. This is particularly impor-tant given that many patients with insomnia do not disclose their CAM use to conventional medical professionals.

LimitationsThe NHIS is a cross-sectional survey that relies on self-re-

port, with potential for misclassification and recall bias. This is of particular concern since insomnia symptoms can be episodic, and participants may not report having had insomnia within the past year if their symptoms have remitted at the time of their interview, but may report having used relaxation techniques



Table 4—Independent correlates of CAM use among adults with insomnia symptoms (n = 4,230)*Factors Adjusted Odds Ratio† 95% Confidence Interval

SociodemographicsSex

Male 0.68 0.57, 0.80Female 1.00 (reference)

RaceNon-Hispanic White 1.00 (reference)Non-Hispanic Black 0.66 0.52, 0.84Hispanic 0.66 0.48, 0.83Asian 1.05 0.65, 1.69Other 1.36 0.72, 2.60

Education< High School Graduate 0.40 0.31, 0.52High School Graduate 0.65 0.53, 0.78> High School Graduate 1.00 (reference)

Income< $20,000 0.67 0.53, 0.86$20,000-$34,999 0.68 0.52, 0.88$35,000-$65,000 0.83 0.66, 1.05> $65,000 1.00 (reference)

RegionNortheast 1.06 0.84, 1.33Midwest 1.28 1.00, 1.62South 1.00 (reference)West 1.55 1.25, 1.92

Health HabitsPhysical Activity Level

Low 0.60 0.50, 0.71Moderate 1.09 0.88, 1.37High 1.00 (reference)

AlcoholAbstainer 1.00 (reference)Light 1.32 1.10, 1.59Moderate 1.61 1.19, 2.18Heavy 1.42 1.02, 1.97

Smoking StatusCurrent 0.76 0.60, 0.97Former 1.25 1.03, 1.51Never 1.00 (reference)

Medical ConditionsHypertension, history 0.81 0.69, 0.95Asthma, history 1.10 0.88, 1.38Heartburn, past 12 months 1.22 1.02, 1.45Dental pain, past 12 months 1.34 1.15, 1.62Neck pain, past 3 months 1.63 1.32, 2.01Jaw pain, past 12 months 1.21 0.89, 1.64Low back pain, past 3 months 1.27 1.06, 1.52Mania/psychosis, history 1.01Anxiety, past 12 months 1.20 0.49, 2.06Excessive daytime sleepiness, past 12 months 1.01 0.82, 1.23Attention Deficit Disorder, history 1.57 1.08, 2.30Musculoskeletal strain, past 12 months 1.32 0.99, 1.76

Bold denotes significant predictors. *Observations with complete data. †Adjusted for sociodemographic factors, health care access and health habits.


SM Bertisch, RE Wells, MT Smith et alor CAM to alleviate their insomnia symptoms. Similarly, mis-classification may have occurred if a respondent reported us-ing relaxation techniques or CAM therapy to treat a medical condition associated with insomnia, even if it they had use it to improve their comorbid insomnia symptoms. Another im-portant limitation is that insomnia was not defined using es-tablished diagnostic criteria, and therefore we were not able to distinguish between insomnia symptoms and insomnia as a clinical condition. Furthermore, NHIS does not assess quantity or duration of relaxation technique or CAM use, which limits our ability to distinguish the characteristics of one-time users to those adults who practice relaxation techniques and CAM regularly; nor were we able to distinguish whether instructions for relaxation techniques use were provided by a health care professional or self-taught. Lastly, the 2007 NHIS was adminis-tered only in English and Spanish, and certain immigrant popu-lations that are less acculturated may have different patterns of use that were not captured.37 Despite these inherent limitations, our findings are based upon the most recent nationally repre-sentative data available on relaxation techniques and CAM use among adults with insomnia symptoms in the United States.

CONCLUSION

In summary, we found that despite high prevalence of use of relaxation techniques and CAM by adults with insomnia symptoms in the United States, few adults used relaxation tech-niques to treat their insomnia. Only one in four adults disclosed use of relaxation techniques to their conventional medical pro-fessionals. Better understanding of the barriers of relaxation techniques use and a refinement in our understanding of their efficacy and neurobiologic underpinnings would broaden our ability to provide targeted behavioral therapies to patients with sleep disturbances.

REFERENCES1. Ohayon MM. Epidemiology of insomnia: what we know and what we still need to

learn. Sleep Med Rev 2002; 6:97-114.2. Bonnet MH, Arand DL. Hyperarousal and insomnia: state of the science. Sleep

Med Rev 2010;14:9-15.3. Winkelman JW, Buxton OM, Jensen JE, et al. Reduced brain GABA in primary

insomnia: preliminary data from 4T proton magnetic resonance spectroscopy (1H-MRS). Sleep 2008;31:1499-506.

4. Kahn M, Baker BL. Weiss JM. Treatment of insomnia by relaxation training. J Abnorm Psychol 1968; 73:556-8.

5. Borkovec TD, Fowles DC. Controlled investigation of the effects of progressive and hypnotic relaxation on insomnia. J Abnorm Psychol 1973;82:153-8.

6. Nicassio P, Bootzin R. A comparison of progressive relaxation and autogenic training as treatments for insomnia. J Abnorm Psychol 1974; 83:253-60.

7. Morgenthaler T, Kramer M, Alessi C, et al. Practice parameters for the psycho-logical and behavioral treatment of insomnia: an update. An American Academy of Sleep Medicine report. Sleep 2006;29:1415-9.

8. Pearson NJ, Johnson LL, Nahin RL. Insomnia, trouble sleeping, and comple-mentary and alternative medicine: Analysis of the 2002 national health interview survey data. Arch Intern Med 2006;166:1775-82.

9. What Is Complementary and Alternative Medicine? National Center for Comple-mentary and Alternative Medicine.Available from: http://nccam.nih.gov/health/whatiscam/#types, accessed April 20, 2011.

10. Wells RE, Phillips RS, Schachter SC, McCarthy EP. Complementary and alter-native medicine use among US adults with common neurological conditions. J Neurol 2010;257:1822-31.

11. Bertisch SM, Wee CC, Phillips RS, McCarthy EP. Alternative mind-body thera-pies used by adults with medical conditions. J Psychosom Res 2009;66:511-9.

12. Sarris J, Byrne GJ. A systematic review of insomnia and complementary medi-cine. Sleep Med Rev 2011;15:99-106.

13. Khalsa SB. Treatment of chronic insomnia with yoga: a preliminary study with sleep-wake diaries. Appl Psychophysiol Biofeedback 2004;29:269-78.

14. Irwin MR, Olmstead R, Motivala SJ. Improving sleep quality in older adults with moderate sleep complaints: A randomized controlled trial of Tai Chi Chih. Sleep 2008;31:1001-8.

15. Wang C, Schmid CH, Rones R, et al. A randomized trial of tai chi for fibromyal-gia. N Engl J Med 2010;363:743-54.

16. Li F, Fisher KJ, Harmer P, Irbe D, Tearse RG, Weimer C. Tai chi and self-rated quality of sleep and daytime sleepiness in older adults: a randomized controlled trial. J Am Geriatr Soc 2004;52:892-900.

17. Centers for Disease Control. 2006, 2007 National Health Interview Survey (NHIS) Survey Description Documents, Appendices III and VII: Variance Esti-mation and Other Analytic Issues. Available from: http://www.cdc.gov/nchs/data/nhis/2006_2007var.pdf, accessed April 20, 2011.

18. National Center for Health Statistics. 2007 National Health Interview Survey (NHIS) Survey Description Document. Available from: ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Dataset_Documentation/NHIS/2007/srvydesc.pdf, ac-cessed April 20, 2011.

19. Bliwise DL, Ansari FP. Insomnia associated with valerian and melatonin usage in the 2002 National Health Interview Survey. Sleep 2007;30:881-4.

20. Eisenberg DM, Kessler RC, Foster C, Norlock FE, Calkins DR, Delbanco TL. Unconventional medicine in the United States. Prevalence, costs, and patterns of use. N Engl J Med 1993;328:246-52.

21. Wolsko PM, Eisenberg DM, Davis RB, Ettner SL, Phillips RS. Insurance cover-age, medical conditions, and visits to alternative medicine providers: results of a national survey. Arch Intern Med 2002;162:281-7.

22. Health, United States, 2007. U.S. Department Health and Human Services, Center for Disease Control and Prevention, National Center for Health Statis-tics. Available from: http://www.cdc.gov/nchs/data/hus/hus07.pdf, accessed April 20, 2011.

23. Kushi LH, Fee RM, Folsom AR, Mink PJ, Anderson KE, Sellers TA. Physical activity and mortality in postmenopausal women. JAMA 1997;277:1287-92.

24. Perlis ML, Smith MT. How can we make CBT-I and other BSM services widely available? J Clin Sleep Med 2008;4:11-3.

25. Pigeon WR, Crabtree VM, Scherer MR. The future of behavioral sleep medicine. J Clin Sleep Med 2007;3:73-9.

26. Ospina MB, Bond K, Karkhaneh M, et al. Meditation practices for health: state of the research. Evid Rep Technol Assess (Full Rep) 2007:1-263.

27. Britton W. Meditation for Sleep: Paradoxes and Promises. 2010 Jun In: Huffington Post. HuffPost Living [Internet]. Available from: http://www.huffingtonpost.com/willoughby-britton/sleep-tips-meditation-for_b_597600.html, accessed April 20, 2011.

28. Britton WB, Haynes PL, Fridel KW, Bootzin RR. Polysomnographic and sub-jective profiles of sleep continuity before and after mindfulness-based cognitive therapy in partially remitted depression. Psychosom Med 2010;72:539-48.

29. Shapiro SL, Bootzin RR, Figueredo AJ, Lopez AM, Schwartz GE. The efficacy of mindfulness-based stress reduction in the treatment of sleep disturbance in wom-en with breast cancer: an exploratory study. J Psychosom Res 2003;54:85-91.

30. Gross CR, Kreitzer MJ, Thomas W, et al. Mindfulness-based stress reduction for solid organ transplant recipients: a randomized controlled trial. Altern Ther Health Med 2010;16:30-8.

31. Gross CR, Kreitzer MJ, Reilly-Spong M, et al. Mindfulness-based stress reduc-tion versus pharmacotherapy for chronic primary insomnia: a randomized con-trolled clinical trial. Explore (NY) 2011;7:76-87.

32. Goldenberg DL, Kaplan KH, Nadeau MG, Brodeur C, Smith S, Schmid CH. A controlled trial of a stress-reduction, cognitive behavioral treatment program in fibromyalgia. J Musculoskel Pain 1994;2:53-6.

33. Roth B, Robbins D. Mindfulness-based stress reduction and health-related qual-ity of life: findings from a bilingual inner-city patient population. Psychosom Med 2004;66:113-23.

34. Fernandez-San-Martin MI, Masa-Font R, Palacios-Soler L, Sancho-Gomez P, Calbo-Caldentey C, Flores-Mateo G. Effectiveness of Valerian on insomnia: a meta-analysis of randomized placebo-controlled trials. Sleep Med 2010;11:505-11.

35. Bent S, Padula A, Moore D, Patterson M, Mehling W. Valerian for sleep: a sys-tematic review and meta-analysis. Am J Med 2006;119:1005-12.

36. Buscemi N, Vandermeer B, Hooton N, et al. The efficacy and safety of exog-enous melatonin for primary sleep disorders. A meta-analysis. J Gen Intern Med 2005;20:1151-8.

37. Hsiao AF, Wong MD, Goldstein MS, et al. Variation in complementary and alterna-tive medicine (CAM) use across racial/ethnic groups and the development of eth-nic-specific measures of CAM use. J Altern Complement Med 2006;12:281-90.



ACKNOWLEDGMENTSThe authors thank Julie K. Smith for her thoughtful editorial review. Work for this

study was conducted at Beth Israel Deaconess Medical Center, Boston, MA. Dr. Ber-tisch was supported by a K23 Career Development Award (K23AT005104). Dr. Wells was supported by an institutional National Research Service Award (T32AT000051) from the National Center for Complementary & Alternative Medicine (NCCAM) at the National Institutes of Health, USA. Dr. Smith was supported by the National Institute of Arthritis and Musculoskeletal Disease (R01s AR054871 and AR059410) Dr. Mc-Carthy was supported by R03AT002236, also from NCCAM. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Complementary & Alternative Medicine or the National Insti-tutes of Health, USA.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication June, 2011Submitted in final revised form May, 2012Accepted for publication May, 2012Address correspondence to: Suzanne M. Bertisch, M.D., M.P.H., Divisions of General Medicine and Primary Care, and Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, 1309 Beacon Street, 2nd Floor, Brookline, MA 02446; Tel: (617)-754-1404; Fax: (617)-754-1440; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. Dr. Smith is an owner and consultant to

B-Med Technologies, Inc. The terms of this arrangement are being managed by the Johns Hopkins University in accordance with its conflict of interest policies. The other authors have indicated no financial conflicts of interest.

APPENDIx

1. Medical conditions included in multivariable model as-sessing the association between insomnia symptoms and relaxation technique use: cardiovascular disease, chronic obstructive pulmonary disease, asthma, hayfever/sinusitis, excessive daytime sleepiness, anxiety, bipolar disorder, mania/psychosis, dental pain, arthritis, neck pain, low back pain, jaw/ear pain, musculoskeletal strain/sprain

2. Medical conditions included in multivariable model as-sessing the association between insomnia symptoms and CAM use: hayfever/sinusitis, asthma, heartburn, excessive daytime sleepiness, anxiety, bipolar disorder, dental pain, arthritis, neck pain, low back pain, jaw/ear pain


Study Objectives: First, to determine whether serum vitamin D levels were correlated with excessive daytime sleepiness (EDS) in patients with or without vitamin D defi ciency (VitDd). Second, to assess whether race affected the relation between vitamin D levels and EDS.Methods: Serum 25-hydroxyvitamin D (25OHD) was mea-sured by immunoassay in a consecutive series of 81 sleep clinic patients who complained of sleep problems and non-specifi c pain (25OHD < 20 ng/mL ≡ VitDd). Sleepiness was determined using the Epworth Sleepiness Scale score ([ESSs] ESSs ≥ 10 ≡ EDS). Correlations were assessed using Pearson r.Results: In patients without VitDd (25OHD ≥ 20 ng/mL), ESSs was inversely correlated with vitamin D concentration (r = 0.45, p < 0.05). The group consisted of 6% black patients, compared with 35% for the entire cohort. Among the patients who had

VitDd (25OHD < 20 ng/mL), ESSs was directly correlated with 25OHD in black (r = 0.48, p < 0.05) but not white patients. In black patients, mean ESSs in patients with VitDd were higher and 25OHD levels were lower p < 0.05).Conclusions: The results suggested the novel possibility that VitDd-related disease has a yet-to-be-identifi ed mechanis-tic role in the presentation of sleepiness, sleep disorders, or both. Further research is needed to clarify the mechanism(s) involved in producing the complex relationships noted.Keywords: Vitamin D, excessive daytime sleepiness, pain, sleep regulating substancesCommentary: A commentary on this article appears in this issue on page 699.Citation: McCarty DE; Reddy A; Keigley Q; Kim PY; Marino AA. Vitamin D, race, and excessive daytime sleepiness. J Clin Sleep Med 2012;8(6):693-697.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Understanding of the metabolic role of vitamin D has ex-panded greatly beyond its classically described effects

on gut and bone, leading to reports of numerous non-classical diseases associated with insuffi cient supply of vitamin D.1 Re-cent reports showed that vitamin D had potent immunomodu-latory activities2 and that low serum levels of vitamin D were linked to pulmonary disease,3 musculoskeletal pain,4 meta-bolic syndrome,5 hypertension,6 poor stress resilience,7 al-tered emotional functioning,8 and cognitive decline.9 Subjects classifi ed as defi cient in vitamin D generally exhibited more severe and/or chronic symptoms than subjects with smaller departures from the age- and gender-adjusted levels found in the clinically normal population.

More than half of the patients seen in our sleep medicine clinic who complained of sleep disruption and nonspecifi c somatic pain also exhibited vitamin D defi ciency.10 The as-sociation was particularly strong among black patients. In a case involving idiopathic central nervous system hypersomnia, treatment of a defi ciency in vitamin D resulted in a resolution of the clinical syndrome.11 These reports together with the rapidly developing perception of the ubiquitous role of vitamin D in metabolism suggest that suboptimal levels of vitamin D could cause or contribute to a pathologic level of centrally induced sleepiness, either directly or by means of chronic pain. Under the hypothesis that insuffi cient serum vitamin D was a causal or contributing factor in the development of excessive daytime sleepiness (EDS), we would expect to fi nd progressively lower levels of vitamin D among patients with progressively higher

Vitamin D, Race, and Excessive Daytime SleepinessDavid E. McCarty, M.D.1; Aronkumar Reddy, M.D.1; Quinton Keigley, B.S.2; Paul Y. Kim, Ph.D.1; Andrew A. Marino, Ph.D.1

1Division of Sleep Medicine, Department of Neurology, Louisiana State University Health Sciences Center, Shreveport, LA; 2School of Medicine, Louisiana State University, Shreveport, LA

levels of EDS, at least in patients who exhibited only moderate hypovitaminosis D.

Our fi rst aim was to determine whether serum vitamin D lev-els were correlated with sleepiness in patients with or without VitDd. Our second aim was to assess the role of race in the relationship between vitamin D levels and sleepiness.

METHODS

PatientsConsecutive new patients seen during routine consultation

visits in an academic sleep medicine clinic between April 2008 and October 2010 were interviewed within the context of a full

BRIEF SUMMARYCurrent Knowledge/Study Rationale: Defi ciency of Vitamin D is known to contribute to somatic pain symptoms and immune dysregulation (in-cluding inducing a relative elevation of circulating TNFα and NFĸB, both of which can result in subjective sleepiness symptoms). It is therefore mechanistically plausible that defi ciency of Vitamin D could contribute to poor quality sleep and/or symptoms of impaired wakefulness (such as excessive daytime sleepiness).Study Impact: We found a signifi cant, but complex, relationship be-tween 25-hydroxyvitamin D and subjective sleepiness, as measured by the Epworth Sleepiness Scale, suggesting the possibility that Vitamin D defi ciency may be a modifi able cofactor in the pathophysiology of exces-sive daytime sleepiness, sleep disorders, or both. Further research is needed to clarify these relationships.


DE McCarty, A Reddy, Q Keigley et al

sleep medicine history and physical exam. During this encoun-ter, patients completed a questionnaire in which they were que-ried about the presence of various symptoms potentially linked to sleep disruption, including moderate to severe musculoskel-etal pain contributing to sleep disruption, daytime discomfort, or both. Affirmative responses were verified during the inter-view by a physician board certified in Sleep Medicine (DM). Epworth Sleepiness Scale (ESS) surveys were completed by all patients as part of the same multi-item questionnaire. Those who answered affirmatively to the presence of pain and who agreed to undergo venous blood sampling for 25-hydroxyvita-min D (25OHD) were included in the study. ESS scores (ESSs) ≥ 10 were considered to indicate excessive daytime sleepiness (EDS). A total of 81 patients, all of whom were ultimately diagnosed with various sleep disorders (mostly obstructive sleep apnea) were included in the final analysis (table 1). All research-related procedures were approved by the institutional review board for human research.

Serum MeasurementsSerum concentration of 25OHD was determined by immu-

noassay employing the manufacturer’s specifications (DiaSorin Liason, Italy). A 25OHD < 20 ng/mL was regarded as indicat-ing a vitamin D deficiency (VitDd).12

StatisticsThe correlation between the ESS score and serum 25OHD

levels was evaluated using Pearson r. Mean serum levels were compared between races using the t-test. The significance level in both cases was p < 0.05.

RESULTS

In patients without VitDd (25OHD ≥ 20 ng/mL), sleepi-ness was inversely correlated with 25OHD levels (Figure 1).

The portion of the cohort without VitDd consisted of only 6% black patients (2 of 34), compared with 35% in the en-tire cohort (table 1). In this (predominantly white) portion of the cohort, vitamin D and ESSs were significantly correlated (r = 0.44, p < 0.05).

Among all patients with VitDd, the ESSs and 25OHD levels were uncorrelated. The VitDd portion of the cohort consisted of 26 black and 21 white patients. Mean ESSs among the black patients was significantly greater than among the white pa-tients, and 25OHD levels tended to be lower (Figure 2). Taken as a group, there was no relation between 25OHD and ESSs (Figure 2A), nor did such a relation exist for the white patients alone. For black patients, in contrast, a direct correlation was seen, with higher 25OHD levels associated with higher ESSs (r = 0.48, p < 0.05) (Figure 2A, black patients). On average, the ESSs was higher among blacks (Figure 2B).

DISCUSSION

Increasing evidence from clinical and basic research suggests that a suboptimal level of vitamin D constitutes a condition that disposes patients to the development of various diseases apart from the classical descriptions of bony demineralization.1,13 The spectrum of disease associated with low vitamin D is likely to include diseases of immune dysregulation, one manifestation of which could be excessive daytime sleepiness mediated by com-ponents of inflammatory cascades (see below). We therefore expected that progressively lower levels of 25OHD would be correlated with increased sleepiness. Further, because increased skin pigmentation is an established risk factor for low vitamin D14—and logically would increase risk for diseases associated with low vitamin D—we also expected that the relation be-tween 25OHD and sleepiness would depend materially on race.

Patients with a chief complaint of chronic nonspecific mus-culoskeletal pain have been shown to have low vitamin D levels

Table 1—Characteristics of cohortNumber of patients 81Sleepy patients (ESS Score ≥ 10) 59Non-sleepy patients (ESS Score < 10) 22Age (y) 47.9 ± 14Gender (N) Male 23 (28.3%) Female 77 (71.7%Race (N) Black 28 (35%) White 53 (65%)BMI (kg/m2) 35.6 ± 8.7Diagnoses (N) Obstructive sleep apnea 60 (74%) Hypersomnia 9 (11%) Insomnia 13 (16%) Restless legs syndrome 24 (30%) Other 4 (5%)

Data are presented as mean (± SD). N, number. Some patients had more than one diagnosis.

Black patientsWhite patients

0

10

20

20 25 30 35 40 45 50 55 60

Slee

pine

ss (E

pwor

th sc

ale)

Vitamin D (ng/mL)

24

r = 0.44 (p < 0.05)

Figure 1—Epworth Sleepiness Scale score (ESSs) as a function of vitamin D concentration in patients without vitamin D deficiency (25OHD ≥ 20 ng/mL)


Vitamin D, Race, and Daytime Sleepinessin some,15-18 but not all19,20 studies. We chose to study the rela-tionship between 25OHD and sleepiness in a group of patients we felt would have an elevated likelihood for low vitamin D. Our cohort consisted of patients seen in a sleep medicine spe-cialty clinic for various reasons, but who all admitted to the presence of moderate to severe pain either disrupting sleep, impairing daytime function, or both. This criterion for inclu-sion has important clinical implications, in that patients were not required to seek care for the pain symptoms, nor were pa-tients required to independently offer the complaint, but instead were only required to admit to its presence when asked. This not only creates a lower threshold for inclusion in the study, it also produces a cohort that is clinically heterogeneous, both in terms of the sleep-related diagnoses and with respect to the ana-tomic localization, duration, and severity of pain symptoms. In this admittedly diverse group, we asked whether sleepiness (as ascertained by ESS) was associated in a race-dependent manner with circulating 25OHD, and we found that it was. This is the first evidence of such a relationship which, if confirmed, has significant public health implications.

Sleepiness was inversely correlated with vitamin D con-centration among those with 25OHD levels ≥ 20 ng/mL, a threshold traditionally felt to represent a lower risk for the classical diseases of bone demineralization (Figure1).1 Black patients were scarcely present in this group, which prevented drawing conclusions regarding the relation between sleepi-ness and 25OHD among black patients. Among white pa-tients, however, the observed inverse relationship between 25OHD and ESSs in a range assumed to carry low risk for classical deficiency disease (Figure 1) suggested that the spectrum of non-classical illness may include magnification of hypersomnia symptoms.

The cumulative burden of VitDd may influence how this de-ficiency state interacts with sleepiness symptoms. This concept was supported by the results from those with 25OHD < 20 ng/mL—the threshold considered to increase the risk for classi-cal disease of vitamin D deficiency (Figure 2). This portion of our cohort was composed of 55% black patients, even though they represented only 35% of the cohort. Among all patients with 25OHD < 20 ng/mL, there was no significant relationship between 25OHD and ESSs. Among black patients, however, a significant direct relationship was found (r = 0.48, p < 0.05). The black patients also were significantly sleepier (Figure 2B).

The direct relationship between 25OHD and ESSs in black patients was unexpected, and ran counter to our original hy-pothesis of increasing sleepiness with lower 25OHD. One pos-sible mechanistic explanation for this result involves a greater degree of sympathetic stimulation and/or activation of the hypothalamic-pituitary-adrenal stress response axis, possibly due to increased pain in the context of a more severe burden of vitamin D deficiency-related disease. Another possibility in-volves an interaction between VitDd and other disorders, par-ticularly obstructive sleep apnea (OSA). VitDd is correlated with chronic rhinitis,21 tonsillar hypertrophy,22,23 and nonspe-cific myopathy24-26—all of which are known to increase the risk for OSA—and therefore represents a plausible novel factor that could lead to more severe OSA, as ascertained by frequency of apneas or hypopneas or severity of intermittent hypoxia. We had no reason to speculate on such a relationship a priori, and

consequently we did not collect data pertinent to OSA severity. The issue remains to be addressed.

The observation that ESSs and 25OHD levels < 20 ng/mL were correlated in black but not white patients suggests that a single blood draw may yield an incomplete picture of the true burden of disease. We previously showed that season of blood draw (summer vs. winter) was uncorrelated with VitDd in (~35% black, ~65% white) patients who admitted to nonspe-cific musculoskeletal pain during evaluation at a sleep medi-cine specialty clinic.27 In that cohort, however, white patients had higher mean summer 25OHD levels compared with winter levels (28.0 ± 13.5 and 23.1 ± 12.7 ng/mL, respectively). In

n = 26

Vitamin D Deficientn = 21

0

10

20

Slee

pine

ss (E

pwor

th sc

ale)

24

*

0

10

20

0 5 10 15 20

Slee

pine

ss (E

pwor

th S

cale)

Vitamin D (ng/mL)

24

Black patients White patients

Black patients White patientsA

B

Figure 2—Epworth Sleepiness Scale score (ESSs) as a function of vitamin D concentration in patients with vitamin D deficiency (25OHD < 20 ng/mL)

(A) Individual patients stratified by race. Black patients, r = 0.48 (p < 0.05). White patients (dashed line), r = -0.27 (NS). (B) Mean sleepiness ± SD. *p < 0.05.


DE McCarty, A Reddy, Q Keigley et alcontrast, there was no seasonal variation seen in black patients (12.7 ± 9.4 and 13.9 ± 6.0 ng/mL, respectively). These observa-tions suggest the possibility that a year-round burden of chronic VitDd may be more severe in black patients, thereby providing a basis for understanding different presentations and/or conse-quences of chronic diseases.

VitDd may contribute to symptoms of sleepiness via known sleep regulating substances, for example, TNFα, whose im-portance has been shown in clinical and experimental stud-ies.28,29 VitDd is also associated with upregulation of NFĸB,30 which functions as a master switch for inflammation and a trigger of the cellular inflammatory cascade resulting from intermittent hypoxia associated with OSA.31 NFĸB is also re-sponsible for the regulation of numerous substances known to exert homeostatic sleep pressure, including prostaglandin D2.32,33 Thus VitDd may not only be a cofactor for the devel-opment of cardiovascular morbidity associated with OSA, but may also play a role in the pathogenesis of EDS associated with the disease.

The study limitations included the small cohort size and the relative clinical heterogeneity of the patients. The post hoc rec-ognition of a potential relationship between vitamin D and OSA highlighted the need for data pertinent to this diagnosis. The lack of uniformity may have distorted the subjective experience of excessive sleepiness in an unpredictably variable manner from patient to patient, thereby obscuring the role of VitDd in the development of EDS symptoms. Further studies involving narrowed sleep disorder entry criteria (including quantification of severity of OSA) and quantification of pain location and se-verity levels would help clarify the connections between sleep disorders, EDS, vitamin D levels, and race.

This study was not designed to identify mechanisms. Low 25OHD may cause daytime impairment, but it is also possi-ble that patients with severe sleepiness may exhibit behaviors that increase the likelihood of low 25OHD. For example, sun avoidance may lessen the biosynthesis of vitamin D, which is related to direct sunlight exposure. In this view, however, a lin-ear inverse relationship would be expected for all patients, with higher ESS scores correlated with lower 25OHD values. Our discovery of a direct relationship among those with a greater burden of deficiency (i.e., blacks with 25OHD < 20 ng/mL) does not support the notion of sleepiness-related behavior pro-voking deficiency. Further studies would help clarify the bio-logic underpinnings relating 25OHD to sleepiness.

In conclusion, this is the first study to demonstrate a signifi-cant relationship between sleepiness and vitamin D. Though ESSs and 25OHD are related, the relationship is complex; the presence of VitDd changes the nature of this relationship com-pared to subjects without VitDd; among subjects with VitDd, the relationship between ESSs and 25OHD is markedly affect-ed by race.

REFERENCES1. Zittermann A, Gummert JF. Nonclassical vitamin D actions. Nutrients

2010;2:408-25.2. Khoo AL, Chai L, Koenen H, Joosten I, Netea M, van der Ven A. Translating

the role of vitamin D(3) in infectious diseases. Crit Rev Microbiol 2012; [Epub ahead of print].

3. Black PN, Scragg R. Relationship between serum 25-hydroxyvitamin D and pul-monary function in the third national health and nutrition examination survey. Chest 2005;128:3792-8.

4. Turner MK, Hooten WM, Schmidt JE, Kerkvliet JL, Townsend CO, Bruce BK. Prevalence and clinical correlates of vitamin D inadequacy among patients with chronic pain. Pain Med 2008;9:979-84.

5. Botella-Carretero JI, Alvarez-Blasco F, Villafruela JJ, Balsa JA, Vázquez C, Escobar-Morreale HF. Vitamin D deficiency is associated with the metabolic syndrome in morbid obesity. Clin Nutr 2007;26:573-80.

6. Forman JP, Giovannucci E, Holmes MD, et al. Plasma 25-hydroxyvitamin D lev-els and risk of incident hypertension. Hypertension 2007;49:1063-9.

7. Thomas MK, Lloyd-Jones DM, Thadhani RI, et al. Hypovitaminosis D in medical inpatients. N Engl J Med 1998;338:777-83.

8. Dean AJ, Bellgrove MA, Hall T, et al. Effects of vitamin D supplementation on cognitive and emotional functioning in young adults—a randomised controlled trial. PLoS One 2011;6:e25966.

9. Annweiler C, Fantino B, Schott AM, Krolak-Salmon P, Allali G, Beauchet O. Vita-min D insufficiency and mild cognitive impairment: cross-sectional association. Eur J Neurol 2012; [Epub ahead of print].

10. McCarty DE, Reddy A. Prevalence of vitamin D insufficiency/deficiency among sleep medicine patients complaining of somatic pain and correlation with day-time sleepiness. Sleep 2011;35 (Abstract Supplement):A230.

11. McCarty DE. Resolution of hypersomnia following identification and treatment of vitamin D deficiency. J Clin Sleep Med 2010;6:605-8.

12. Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266-81.13. Hoeck AD, Pall ML. Will vitamin D supplementation ameliorate diseases charac-

terized by chronic inflammation and fatigue? Med Hypotheses 2011;76:208-13.14. Clemens TL, Adams JS, Henderson SL, Holick MF. Increased skin pigment re-

duces the capacity of skin to synthesise vitamin D3. Lancet 1982;1:74-6.15. Lotfi A, Abdel-Nasser AM, Hamdy A, Omran AA, El-Rehany MA. Hypovita-

minosis D in female patients with chronic low back pain. Clin Rheumatol 2007;26:1895-901.

16. Plotnikoff GA, Quigley JM. Prevalence of severe hypovitaminosis D in pa-tients with persistent, nonspecific musculoskeletal pain. Mayo Clin Proc 2003;78:1463-70.

17. Heidari B, Shirvani JS, Firouzjahi A, Heidari P, Hajian-Tilaki KO. Association between nonspecific skeletal pain and vitamin D deficiency. Int J Rheum Dis 2010;13:340-6.

18. Badsha H, Daher M, Ooi Kong K. Myalgias or non-specific muscle pain in Arab or Indo-Pakistani patients may indicate vitamin D deficiency. Clin Rheumatol 2009;28:971-3.

19. de Rezende Pena C, Grillo LP, das Chagas Medeiros MM. Evaluation of 25-hy-droxyvitamin D serum levels in patients with fibromyalgia. J Clin Rheumatol 2010;16:365-9.

20. Tandeter H, Grynbaum M, Zuili I, Shany S, Shvartzman P. Serum 25-OH vitamin D levels in patients with fibromyalgia. Isr Med Assoc J 2009;11:339-42.

21. Abuzeid W, Akbar N, Zacharek M. Vitamin D and chronic rhinitis (publish ahead of print). Curr Opin Allergy Clin Immunol 2012;12:13-7.

22. Nunn JD, Katz DR, Barker S, et al. Regulation of human tonsillar T-cell prolifera-tion by the active metabolite of vitamin D3. Immunology 1986;59:479-84.

23. Reid D, Morton R, Salkeld L, Bartley J. Vitamin D and tonsil disease--preliminary observations. Int J Pediatr Otorhinolaryngol 2011;75:261-4.

24. Glerup H, Mikkelsen K, Poulsen L, et al. Hypovitaminosis D myopathy with-out biochemical signs of osteomalacic bone involvement. Calcif Tissue Int 2000;66:419-24.

25. Prabhala A, Garg R, Dandona P. Severe myopathy associated with vitamin D deficiency in western New York. Arch Intern Med 2000;160:1199-203.

26. Russell JA. Osteomalacic myopathy. Muscle Nerve 1994;17:578-80.27. McCarty DE, Reddy A, Kiegley Q, Kim PY, Cohen S, Marino A. Vitamin D de-

ficiency is common in southern U.S. ambulatory sleep medicine clinic patients who complain of somatic pain, irrespective of season. In: Poster presented at: Southern Sleep Society; 2012; Miramar Beach, Florida, 2012.

28. Churchill L, Rector DM, Yasuda K, et al. Tumor necrosis factor alpha: activity dependent expression and promotion of cortical column sleep in rats. Neurosci-ence 2008;156:71-80.

29. Peterson CA, Heffernan ME. Serum tumor necrosis factor-alpha concentrations are negatively correlated with serum 25(OH)D concentrations in healthy women. J Inflamm (Lond) 2008;5:10.

30. Jablonski KL, Chonchol M, Pierce GL, Walker AE, Seals DR. 25-Hydroxyvitamin D deficiency is associated with inflammation-linked vascular endothelial dys-function in middle-aged and older adults. Hypertension 2011;57:63-9.


Vitamin D, Race, and Daytime Sleepiness31. Ryan S, Taylor CT, McNicholas WT, Ryan S, Taylor CT, McNicholas WT. Selec-

tive activation of inflammatory pathways by intermittent hypoxia in obstructive sleep apnea syndrome. Circulation 2005;112:2660-7.

32. Chen Z, Gardi J, Kushikata T, Fang J, Krueger JM. Nuclear factor-kappaB-like activity increases in murine cerebral cortex after sleep deprivation. Am J Physiol 1999;276:R1812-8.

33. Krueger JM, Szentirmai E, Kapas L. Biochemistry of sleep function: a paradigm for brain organization of sleep. In: Amlaner CJ, Fuller PM, eds. Basics of sleep guide. 2nd ed. Westchester, IL: Sleep Research Society, 2009:69-74.

ACKNOWLEDGMENTSThis work was supported by the Department of Neurology, LSUHSC-Shreveport.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication March, 2012Submitted in final revised form May, 2012Accepted for publication May, 2012Address correspondence to: David E. McCarty, M.D., Division of Sleep Medicine, LSUHSC-Shreveport, P.O. Box 33932, 1501 Kings Highway, Shreveport, LA 71130-3932; Tel: (318) 675-8568; Fax: (318) 675-6382; E-mail: [email protected]





CO

MM

EN

TAR

Y

The paper in this issue by McCarty and colleagues1 is note-worthy. The authors reported that patients without defi cien-

cy in vitamin D showed an inverse correlation with excessive daytime sleepiness (EDS), and a minority of this sample was black. In those patients with vitamin D defi ciency, the som-nolence was directly correlated in black individuals. Thus, it is important to stress that the remarkable observation that ex-cessive daytime somnolence may be associated with vitamin D defi cits does indeed open a broad range of possibilities that radically depart from the classic context of bone demineraliza-tion. Furthermore, this article is in accordance with recent stud-ies, in which vitamin D is discussed as an underlying factor in several sleep- and chronobiology-related issues.2-4

Although it is too early to make assertive statements regard-ing the multiple consequences vitamin D defi cits might engen-der, such outcomes factually extend beyond the continuous fi xation of calcium by osteoblasts and its sequestering from the bone by osteoclasts to balance body acidity. A possible media-tor of such consequences could be melanin, which plays a key role in the synthesis of the fi rst form in the vitamin D chain. Melamine is known to affect sleep patterns, making it all the more reasonable to assume that altered concentrations of vita-min D, be they caused by skin complexion or otherwise, will affect the sleep pattern, particularly in light of the observation that symptoms of sleep deprivation are present in the form of excessive daytime somnolence.

We congratulate the authors for their insight and remain hopeful that the observations made will help spur a new line of research to determine whether it is sleep deprivation that causes the defi ciency in vitamin D, or if it is the shortage of vitamin D

Vitamin D as an Underlying Factor in Sleep-Related IssuesCommentary on McCarty et al. Vitamin D, race, and excessive daytime sleepiness.

J Clin Sleep Med 2012;8:693-697.Monica Levy Andersen, M.Sci., Ph.D.; Sergio Tufi k, M.D., Ph.D.

Departamento de Psicobiologia – Universidade Federal de São Paulo, São Paulo, Brazil

that interferes with the sleep pattern. In conclusion, this study is timely and presents results that will increase the interest in a broad range of fi elds.

CITATIONAndersen ML; Tufi k S. Vitamin D as an underlying factor in sleep-related issues. J Clin Sleep Med 2012;8(6):699.

REFERENCES1. McCarty DE, Reddy A, Keigley Q, Kim PY, Marino AA. Vitamin D, race, and

excessive daytime sleepiness. J Clin Sleep Med 2012;8:693-7.2. Santos LG, Pires GN, Azeredo Bittencourt LR, Tufi k S, Andersen ML. Chronobi-

ology: relevance for tuberculosis. Tuberculosis (Edinb) 2012;92:293-300.3. Gominak SC, Stumpf WE. The world epidemic of sleep disorders is linked to

vitamin D defi ciency. Med Hypotheses 2012;79:132-5.4. Huang W, Shah S, Long Q, Crankshaw AK, Tangpricha V. Improvement of pain,

sleep, and quality of life in chronic pain patients with Vitamin D supplementation. Clin J Pain 2012 June 13. [Epub ahead of print].

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication October, 2012Accepted for publication October, 2012Address correspondence to: Monica Levy Andersen, M.Sci., Ph.D., Department of Psychobiology - Universidade Federal de São Paulo, Rua Napoleão de Barros, 925 Vila Clementino - SP- 04024-002, São Paulo, Brazil; Tel: (55-11) 2149-0155; Fax: (55-11) 5572-5092; E-mail: [email protected]/[email protected]

DISCLOSURE STATEMENTThe authors have indicated no fi nancial confl icts of interest.


and the Epworth Sleepiness Scale [ESS]5,6), to the exclusion of wakefulness inability (diffi culty maintaining wakefulness) and fatigue, in sleep disordered patients.

The MSLT defi nes sleepiness as the ability to fall asleep in a dark room when asked to do so. According to a review paper to establish Standards of Practice for the clinical use of the MSLT and MWT (Maintenance of Wakefulness Test), “the wide range in MSL makes it diffi cult to establish a spe-cifi c threshold value for excessive sleepiness or to discriminate

Study Objectives: Routine assessment of daytime function in Sleep Medicine has focused on “tendency to fall asleep” in soporifi c circumstances, to the exclusion of “wakefulness inability” or inability to maintain wakefulness, and fatigue/tiredness/lack of energy. The objective was to establish re-liability and discriminant validity of a test for wakefulness inability and fatigue, and to test its superiority against the criterion standard for evaluation of sleepiness—the Epworth Sleepiness Scale (ESS).Methods: A 12-item self-administered instrument, the Sleepi-ness-Wakefulness Inability and Fatigue Test (SWIFT), was de-veloped and administered, with ESS, to 256 adults ≥ 18 years of age (44 retook the tests a month later); consecutive patients with symptoms of sleep disorders including 286 with obstruc-tive sleep apnea ([OSA], apnea-hypopnea index ≥ 5/h sleep on polysomnography [PSG]), 49 evaluated with PSG and mul-tiple sleep latency test for narcolepsy and 137 OSA patients treated with continuous positive airway pressure (CPAP).

Results: SWIFT had internal consistency 0.87 and retest in-traclass coeffi cient 0.82. Factor analysis revealed 2 factors—general wakefulness inability and fatigue (GWIF) and driving wakefulness inability and fatigue (DWIF). Normal subjects dif-fered from patients in ESS, SWIFT, GWIF, and DWIF. SWIFT and GWIF (but not DWIF) had higher area under ROC curve, Youden’s index, and better positive and negative likelihood ra-tios than ESS. ESS, SWIFT, GWIF, and DWIF improved with CPAP. Improvements in SWIFT, GWIF, and DWIF (but not ESS) were signifi cantly correlated with CPAP compliance.Conclusions: SWIFT is reliable and valid. SWIFT and its fac-tor GWIF have a discriminant ability superior to that of the ESS.Keywords: Sleepiness, wakefulness inability, fatigue, obstruc-tive sleep apnea, Epworth Sleepiness ScaleCitation: Sangal RB. Evaluating sleepiness-related daytime function by querying wakefulness inability and fatigue: Sleep-iness-Wakefulness Inability and Fatigue Test (SWIFT). J Clin Sleep Med 2012;8(6):701-711.


SC

IEN

TIFI

C I

NV

ES

TIG

ATI

ON

S

Sleepiness is a commonly reported public health problem. In a 2002 National Sleep Foundation (NSF) poll,1 7% reported

sleepiness almost every day and another 9% a few days a week, for a total of 16%. Seventeen percent reported having dozed off while at the wheel of a vehicle, and 1% reported having an accident because they dozed off or were too tired (emphasis added). In other words, accidents were attributed to either doz-ing off or being tired. According to the National Highway Traf-fi c Safety Administration (NHTSA),2 “NHTSA data indicate in recent years there have been about 56,000 crashes annually in which driver drowsiness/fatigue (emphasis added) was cited by police. Annual averages of roughly 40,000 nonfatal injuries and 1,550 fatalities result from these crashes.” Sleep apnea patients report not just daytime sleepiness, but also being tired, fatigued, or having a lack of energy, and these complaints may be more frequent than sleepiness in sleep apnea.3 Sleepiness (and its adverse effect of auto accidents) may be multifactorial, with elements of inability maintaining wakefulness when necessary or desired (wakefulness inability), tendency to doze off in sopo-rifi c circumstances, and fatigue/tiredness/lack of energy. How-ever, the specialty of Sleep Medicine has tended to focus rather exclusively on routine evaluation of daytime function by mea-suring sleepiness defi ned as tendency to fall asleep in soporifi c circumstances (using the multiple sleep latency test [MSLT]4

Evaluating Sleepiness-Related Daytime Function by Querying Wakefulness Inability and Fatigue: Sleepiness-Wakefulness

Inability and Fatigue Test (SWIFT)R. Bart Sangal, M.D., F.A.A.S.M.

Sleep Disorders Institute, Sterling Heights, MI; Oakland University William Beaumont School of Medicine, Rochester, MI

BRIEF SUMMARYCurrent Knowledge/Study Rationale: There is not a questionnaire in-strument to measure wakefulness inability or diffi culty staying awake in situations where staying awake is desirable, or one that simultaneously addresses symptoms related to pathological sleepiness and to fatigue/tiredness/lack of energy in sleep disorders patients. Conceivably, what could be better than being able to fall asleep when one wants to (low MSLT [Mean Sleep Latency Test] and even high ESS [Epworth Sleepi-ness Scale]), but be able to stay awake when one wants to and feel refreshed (not tired) during the day?Study Impact: A reliable and valid self-rating instrument (Sleepiness-Wakefulness Inability and Fatigue Test or SWIFT) was created and shown to be superior to the criterion standard for sleepiness (ESS) with regard to specifi city, sensitivity and discriminate ability. It should be add-ed to the ESS in evaluating daytime consequences of sleep disorders.


RB Sangalpatients with sleep disorders from non-patients.”7 Further, “the MSL change between pre- and post-treatment for an individual is probably meaningful, although comparison of these data with the normative values is not helpful.”7 This may be related to the use of a behavior (the ability to fall asleep quickly when lying down in a dark room) that may be a desirable and adap-tive trait rather than abnormal state. The MWT8,9 sought to cor-rect this by asking subjects to try and stay awake in a dimly lit room. However, staying awake sitting in a dimly lit room doing nothing is not particularly advantageous (as opposed to staying awake when sitting in a dimly lit car and driving). The MSLT and MWT measure different abilities,10,11 and treatment may improve “wakefulness inability” (the MWT) more than ”sleep tendency” (MSLT).12 Both the MSLT and MWT require a large investment in time and resources.

There are self-rating questionnaire instruments that aim to measure sleepiness, the most commonly used being the ESS. The ESS queries for tendency to fall asleep in a variety of circumstances, often soporific. Sanford et al.13 have reported the distribution of the ESS. In their sample of normal subjects, median ESS was 7-8, and 30.7% of normal subjects without insomnia reported an ESS score ≥ 10, the widely used cutoff for abnormal excessive sleepiness. It has been shown that the ESS does not measure the same ability as the MWT.14,15 In patients who are severely sleepy on the MWT, the ESS was in-sensitive to the level of sleepiness as measured by the MWT. The ESS may16 or may not17 be correlated with the MSLT. A sample of 10,000 subjects with 71% response rate showed no correlation between the ESS and the adverse consequence of automobile accidents, although there was a correlation be-tween dozing off while stopped in traffic (item 8 on ESS) and automobile accidents.18

Sleep disordered patients report fatigue, lack of energy, and tiredness in addition to sleepiness.3 Fatigue, tiredness and lack of energy are largely interchangeable terms, as suggested by Merriam-Webster Dictionary’s19 definitions of fatigue as “wea-riness or exhaustion from labor, exertion, or stress”, and tired as “drained of strength and energy: fatigued often to the point of exhaustion.” Although these symptoms may be separable from sleepiness/wakefulness inability (wakeful being defined by Merriam-Webster Dictionary as “not sleeping or able to sleep”), it is not clear that sleep disordered patients, the general public and the NSF,1 or the police and the NHTSA,2 can sepa-rate these symptoms clearly.

Conceivably, what could be better than being able to fall asleep when one wants to (low MSLT and even high ESS), but be able to stay awake when one wants to and feel refreshed (not fatigued) during the day? Is the ability to fall asleep easily a pathological problem or an adaptive ability?20 This leads to the question of whether we should be querying instead the ability to stay awake when desired, along with fatigue.

There does not seem to be a questionnaire instrument to measure wakefulness inability or difficulty staying awake in situations where staying awake is desirable. Although fatigue inventories such as the 83-item Multidimensional Fatigue In-ventory (MFI)21 and its 30-item short form (MFSI-sf)22 exist, there is no single short questionnaire that simultaneously ad-dresses symptoms related to pathological sleepiness and to fatigue/tiredness/lack of energy in sleep disordered patients.

Thus, these other domains of sleep disorder complaints are not routinely queried or measured.

The hypothesis was that a self-rating instrument for assess-ing wakefulness inability and fatigue can be created that is reli-able (with good internal consistency and test-retest reliability in normal subjects) and valid (with good ability to discriminate between normal individuals and sleep disordered patients, and to show improvement with treatment of sleep disorders such as obstructive sleep apnea [OSA]), and that such a test incorporat-ing wakefulness inability and fatigue is superior to the criterion standard for sleepiness (ESS) with regard to specificity, sensi-tivity, and discriminant ability.

METHODS

A 12-item questionnaire (Sleepiness-Wakefulness Inability and Fatigue Test, or SWIFT) was developed. Subscale A has 6 questions related to difficulty staying awake/wakefulness in-ability in different situations that might affect performance or cause adverse consequences; subscale B has 6 questions related to fatigue, tiredness or lack of energy in different situations that might affect performance or cause adverse consequences, all answered on a 4-level (scored 0-3) Likert scale. Items were pre-pared by the author based on apparent face validity, with the inclusion of more than one item related to driving. The SWIFT is shown in table 1.

Normal SubjectsAfter obtaining approval from the Wayne State University

Human Investigations Committee, adult subjects (age ≥ 18 years) were recruited over a period of 10 weeks by means of a group e-mail to medical students at Wayne State University as well as by personal solicitation of subjects in public places such as malls and parks. After reading an informational sheet, they were asked to fill out questionnaires seeking their gen-der, age, educational level, occupation, race, height, weight, medical/psychiatric problems, medicines taken, sleep habits, and presence or absence of sleep symptoms including snor-ing, observed or perceived apneic episodes in sleep, insomnia, fatigue, and sleepiness. They were also asked to complete the SWIFT and the ESS. Subjects willing to be contacted again in a month to retake the questionnaire were asked for contact in-formation, and were contacted after a month to again complete the questionnaire.

A total of 403 subjects filled out the questionnaire. Subjects with incomplete questionnaires (49) were excluded. In order to examine the normal range of sleepiness, wakefulness inabil-ity, and fatigue, it was decided to exclude subjects with issues known to affect sleepiness, wakefulness inability, and fatigue, such as CNS-active or psychotropic medicines (53), CNS dis-orders (3), untreated depression (19), and history of observed/perceived apneic episodes in sleep (23). This left 256 normal subjects (87 male, 169 female; age range 18-92 years; 190 White, 26 Black, 5 Hispanic, 11 Asian American, 6 South Asian American, 18 Other; 4 with less than high school education, 23 high school graduates, 52 with some college, 101 with college degrees, and 76 with graduate degrees; National Statistics so-cioeconomic classification included: 10 higher professional or managerial, 50 lower professional or managerial, 26 intermedi-


Sleepiness-Wakefulness Inability and Fatigue Test

ate occupations, 3 small employers and own account workers, 12 lower supervisory and technical, 4 routine occupations, 36 retired or unemployed, and 115 students. Forty-four of them retook the SWIFT and ESS a month later.

Determining ReliabilityTo determine internal consistency, Cronbach α was calcu-

lated for the SWIFT (and the ESS) using data from the normal subjects. To determine test-retest reliability, intraclass coeffi-cients were calculated using the normal subjects with test and retest data. If these tests showed good reliability (Cronbach α and intraclass coefficient > 0.8), factor analysis with varimax rotation was performed using SWIFT data.

Additional Analyses in Normal SubjectsCorrelation between ESS and SWIFT was calculated. Males

and females were compared with regard to age, BMI, time in bed, SWIFT and ESS, as were subjects who completed the questionnaire again and those who did not. Correlations were calculated between SWIFT and ESS on the one hand, and age, time in bed and BMI on the other. If there was a significant cor-relation between SWIFT and age, normal subjects were divided into two age groups to determine if the correlation persisted. If it did not, factor analysis and correlations were performed again by age group.

Correction for Multiple Statistical AnalysesIn order to correct for multiple statistical analyses, the false

discovery rate method was applied to primary but not to condi-

tional analyses (analyses performed only as a result of another statistically significant analysis).23 This method rank orders the p-values of the analyses. For k analyses, one p value of 0.05/k was accepted as significant, one p value of 0.05/k-1 was ac-cepted as significant, and so on.

Determining ValidityAfter establishing good reliability for the SWIFT, validity

was determined for SWIFT and its factors using data from nor-mal subjects and sleep disordered patients, and SWIFT and its factors were compared with ESS to establish superiority.

Sleep Disordered PatientsAll new patients presenting with sleep disorder symptoms to

an AASM accredited Sleep Disorders Center during a 15-month period were administered the SWIFT and ESS at the time of the initial evaluation. If appropriate, they underwent a polysom-nography (PSG) to evaluate for OSA using American Academy of Sleep Medicine (AASM) scoring criteria,24 or a PSG with MSLT to evaluate for Narcolepsy. New patients who had previ-ously been evaluated/treated for OSA anywhere were excluded from analysis. All patients with significant OSA (apnea-hypop-nea index (AHI) ≥ 15/h sleep, or ≥ 5/h sleep with comorbid sleepiness, hypertension, or cardiovascular disease, evaluated as new patients over the 15-month period, were offered stan-dard treatment. Patients opting for continuous positive airway pressure (CPAP) were administered a PSG with CPAP titration, and were prescribed CPAP at the optimum determined pressure. At a follow-up office visit between 1 and 3 months after CPAP

Table 1—Sleepiness-Wakefulness Inability and Fatigue Test (SWIFT)Name: Date:

If this is the first time you are filling this out for this study, answer keeping in mind the last one month. Mark symptoms present only if they have been present for at least one month. If you have filled out this questionnaire before for this study, answer based on the period of time since you last filled this out. Be sure to answer every question to the best of your ability. This questionnaire refers to your usual way of life in recent times. Even if you have not done some of these things recently try to work out how they would have affected you. Use the following scale to choose the most appropriate number for each situation:

A How much of a problem is it to stay awake during the day (or your usual wake period if you sleep during the day)?Not at all Just a little Pretty Much Very Much

1 Struggling to stay awake during the day 0 1 2 32 Difficulty staying awake while driving 0 1 2 33 Difficulty staying awake stopped at a traffic signal 0 1 2 34 Difficulty staying awake at work or while doing tasks 0 1 2 35 Difficulty staying awake while reading or studying 0 1 2 36 Difficulty staying awake in social situations 0 1 2 3

B How much of a problem has fatigue, tiredness or lack of energy been for you?Not at all Just a little Pretty Much Very Much

1 Feeling tired when at work or while doing tasks 0 1 2 32 Lack of energy during social situations 0 1 2 33 Struggling with fatigue during the day 0 1 2 34 Feeling tired while reading or studying 0 1 2 35 No energy to do tasks that do not absolutely have to be done 0 1 2 36 Difficulty driving because of fatigue 0 1 2 3© R. Bart Sangal, M.D., 2011. Permission is given to copy for use in medical care of personal patients.

The total score is the sum of the scores of all 12 items (A1-6, B1-6), with a maximum possible score of 36.


RB Sangalprescription, they were again administered the SWIFT and ESS, and CPAP compliance data were downloaded if available.

Data were available for 286 adult subjects (age ≥ 18 years, 192 males, 94 females) who presented with sleep disorder symptoms and had documented OSA (AHI ≥ 5/h sleep). After excluding subjects with AHI ≥ 5/h sleep on the PSG preceding the MSLT (who were counted among the adult OSA subjects), and patients who were administered the MSLT on CPAP, data were available for 49 adult subjects (17 males, 32 females) who were adminis-tered PSG with MSLT for suspicion of narcolepsy (because of unexplained sleepiness with no clinical evidence of OSA, off CNS-active medicines, including psychotropic medicines, for five half-lives). These 49 subjects were independent of and not a subset of the 286 OSA patients. Repeat ESS and SWIFT and compliance data from follow-up visits after CPAP initiation were available for 137 adult OSA patients (98 males, 39 females).

Determining Discriminant ValidityTo determine discriminant validity, SWIFT (and SWIFT fac-

tor) and ESS scores were compared between the normal sub-jects and the OSA patients, as well as between normal subjects and patients evaluated for suspicion of narcolepsy. To further determine discriminant validity, SWIFT (and SWIFT factor) and ESS scores were compared in OSA patients before and af-ter CPAP treatment, and the number of patients with abnormal SWIFT (and SWIFT factor) and ESS scores before and after CPAP treatment were compared. Correlations were calculated between compliance and improvement in ESS, SWIFT, and SWIFT factors identified by factor analysis.

Statistics of Diagnostic TestsA diagnostic test identifies 2 groups: those with the disor-

der and those without the disorder. The sensitivity (also called true-positive rate) of a test is the probability of a positive test in the disordered population, whereas the specificity is the prob-ability of a negative test in a disorder-free population; and the value (1-specificity) is also called the false-positive rate. The positive predictive value is the probability of a subject with a positive test having the disorder. The negative predictive value is the probability that a subject with a negative test does not have the disorder. Sensitivity and specificity are not affected by prevalence of the disorder, whereas positive and negative predictive values are affected. This means sensitivity and speci-ficity can be accurately calculated when using a normal sample and a sample of disordered subjects, but predictive values can-not (they require a population sample for accurate calculations). A test with higher sensitivity and specificity than another is the superior test. However, a test may have higher sensitivity but lower specificity than another, or vice versa. Therefore, com-paring 2 tests requires combining specificity and sensitivity. The likelihood ratio of a positive test or positive likelihood ratio (ρ+) is the ratio of the probability of a positive test in a disor-dered subject (true-positive rate) to the probability of a positive test in a normal subject (false-positive rate), and is calculated as [sensitivity/(1-specificity)]. The likelihood ratio of a negative test or negative likelihood ratio (ρ–) is the ratio of the probabil-ity of a negative test in a disordered subject to the probability of a negative test in a normal subject, calculated as [(1-sensitiv-ity)/specificity]. Both likelihood ratios may range from 0 to α.

A positive likelihood ratio < 1 indicates a useless test, as does a negative likelihood ratio > 1. With a diagnostic test based on a continuously measured variable, a decision or cutoff thresh-old allows sensitivity and specificity to be combined into the Youden’s index γ, which is the true positive rate minus the false positive rate, calculated as [sensitivity-(1-specifity)] (also writ-ten as [sensitivity + specificity -1]). A perfect test (with sen-sitivity and specificity of 1) results in a Youden’s index of 1, whereas a useless test has a Youden’s index of 0. When the cut-off threshold is increased, the proportions of both true positives (sensitivity) and false positives (1-specificity) will increase. The receiver operating characteristic (ROC) is a graph of sen-sitivity against (1-specificity). A perfect test has an area under the ROC (AUC) of 1, a useless test has an AUC of 0.5. Bewick et al. have written an excellent but concise and simple discus-sion of these tests.25 Since neither specificity nor sensitivity are affected by prevalence of the disorder, therefore positive and negative likelihood ratios, Youden’s index and AUC are also not affected by prevalence when they are applied to population-based samples.

Determining Test SuperiorityTo determine which test is superior in discriminating nor-

mal subjects from sleep disordered patients, the AUC for the 2 tests can be compared. Visually, if the ROC for one test is entirely within the ROC of another test, then the second test seems certainly superior. Confidence intervals can be obtained for the AUC and statistical comparisons performed between the AUC for 2 tests using various nonparametric and binormal methods.26,27 A nonparametric distribution and correlated ROCs were assumed for this report. Different methods for calculating confidence intervals of Youden’s index and likelihood ratios also exist, and a general method based on constant χ2 boundaries was used for this analysis.28 However, since the AUC is not depen-dent on a cutoff threshold, and diagnostic decisions are based on cutoff thresholds, a test may have a smaller AUC yet be more suitable than another, and the AUC is not a suitable measure of diagnostic excellence. Youden’s index is a more suitable mea-sure of diagnostic superiority.29 Although Youden’s index is a good single summary measure of comparison between two tests, positive and negative likelihood ratios are an even better test of superiority.30 If test A has positive likelihood ratio greater than that for test B, and negative likelihood ratio lesser than that for test B, then test A is superior overall to test B. AUC, Youden’s index, and positive and negative likelihood ratios (along with confidence intervals) were calculated for the ESS, SWIFT and its factors, using data from the normal subjects and OSA pa-tients, as well as data from the normal subjects and patients eval-uated for suspicion of narcolepsy. The AUC for SWIFT and ESS were compared. If there was a significant difference in favor of SWIFT, then the AUC for its factors were also compared with the AUC for ESS.

To determine which test is superior in showing improvement with treatment, effect sizes may be used. However, it is more important clinically to have a cutoff score (such as mean + 1 SD), above which the test is considered high and below which it is considered normal, and to show superiority in conversion of patients from abnormal scores before treatment to normal scores after treatment, therefore χ2 analyses were also performed.


Sleepiness-Wakefulness Inability and Fatigue TestAdditional Analyses in Sleep Disordered Patients

Correlation coefficients were calculated between ESS and SWIFT and its factors on the one hand, and sleep effi-ciency, arousal index, periodic limb movement arousal index (PLMAI), and AHI and lowest oxygen saturation (for OSA pa-tients), mean sleep latency (MSL) and sleep onset REM periods [SOREMPS] (for patients evaluated for narcolepsy), with cor-rections for false discovery rate for multiple tests.

RESULTS

ReliabilityCronbach α using data from the 256 normal subjects was

0.87 for SWIFT and 0.80 for ESS. Upon retest, the intraclass correlation coefficient for SWIFT was 0.82 (p < 0.001), and for ESS 0.91 (p < 0.001).

Factor AnalysisFactor analysis of SWIFT with varimax rotation revealed 2

factors: Factor 1 (36% of variance) included 9 items (A1, A4, A5, A6, B1, B2, B3, B4, B5), and was called general wakeful-ness inability and fatigue (GWIF) based on the generality of the items. Factor 2 (20% of variance) included 3 items (A2, A3, B6), and was called driving wakefulness inability and fatigue (DWIF) based on these items being related to driving. table 2 gives the factor loadings.

Additional Analysis of Normal SubjectsESS was correlated with SWIFT (r = 0.64, p < 0.001). There

was no difference between males and females in age, BMI, hours in bed, SWIFT, or ESS. Those who completed the ques-tionnaires again had a lower BMI (24.1 vs. 26.6, equal vari-ances not assumed, t = 2.9, df = 68.6, p = 0.005) than those who did not, but there were no other significant differences. After

the false discovery rate correction, there were significant nega-tive correlations between age and SWIFT (r = -0.25, p < 0.001) as well as ESS (r = -0.14, p = 0.024). There were no other sig-nificant corrected correlations.

Upon dividing the subject group into young adults (ages 18-45, n = 188) and middle-aged to older adults (age > 45, n = 68), ESS and SWIFT were no longer correlated with age in either group. table 3 gives the measures of central tendency and dis-persion for age, hours in bed, BMI, SWIFT, ESS, and the GWIF and DWIF factors for the 188 young adults and 68 middle-aged to older adults; the 85th percentile generally corresponds very closely to mean + 1 SD, and 95th percentile to mean + 2 SD.

Table 3—Normal subjects: mean, SD, medians and percentilesMean SD Median Mean + 1 SD 85th percentile Mean + 2 SD 95th percentile

Young adults (18-45 y)Age 27.4 6.4 26.0 33.8 35.0 40.2 41.6BMI 25.6 6.0 24.4 31.6 30.9 37.6 37.6Time in bed (h) 7.7 1.6 7.5 9.3 8.5 10.9 11.0ESS 6.8 4.1 5.0 10.9 11.0 15.0 15.0SWIFT 7.1 4.9 6.0 12.0 12.0 16.9 17.6GWIF 6.6 4.4 6.0 11.0 10.0 15.4 15.6DWIF 0.6 1.1 0.0 1.7 1.8 2.8 3.0

Middle-aged to older adults (> 45 y)Age 63.2 12.8 26.0 76.0 79.3 88.8 87.1BMI 27.8 5.9 24.4 33.7 33.3 39.6 40.8Time in bed (h) 7.8 1.4 7.5 9.2 9.0 10.6 10.1ESS 5.8 4.0 5.0 9.8 10.7 13.8 14.6SWIFT 4.7 4.3 6.0 9.0 9.0 13.3 13.6GWIF 4.2 3.9 6.0 8.1 8.0 12.0 13.0DWIF 0.5 0.9 0.0 1.4 1.7 2.3 3.0

Mean, median, standard deviation, mean + 1 and + 2 standard deviations, and 85th and 95th percentiles are shown by age group for age, BMI, time in bed, the ESS, SWIFT, and the general wakefulness inability and fatigue (GWIF), and driving wakefulness inability and fatigue (DWIF) factors of the SWIFT.

Table 2—Factor analysis matrix of SWIFT

Factor 1 Factor 2B4 0.79 0.25B3 0.77 0.19B1 0.74 0.02B2 0.73 0.09A5 0.73 0.21B5 0.72 0.04A1 0.69 0.22A4 0.56 0.10A6 0.46 0.32A2 0.04 0.90A3 0.15 0.84B6 0.26 0.80

Rotated component matrix for normal subjects by age group: A1, A4, A5, A6, B1, B2, B3, B4, B5 load on Factor 1 (GWIF: general wakefulness inability and fatigue), with a maximum possible score of 27. A2, A3 and B6 load on Factor 2 (DWIF: driving wakefulness inability and fatigue), with a maximum possible score of 9.


RB SangalUpon performing factor analysis separately for each age group, there were the same 2 factors for young adults, accounting for the same 36% and 20% of variance. For middle-aged to older adults, Factor 2 remained the same (A2, A3, B6) and accounted for 19% of variance. Factor 1 separated into 3 factors. The new Factor 1 (A1, A4, B1, B2, B3, B5) accounted for 27% of variance, while A5 and B4 (17% of variance), and A6 (10% of variance) became new separate factors, suggesting that wakefulness inability/fa-tigue while reading or studying may separate from general wake-fulness inability/fatigue in middle-aged to older adults.

OSA PatientsOf the 286 patients with AHI ≥ 5, 86 were young adults (ages

18-45 years) and 200 were middle-aged to older adults (age > 45 years). The 188 normal young adults differed significantly from the 86 young adults with AHI ≥ 5 in age, SWIFT, GWIF, DWIF, and ESS. table 4 gives the means and standard devia-tions. table 5 gives the AUC and, using cutoffs at greater than mean + 1 SD (> 10 for ESS, > 12 for SWIFT, > 11 for GWIF, and > 1 for DWIF), the sensitivity, specificity, positive like-lihood ratio, negative likelihood ratio, and Youden’s index. SWIFT and GWIF but not DWIF had better AUC, positive and negative likelihood ratios and Youden’s index than ESS. Figure 1 shows that the ROC for ESS was entirely within the ROC for SWIFT and GWIF. However, there was no significant difference between AUC for ESS and SWIFT.

The 68 normal middle-aged to older adults differed signifi-cantly from the 200 middle-aged to older adults with AHI ≥ 5 in age, SWIFT, GWIF, DWIF, and ESS. table 4 gives the means and standard deviations. table 5 gives the AUC and, using cut-offs at greater than mean + 1 SD (> 9 for ESS, > 9 for SWIFT, > 8 for GWIF, and > 1 for DWIF), the sensitivity, specific-ity, positive likelihood ratio, negative likelihood ratio, and Youden’s index. SWIFT, GWIF, and DWIF had better AUC, positive and negative likelihood ratios, and Youden’s index than ESS. Figure 2 shows the ROC for ESS was entirely within the ROC for SWIFT and GWIF. The AUC was significantly higher

Table 5—Normal subjects vs. OSA patients: indices of test superiorityYoung adults (18-45 y) Middle-aged to older adults (> 45 y)

Value 95% CI Value 95% CIAUC for ESS 0.660 0.585-0.734 0.688 0.620-0.757AUC for SWIFT*† 0.743 0.676-0.809 0.793 0.736-0.850AUC for GWIF*† 0.743 0.677-0.810 0.793 0.733-0.851AUC for DWIF 0.652 0.578-0.725 0.669 0.602-0.737sensitivity for ESS 0.453 0.368-0.535 0.430 0.395-0.457sensitivity for SWIFT* 0.488 0.405-0.563 0.565 0.532-0.587sensitivity for GWIF* 0.453 0.372-0.525 0.540 0.508-0.561sensitivity for DWIF 0.384 0.302-0.462 0.430 0.397-0.453specificity for ESS 0.819 0.780-0.857 0.809 0.706-0.887specificity for SWIFT* 0.872 0.834-0.906 0.868 0.771-0.932specificity for GWIF* 0.888 0.851-0.921 0.882 0.787-0.943specificity for DWIF* 0.846 0.808-0.882 0.853 0.755-0.921ρ+ for ESS 2.508 1.670-3.731 2.249 1.364-4.093ρ+ for SWIFT* 3.826 2.445-6.016 4.269 2.351-8.794ρ+ for GWIF* 4.060 2.503-6.654 4.590 2.419-9.967ρ+ for DWIF 2.488 1.573-3.910 2.924 1.638-5.848ρ– for ESS 0.667 0.543-0.811 0.705 0.613-0.856ρ– for SWIFT* 0.587 0.482-0.713 0.501 0.443-0.607ρ– for GWIF* 0.615 0.516-0.737 0.525 0.466-0.625ρ– for DWIF 0.729 0.610-0.864 0.668 0.593-0.800γ for ESS 0.273 0.147-0.392 0.239 0.102-0.343γ for SWIFT* 0.361 0.239-0.469 0.433 0.303-0.519γ for GWIF* 0.342 0.224-0.446 0.422 0.295-0.503γ for DWIF 0.229 0.110-0.344 0.283 0.151-0.375

*Value superior to value for ESS in both age groups. †AUC significantly higher than that for ESS in middle-aged to older adults. ρ+ and ρ– are likelihood ratios for positive and negative test. γ, Youden’s index. Mean + 1 SD used as cut-offs for calculation of sensitivity, specificity, ρ+, ρ– and γ. Young adults: ESS > 10, SWIFT > 12, GWIF > 11, DWIF > 1. Middle-aged to older adults: ESS > 9, SWIFT > 9, GWIF > 8, DWIF > 1.

Table 4—Normal subjects vs. OSA patients: means and SDYoung adults

(18-45 y)Middle-aged to older

adults (> 45 y)

NormalWith OSA (AHI ≥ 5) Normal

With OSA (AHI ≥ 5)

Age* 27.4 (6.4) 37.3 (6.2) 63.2 (12.8) 58.5 (8.9)ESS* 6.8 (4.1) 9.7 (5.5) 5.8 (4.0) 9.2 (5.5)SWIFT* 7.1 (4.9) 12.9 (7.3) 4.7 (4.3) 12.0 (7.9)GWIF* 6.6 (4.4) 11.4 (6.2) 4.2 (3.9) 10.4 (6.6)DWIF* 0.6 (1.1) 1.4 (1.8) 0.5 (0.9) 1.6 (2.0)

*Normal different from OSA patients at p < 0.001 for both age groups. All values mean (SD).



for SWIFT (z = 2.36, p = 0.018) than for ESS, and the AUC was also significantly higher for GWIF than for ESS (z = 2.35, p = 0.019), but not for DWIF.

There were no significant correlations found in the OSA pa-tients between ESS, SWIFT, GWIF, or DWIF on the one hand, and sleep efficiency, PLMAI, or lowest oxygen saturation on the other. SWIFT (r = 0.16, p = 0.006), GWIF(r = 0.15, p = 0.009) and DWIF (r = 0.14, p = 0.023), but not ESS, were significantly correlated with arousal index. ESS (r = 0.14, p = 0.018) and GWIF (r = 0.14, p = 0.022), but not SWIFT or DWIF, were significantly correlated with AHI.

CPAP TreatmentESS, SWIFT, GWIF, and DWIF improved significantly in

patients on CPAP in both age groups (36 young adults: t = 7.1, df = 35, p < 0.001 for ESS, t = 7.0, df = 35, p < 0.001 for SWIFT, t = 7.4, df = 35, p < 0.001 for GWIF, t = 3.4, df = 35, p = 0.002 for DWIF; 101 middle-aged to older adults: t = 9.7, df = 100, p < 0.001 for ESS, t = 12.2, df = 100, p < 0.001 for SWIFT, t = 11.5, df = 100, p < 0.001 for DWIF, t = 7.7, df = 100, p < 0.001 for DWIF). Effect sizes and 95% confidence intervals for the 137 subjects were as follows: ESS 0.96 (0.07, 1.63), SWIFT 1.07 (-0.20, 1.98), GWIF 1.04 (-0.03, 1.82), DWIF 0.75 (0.43, 0.93). One hundred fourteen of 137 (83.2%) subjects were compliant (use ≥ 4 h/night) for ≥ 70% of nights. Compli-ance was significantly correlated with improvement in SWIFT (r = 0.21, p = 0.015), GWIF (r = 0.18, p = 0.034) and DWIF (r = 0.18, p = 0.032), but not ESS (r = 0.11, p = 0.216). Im-provement in SWIFT (r = 0.22, p = 0.011) and GWIF (r = 0.24, p = 0.004) were also significantly correlated with AHI, but im-

provement in DWIF or ESS were not. table 6 gives by age group the pre- and post-treatment data, as well as numbers above and below the cutoffs before and after treatment, effect sizes, and χ2 statistics. SWIFT, GWIF, DWIF, and ESS were all valuable in demonstrating conversion from abnormal to normal values with CPAP use.

Patients Evaluated for NarcolepsyOf 49 patients evaluated with PSG and MSLT for evaluation

of narcolepsy, 37 were young adults (ages 18-45 years), and 12 were middle-aged to older adults (age > 45 years). Ten of the young adults and none of the middle-aged to older adults met MSLT criteria for diagnosis of narcolepsy—a 20% positive diag-nostic rate, which is comparable to the 20% (170 of 832) positive diagnostic rate for the MSLT reported earlier in sleepy patients without OSA.31 The young adults with narcolepsy were signifi-cantly younger than the young adults without narcolepsy (24.1, SD 5.3 vs. 38.4, SD 13.6), but did not significantly differ from them in ESS, SWIFT, GWIF, or DWIF. The 188 normal young adults differed significantly from the 37 young adults evaluated for suspicion of narcolepsy in SWIFT, GWIF, DWIF, and ESS, but not in age. table 7 gives the means and standard deviations. table 8 gives the AUC and, using cutoffs at greater than mean + 1 SD (> 10 for ESS, > 12 for SWIFT, > 11 for GWIF, and > 1 for DWIF), the sensitivity, specificity, positive likelihood ratio, neg-ative likelihood ratio, and Youden’s index. SWIFT, GWIF, and DWIF had better AUC, positive and negative likelihood ratios, and Youden’s index than ESS. Figure 3 shows the ROC for ESS was entirely within the ROC for SWIFT and GWIF. The AUC was significantly higher for SWIFT (z = 2.29, p = 0.022) than for

ROC CurveNormal subjects vs. OSA pts (18-45 yrs)

Source of the curve:

ESSSWIFTGWIFDWIF

1 - Specificity

Sens

itivit

y

0.0 0.2 0.4 0.6 0.8 1.0

1.0

0.8

0.6

0.4

0.2

0.0

Figure 1—ROC curves for ESS, SWIFT, GWIF, and DWIF for normal subjects vs. OSA patients in age group 18-45 years

ESS, Epworth Sleepiness Scale; SWIFT, Sleepiness-Wakefulness Inability and Fatigue Test; GWIF, general wakefulness inability and fatigue Factor; DWIF, driving wakefulness inability and fatigue factor.

ROC CurveNormal subjects vs. OSA pts (age > 45 yrs)


ESSSWIFTGWIFDWIF

1 - SpecificitySe

nsiti

vity

0.0 0.2 0.4 0.6 0.8 1.0

1.0

0.8

0.6

0.4

0.2

0.0

Figure 2—ROC curves for ESS, SWIFT, GWIF, and DWIF for normal subjects vs. OSA patients in age group > 45 years

ESS, Epworth Sleepiness Scale; SWIFT, Sleepiness-Wakefulness Inability and Fatigue Test; GWIF, general wakefulness inability and fatigue factor; DWIF, driving wakefulness inability and fatigue factor.


RB Sangal

ESS. The AUC was also significantly higher for GWIF than for ESS (z = 2.07, p = 0.038), but not for DWIF.

The 68 normal middle-aged to older adults differed sig-nificantly from the 12 middle-aged to older adults evaluated for suspicion of narcolepsy in age, SWIFT, GWIF, DWIF, and ESS. table 7 gives the means and standard deviations. table 8 gives the AUC and, using cutoffs at greater than mean

+ 1 SD (> 9 for ESS, > 9 for SWIFT, > 8 for GWIF, and > 1 for DWIF), the sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and Youden’s index. SWIFT, GWIF and DWIF had better AUC, positive and negative likelihood ratios, and Youden’s index than ESS. Figure 4 shows the ROC for ESS was entirely within the ROC for SWIFT and GWIF. However, there was no significant difference between AUC for ESS and SWIFT.

There were no significant correlations found in the pa-tients evaluated for narcolepsy between ESS, SWIFT, GWIF, or DWIF on the one hand and sleep efficiency, AHI, PLMAI, arousal index, or number of SOREMPs on the MSLT, on the other. ESS but not SWIFT, GWIF, or DWIF was significantly negatively correlated with mean sleep latency on the MSLT (r = -0.408, p = 0.004).

DISCUSSION

The SWIFT has high internal consistency as shown by high Cronbach α, and high test-retest reliability shown by high intra-class coefficient. Thus, the SWIFT is a reliable test.

Table 6—Before and on CPAP treatment

Mean SD# with

high scores# with normal

scores

Effect size (95% confidence intervals)

χ2 for high and normal scores p-value for χ2

Young adults (age 18-45 y)Before CPAP

Sample size 36Age 37.56 6.20AHI 28.77 28.78ESS 10.33 5.16 18 18SWIFT 13.44 7.04 19 17GWIF 11.97 6.11 16 20DWIF 1.47 1.68 14 22

On CPAPCompliance % 75.2 24.8ESS* 4.67 4.15 3 33 1.23 (-0.46,2.58) 13.2 < 0.001SWIFT* 5.67 6.07 5 31 1.20 (-1.10,3.18) 10.6 0.001GWIF* 5.28 5.46 5 31 1.17 (-0.83,2.95) 6.7 0.010DWIF* 0.39 0.90 3 33 0.81 (-0.13,1.01) 7.7 0.006

Middle-aged to older adults (age > 45 y)Before CPAP

Sample size 101Age 58.92 8.98AHI 47.11 3ESS 9.78 5.45 47 54SWIFT 12.83 7.74 62 39GWIF 11.02 6.49 61 40DWIF 1.81 2.00 48 53

On CPAPCompliance % 81.7 22.3ESS* 5.64 3.93 19 82 0.88 (-0.19,1.64) 16.4 < 0.001SWIFT* 6.13 5.22 26 75 1.02 (-0.49,2.04) 24.7 < 0.001GWIF* 5.52 4.43 25 76 0.99 (-0.27,1.86) 24.8 < 0.001DWIF* 0.60 1.18 16 85 0.74 (0.35-0.97) 22.0 < 0.001

*Significant improvement, p < 0.01.

Table 7—Normal subjects vs. MSLT patients: means and SDYoung adults

(18-45 y)Middle-aged to older

adults (> 45 y)

NormalEvaluated for narcolepsy Normal

Evaluated for narcolepsy

Age 27.4 (6.4) 29.2 (8.1) 63.2 (12.8) 55.0 (6.7)ESS* 6.8 (4.1) 11.7 (5.1) 5.8 (4.0) 11.6 (6.9)SWIFT* 7.1 (4.9) 19.8 (7.8) 4.7 (4.3) 16.8 (7.0)GWIF* 6.6 (4.4) 16.8 (6.3) 4.2 (3.9) 14.5 (5.5)DWIF* 0.6 (1.1) 3.0 (2.2) 0.5 (0.9) 2.3 (2.3)

*Normal different from patients evaluated for narcolepsy, significant at p < 0.05 for both age groups. All values mean (SD).



Table 8—Normal subjects vs. MSLT patients: indices of test superiorityYoung adults (18-45 y) Middle-aged to older adults (> 45 y)

Value 95% CI Value 95% CIAUC for ESS 0.767 0.676-0.857 0.761 0.595-0.927AUC for SWIFT*† 0.898 0.831-0.965 0.938 0.877-0.999AUC for GWIF*† 0.888 0.818-0.958 0.939 0.882-0.996AUC for DWIF* 0.833 0.748-0.918 0.780 0.618-0.942sensitivity for ESS 0.649 0.494-0.781 0.500 0.237-0.758sensitivity for SWIFT* 0.838 0.695-0.929 0.833 0.544-0.970sensitivity for GWIF* 0.784 0.638-0.889 0.833 0.546-0.969sensitivity for DWIF* 0.730 0.578-0.848 0.583 0.307-0.818specificity for ESS 0.818 0.789-0.845 0.809 0.762-0.854specificity for SWIFT* 0.872 0.844-0.890 0.868 0.817-0.892specificity for GWIF* 0.888 0.860-0.909 0.882 0.832-0.906specificity for DWIF* 0.846 0.816-0.869 0.853 0.804-0.894ρ+ for ESS 3.587 2.335-5.045 2.615 0.996-5.200ρ+ for SWIFT* 6.563 4.463-8.469 6.296 2.964-8.953ρ+ for GWIF* 7.017 4.549-9.755 7.083 3.241-10.355ρ+ for DWIF* 4.731 3.136-6.472 3.967 1.569-7.745ρ– for ESS 0.468 0.259-0.642 0.618 0.284-1.001ρ– for SWIFT* 0.186 0.080-0.361 0.192 0.034-0.559ρ– for GWIF* 0.243 0.123-0.421 0.189 0.034-0.546ρ– for DWIF* 0.320 0.175-0.518 0.489 0.203-0.862γ for ESS 0.468 0.282-0.626 0.309 -0.001-0.612γ for SWIFT* 0.710 0.539-0.819 0.701 0.360-0.861γ for GWIF* 0.672 0.498-0.797 0.716 0.377-0.876γ for DWIF* 0.575 0.393-0.717 0.436 0.111-0.712

*Value superior to value for ESS in both age groups. †AUC significantly higher than AUC for ESS in young adults. ρ+ and ρ– are likelihood ratios for positive and negative test. γ, Youden’s index. Mean + 1 SD used as cut-offs for calculation of sensitivity, specificity, ρ+, ρ– and γ. Young adults: ESS > 10, SWIFT > 12, GWIF > 11, DWIF > 1. Middle-aged to older adults: ESS > 9, SWIFT > 9, GWIF > 8, DWIF > 1.

ROC CurveNormal subjects vs. pts evaluated for narcolepsy (age > 45 yrs)


ESSSWIFTGWIFDWIF

1 - Specificity

Sens

itivit

y

0.0 0.2 0.4 0.6 0.8 1.0

1.0

0.8

0.6

0.4

0.2

0.0

Figure 4—ROC curves for ESS, SWIFT, GWIF, and DWIF for normal subjects vs. patients evaluated for narcolepsy in age group > 45 years


ROC CurveNormal subjects vs. pts evaluated for narcolepsy (18-45 yrs)


ESSSWIFTGWIFDWIF

1 - Specificity

Sens

itivit

y

0.0 0.2 0.4 0.6 0.8 1.0

1.0

0.8

0.6

0.4

0.2

0.0

Figure 3—ROC curves for ESS, SWIFT, GWIF, and DWIF for normal subjects vs. patients evaluated for narcolepsy in age group 18-45 years



RB SangalThe twelve test items of the SWIFT load on to two differ-

ent factors. Factor 1 seems to be a measure of general wake-fulness inability and fatigue, while Factor 2 seems to measure driving wakefulness inability and fatigue, indicating that it may be possible to measure separately general and driving related concepts/symptoms.

The ability of SWIFT, GWIF, and DWIF to discriminate be-tween normal subjects and patients with OSA, as well as pa-tients presenting with symptoms suggesting narcolepsy, shows that the SWIFT is a valid test, as does the ability to show sig-nificant improvement with CPAP treatment of OSA.

The SWIFT and GWIF are superior to the ESS (the criterion standard) in discriminating between normal subjects and pa-tients with OSA in both age groups, with regard to sensitivity/specificity/discriminant validity, as shown by AUC (statistical-ly significantly so for middle-aged and older adults), Youden’s index, as well as the positive and negative likelihood ratios. The SWIFT and GWIF are also superior to the ESS in discriminating between normal subjects and patients evaluated for narcolepsy in both age groups, with regard to sensitivity/specificity/dis-criminant validity, as shown by AUC (statistically significantly for young adults), Youden’s index, and positive and negative likelihood ratios. All the ROCs for ESS (young and middle-aged to older adults, patients with OSA, and patients evaluated with narcolepsy) were entirely within the ROCs for SWIFT and GWIF. Given the rarity of narcolepsy, comparisons were made using patients evaluated for narcolepsy rather than patients di-agnosed with narcolepsy. However, patients with OSA were excluded from this group; patients not positive for narcolepsy were as sleepy, fatigued, and unable to maintain wakefulness as the patients with narcolepsy. Thus, SWIFT and GWIF may be more useful than ESS in terms of clinical utility in discriminat-ing between normal and sleep disordered subjects.

Effect sizes were similar for improvement in ESS, SWIFT, and GWIF (but lower for DWIF) with CPAP treatment in young adults. In middle-aged to older adults, effect sizes for SWIFT and GWIF were higher than those for ESS and DWIF. Compar-isons of the number of patients with high ESS, SWIFT, GWIF, and DWIF before and after CPAP treatment revealed signifi-cant differences in both age groups. Improvement in SWIFT, GWIF, and DWIF, but not ESS, was significantly correlated with compliance. This compliance-response relationship lends more confidence in the use of the SWIFT or GWIF rather than the ESS is assessing treatment response with CPAP despite similar effect sizes for SWIFT, GWIF, and ESS. The finding that only ESS, but not SWIFT, GWIF, or DWIF, is correlated with MSLT suggests that subjects were able to separate the con-cept of tendency to fall asleep (as measured by the MSLT and the ESS) from wakefulness inability and fatigue. The finding that SWIFT, GWIF, and DWIF, but not ESS are correlated with arousal index suggests that they are a better measure of lack of sleep quality than the ESS. The finding that wakefulness in-ability and fatigue did not load on to separate factors on factor analysis suggests that subjects may have a hard time separating these two concepts.

The separation of data into two groups by age necessitated by a correlation between age and SWIFT as well as ESS in the combined group, provides a built-in replication, and similar findings in the two independent age groups (though more robust

in the middle-aged to older adults in the case of OSA and in young adults in the case of patients evaluated for narcolepsy) lend increased confidence to the results.

Mills et al. have reported that predictors of fatigue in OSA in-clude BMI, depression scores, and soluble tumor necrosis factor receptor I (sTNF-RI), but not the severity of OSA as measured by AHI or mean oxygen saturation.32 Tumor necrosis factor-α (TNF- α) and interleukin-6 (IL-6) are increased in OSA and narcolepsy.33 Adding measures of fatigue to the measurement scale for daytime functioning, and changing the measurement scale to measure wakefulness inability rather than tendency to fall asleep may improve the measurement of the daytime con-sequences of sleep disorders.

Masa et al.34 reported habitual sleepiness affecting 3.6% of drivers, with an odds ratio of 13.7 for highway automobile ac-cidents, and with considerable ESS overlap between these sub-jects and controls. 50% of habitually sleepy drivers had ESS < 9. This suggests that propensity to fall asleep in other cir-cumstances (as measured by ESS) is neither necessary nor suf-ficient to cause increased risk for auto accidents. Although a sample of 10,000 subjects with 71% response rate showed no correlation between the ESS and the adverse consequence of automobile accidents, there was a correlation with dozing off while stopped in traffic.18 Increased risk for auto accidents may be the result of a complex mix of wakefulness inability, fatigue, and inattention/cognitive impairment, all of which may occur in sleep disordered or sleep deprived subjects. Measurement of increased risk for auto accidents may require questions directly related to wakefulness inability and fatigue while driving, as in the DWIF factor of the SWIFT. The question whether DWIF might be predictive of risk for auto accidents needs to be eluci-dated in further research.

This study was designed to determine if the SWIFT is a reli-able and valid instrument, and if it is superior to the criterion standard, ESS in terms of specificity/sensitivity/discriminant ability, and therefore, possibly, clinical utility. We have shown that the SWIFT is reliable and has discriminant validity, that it has two factors (GWIF and DWIF), and that the SWIFT and its GWIF factor are superior to the ESS in discriminating between normal subjects and sleep disordered patients. These tests mea-sure sleepiness/wake inability and assist in screening/diagnosis. However, they are not meant to discriminate between different causes of difficulties with wakefulness inability or sleepiness. SWIFT should be added to ESS in evaluating daytime conse-quences of sleep disorders. The two tests together comprise 20 questions and can form a quick questionnaire for use in the of-fice to screen for sleepiness, wakefulness inability, and fatigue, with cutoffs of > 10 for ESS, > 12 for SWIFT, > 11 for GWIF, > 1 for DWIF in young adults (ages 18-45 years), and with cut-offs of > 9 for ESS, > 9 for SWIFT, > 8 for GWIF, > 1 for DWIF in middle-aged to older adults (age > 45 years).

A limitation of this study is that item selection was based on face validity rather than qualitative evaluation using patient focus groups. Another limitation is that the control group was recruited by means of a group e-mail to medical students and by personal solicitation in public places, and it is not clear whether this cohort of normal subjects generalizes to the population and whether it is comparable to the patient groups presented. Further, although the SWIFT and its factors are a better mea-


Sleepiness-Wakefulness Inability and Fatigue Testsure for differentiating between normal subjects and sleep dis-ordered patients than the ESS, the areas under the curve still leave a lot to be desired. However, though there may eventu-ally be a simple blood test to measure sleepiness, wakefulness inability, and fatigue, for now we are left with questionnaires as possibly the best proxies, though objective measures such as the psychomotor vigilance test or the divided attention driv-ing test are other candidates.35 This study was a clinical rather than an experimental study. Since the MWT is not routinely performed clinically, this study did not compare the SWIFT with the MWT. Future directions might include a study of the SWIFT using the MWT.

REFERENCES1. 2002 “Sleep in America” Poll. National Sleep Foundation website. http://www.

sleepfoundation.org/sites/default/files/2002SleepInAmericaPoll.pdf. Published 2002. Accessed February 11, 2011.

2. Drowsy Driving and Automobile Crashes. National Highway Safety Administra-tion website. http://www.nhtsa.gov/people/injury/drowsy_driving1/drowsy.html. I: Introduction. Accessed May 22, 2012.

3. Chervin RD. Sleepiness, fatigue, tiredness, and lack of energy in obstructive sleep apnea. Chest 2000;118:372-9.

4. Carskadon MA, Dement WC, Mitler MM, Roth T, Westbrook PR, Keenan S. Guidelines for the multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep 1986;9:519-24.

5. Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540-5.

6. Johns MW. Reliability and factor analysis of the Epworth Sleepiness Scale. Sleep 1992;15:376-81.

7. Arand D, Bonnet M, Hurwitz T, Mitler M, Rosa R, Sangal RB. The clinical use of the MSLT and MWT. Sleep 2005;28:123-44.

8. Mitler MM, Gujavarty KS, Browman CP. Maintenance of wakefulness test: a polysomnographic technique for evaluating treatment in patients with excessive sleepiness. Electroencpahlogr Clin Neurophysiol 1982;53:658-61.

9. Doghramji K, Mitler MM, Sangal RB, et al. A normative study of the mainte-nance of wakefulness test (MWT). Electroencpahlogr Clin Neurophysiol 1997;103:554-62.

10. Sangal RB, Thomas L, Mitler MM. The maintenance of wakefulness test (MWT) and the multiple sleep latency test (MSLT) measure different abilities in patients with sleep disorders. Chest 1992;101:898-902.

11. Browman CP, Gujavarty KS, Sampson MG, Mitler MM. REM sleep episodes dur-ing the maintenance of wakefulness test in patients with sleep apnea syndrome and patients with narcolepsy. Sleep 1983;6:23-8.

12. Sangal RB, Thomas L, Mitler MM. Disorders of excessive sleepiness: treat-ment improves ability to stay awake but does not reduce sleepiness. Chest 1992;102:699-703.

13. Sanford SD, Lichstein KL, Durrence HH, Riedel BW, Taylor DJ, Bush AJ. The influence of age, gender, ethnicity, and insomnia or Epworth Sleepiness Scores: a normative U.S. population. Sleep Med 2006;7:319-26.

14. Sangal RB, Sangal JM, Belisle C. Subjective and objective indices of sleepiness (ESS and MWT) are not equally useful in patients with sleep apnea. Clin Elec-troencephalogr 1999;30:73-5.

15. Sangal RB, Mitler MM, Sangal JM. Subjective sleepiness ratings (Epworth Sleepiness Scale) do not reflect the same parameter of sleepiness as objective sleepiness (maintenance of wakefulness test) in patients with narcolepsy. Clin Neurophysiol 1999;110:2131-5.

16. Johns MW. Sleepiness in different situations measured by the Epworth Sleepi-ness Scale. Sleep 1994;17:703-10.

17. Benbadis SR, Mascha E, Perry MC, et al. Association between the Epworth sleepiness scale and the multiple sleep latency test in a clinical population. Ann Intern Med 1999;130:289-92.

18. Gander PH, Marshall NS, Harris RB, Reid P. Sleep, sleepiness and motor ve-hicle accidents: a national survey. Austr N Z J Public Health 2005;29:16-21.

19. Merriam-Webster Dictionary. http://www.merriam-webster.com. Accessed Janu-ary 20, 2012.

20. Sangal RB. When is sleepiness a disease? How do we measure it? Sleep Med 2006;7:310-1.

21. Stein KD, Martin SC, Hann DM, Jacobsen PB. A multidimensional measure of fatigue for use with cancer patients. Cancer Pract 1998;6:143-52.

22. Stein, KD, Jacobsen PB, Blanchard CM, Thors CT. Further validation of the Multidimensional Fatigue Symptom Inventory-Short Form (MFSI-SF). J Pain Symptom Manage 2004;27:14-23.

23. Curran-Everett D. Multiple comparisons: philosophies and illustrations. J Physiol Regul Integr Comp Physiol 2000;279:R1-R8.

24. Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. American Academy of Sleep Medicine, Westchester, IL, 2007.

25. Bewick V, Cheek L, Ball J. Statistics review 13: Receiver operating characteristic curves. Crit Care 2004;8:508-12.

26. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the area under two or more correlated receiver operating characteristic curves: a nonparametric ap-proach. Biometrics 1988;44:837-45.

27. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operat-ing characteristic (ROC) curve. Radiology 1982;143:29-36.

28. Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley & Sons, 1981:sec 5.6.

29. Hilden J, Glasziou P. Regret graphs, diagnostic uncertainty and Youden’s Index. Stat Med 1996;15:969-86.

30. Biggerstaff BJ. Comparing diagnostic tests: a simple graphic using likelihood ratios. Stat Med 2000;19:649-63.

31. Aldrich MS, Chervin RD, Malow BA. Value of the multiple sleep latency test (MSLT) for the diagnosis of narcolepsy. Sleep 1997;20:620-9.

32. Vgontzas AN, Papanicalou DA, Bixler EO, Kales A, Tysom K, Chrousos GP. Elevation of plasma cytokines in disorders of excessive daytime sleepiness: role of sleep disturbance and obesity. J Clin Endocrinol Metab 1997;82:1313-6

33. Mills PJ, Kim J-H, Bardwell W, Hong S, Dimsdale JE. Predictors of fatigue in obstructive sleep apnea. Sleep Breath 2008;12:397-9.

34. Masa JF, Rubio M, Findley LJ. Habitually sleepy drivers have a higher frequency of automobile crashes associated with respiratory disorders during sleep. Am J Respir Crit Care Med 2000;162:1407-12.

35. Sunwoo BY, Jackson N, Maislin G, Gurubhagavatula I, Goerge CF, Pack AI. Reliability of a single objective measure in assessing sleepiness. Sleep 2012;35:149-58.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication October, 2011Submitted in final revised form May, 2012Accepted for publication May, 2012Address correspondence to: R. Bart Sangal, M.D., Professor, Oakland University William Beaumont School of Medicine, Director, Sleep Disorders Institute, 44344 Dequindre Rd. Ste 360, Sterling Heights, MI 48314; Tel: (586) 254-0707; Fax: (586) 254-7207; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. The author has indicated no financial



Respiratory stridor in patients with multiple system atrophy is a complication that occasionally causes nocturnal sudden death. Continuous positive airway pressure (CPAP) therapy has been proposed as an alternative to tracheostomy to treat nocturnal stridor associated with multiple system atrophy. However, some patients cannot tolerate CPAP therapy and experience sleep dis-turbances, even if the pressure is controlled; also, CPAP therapy can be less effective in patients with a narrow glottic opening

during sleep. This report describes the effect of laser arytenoid-ectomy on respiratory stridor caused by multiple system atrophy.Keywords: CPAP, endoscopic surgery, laser arytenoidectomy, multiple system atrophy, respiratory stridorCitation: Chitose S; Kikuchi A; Ikezono K; Umeno H; Na-kashima T. Effect of laser arytenoidectomy on respiratory stridor caused by multiple system atrophy. J Clin Sleep Med 2012;8(6):713-715.


CA

SE

RE

PO

RTS

Respiratory stridor in patients with multiple system atrophy is a complication that occasionally causes nocturnal sudden

death.1 Continuous positive airway pressure (CPAP) therapy has been proposed as an alternative to tracheostomy to treat noctur-nal stridor associated with multiple system atrophy.2,3 However, some patients cannot tolerate CPAP therapy and experience sleep disturbances, even if the pressure is controlled; also, CPAP thera-py can be less effective in patients with a narrow glottic opening during sleep. This report describes the effect of laser arytenoidec-tomy on respiratory stridor caused by multiple system atrophy.

REPORT OF CASE

A 55-year-old woman presented with a 2-year history of snoring and sleep apnea. She also had a 6-month history of oc-casional stridor during wakefulness and other symptoms, such as vesicorectal failure. Her voice quality was normal and scored as G0 using the grade-roughness-breathiness-asthenicity-strain system. Laryngeal fi berscopy revealed that vocal fold abduction was restricted on inspiration during wakefulness, but functioned normally on phonation. In contrast, during sleep induced by in-travenous injection of 5 mg diazepam, we observed a very nar-row glottic space on inspiration due to vocal fold adduction, but vocal fold abduction functioned normally on expiration, indi-cating the presence of paradoxical movement (Figures 1A, B).4

There was no fl oppy epiglottis5 or airway obstruction at the ary-tenoids. Brain and neck magnetic resonance imaging revealed no abnormal fi ndings. She was referred to the neurology clinic at the hospital and diagnosed with Shy-Drager Syndrome.

CPAP was applied with AutoSet titration. The 95th percen-tile pressure (the pressure at or below which the patient spent 95% of the time that night) was 10 cm H2O, and the median pressure was 9.8 cm H2O. This improved the SpO2 and reduc ed the volume of stridor. Polysomnography showed improvement in the apnea index, apnea-hypopnea index, lowest SpO2, and

Effect of Laser Arytenoidectomy on Respiratory Stridor Caused by Multiple System Atrophy

Shun-ichi Chitose, M.D.; Atsushi Kikuchi, M.D.; Keiko Ikezono, M.D.; Hirohito Umeno, M.D.; Tadashi Nakashima, M.D.Department of Otolaryngology-Head and Neck Surgery, Kurume University School of Medicine, Kurume, Japan

duration of SpO2 < 90% (table 1). However, her arousal index increased, and sleep architecture and daytime sleepiness wors-ened. Therefore, CPAP therapy was abandoned.

She underwent laser arytenoidectomy and tracheostomy (Figure 2). The posterior glottis was exposed under a micro-scope, and an incision was made along the superior portion of the right arytenoid cartilage from the tip of the vocal process to the apex, using a CO2 laser at 6-10 W continuous mode. The vo-cal process of the arytenoid cartilage was vaporized anteriorly to the posterior macula fl ava of the vocal fold. The arytenoid carti-lage was vaporized laterally near the muscular process. The pos-terior portion of the thyroarytenoid muscle was then vaporized. The wound was covered with preserved mucosa of the posterior glottis using fi brin glue. The laryngeal wound stabilized 7 weeks postoperatively and the tracheal stoma was closed.

The procedure widened the posterior glottal airway by re-moving the body and vocal process of the arytenoid cartilage, but left the muscular process of the arytenoid cartilage intact. The widened posterior glottis was recorded during diazepam-induced sleep (Figure 3). Her voice quality was G0 after clo-sure of the stoma. Polysomnography 7 weeks postoperatively showed that the procedure improved not only the apnea and apnea-hypopnea indices, but also the arousal index both with and without CPAP (table 1). Furthermore, the percentage of total sleep time spent in stage 2 sleep increased. Surgery alone was suffi cient for her to sleep without CPAP, and her daytime sleepiness improved. However, snoring caused by glossoptosis gradually appeared 7 years postoperatively. CPAP with a titrat-ed 95th percentile pressure of 10 cm H2O controlled her sleep disordered breathing without tracheostomy.

DISCUSSION

Respiratory stridor occurs in up to 30% of patients with multiple system atrophy. It traditionally has a poor prog nosis


S Chitose, A Kikuchi, K Ikezono et al

and is managed by tracheostomy. Isozaki et al. found that the stridor is caused by persistent hyperactivity of the intrinsic la-ryngeal muscle and neurogenic atrophy of the posterior crico-arytenoid muscle. There are two patterns of inspiratory vocal cord position during sleep: one in which the posterior glottis is still open and another in which it is almost completely closed over the total length of the cords. Tracheostomy should be considered in the latter pattern.4 Permanent tracheostomy is a logical and effective treatment for removing airflow obstruc-tion. However, it can limit daytime activity in patients with other disorders, as in patients with multiple system atrophy. Accordingly, we felt it worthwhile to prevent airflow obstruc-tion by widening the glottis, a method that does not require permanent tracheostomy.

Vocal cord lateralization6 and laser arytenoidectomy7,8 can successfully treat nocturnal stridor. These procedures are mini-mally invasive and are generally used for bilateral median vocal fold fixation in bilateral vocal fold paralysis. These procedures

restrict the movement of a movable vocal fold.6 One unsuc-cessful case of vocal cord lateralization has been reported.8 The thread used in lateralization may break following this procedure due to the glottic closing force. Widening of the posterior glottis by laser arytenoidectomy is more reliable and physiologically appropriate because the posterior glottis is respiratory,7 while the anterior glottis is phonatory. Laser arytenoidectomy can restore an adequate airway in patients with respiratory distur-bance due to multiple system atrophy without disrupting voice quality. Laser arytenoidectomy to widen the posterior glottis is highly recommended in cases of sleep disordered breathing that

Figure 1—Fiberoptic findings during sleep on expiration (A) and inspiration (B)

Figure 2—Intraoperative photograph

The body and vocal process of the right arytenoid cartilage were vaporized submucosally with the muscular process of the arytenoid cartilage left intact. The wound was covered with preserved mucosa from the posterior glottis using fibrin glue (white arrow). PM, preserved mucosa, PA, posterior airway, R, right vocal fold, L, left vocal fold.

Figure 3—Fiberoptic findings 7 weeks postoperatively

The widened posterior glottis after arytenoidectomy was recorded on inspiration during sleep. PA, posterior airway.

Table 1—Polysomnographic results before and after Laser arytenoidectomy

Before surgery After surgeryNo

treatment CPAPSurgery

alone CPAPTST (m) 392 296.5 437 356SE (%) 65.2 63.7 66.2 58.7%Stage1 28.6 22.4 24.3 15.7%Stage2 35.1 38 49.9 48.3%Stage3+4 1.6 3.2 0.3 4.6%REM 7.6 5.7 6.7 10.5Ar-I 47 49.6 27.3 24.7AI 18.2 7.9 0 0AHI 50.1 27.5 3.8 0.5Lowest SpO2 (%) 87 94 94 96SpO2 < 90% (%) 1.8 0.8 0 0

CPAP, continuous positive airway pressure; TST, total sleep time; SE, sleep efficiency; %Stage1, percentage of stage 1 sleep time in TST; %Stage2, percentage of stage 2 sleep time in TST; %Stage3+4, percentage of stage 3 and stage 4 sleep time in TST; Ar-I, arousal index; AI, apnea index; AHI, apnea hypopnea index; SpO2 < 90%, duration of SpO2 less than 90%.


Case Reportscannot be controlled by CPAP or those with respiratory stridor during wakefulness.

CPAP is a safe, noninvasive, and effective long-term therapy for nocturnal stridor and allows avoiding tracheostomy in some cases.2,3 A small number of patients with stridor cannot tolerate CPAP, and the pressure of CPAP on the closed glottis may cause the observed difficulties. The widening procedure may promote tolerance to CPAP therapy, even if sleep disordered breathing arises due to other conditions. However, the disappearance of sleep disordered breathing does not treat multiple system atro-phy. Severe depletion of neurons in the ventral medullary surface can occur, leading to dysfunction of the pontomedullary network that plays a critical role in the respiratory control of automatic breathing. Even after elimination of stridor, patients with mul-tiple system atrophy can die from respiratory arrest of central ori-gin, cerebellar ataxia, autonomic dysfunction, or parkinsonism.

CONCLUSION

Laser arytenoidectomy for respiratory stridor caused by mul-tiple system atrophy is recommended as an effective treatment for patients who cannot tolerate CPAP.

REFERENCES1. Silber MH, Levine S. Stridor and death in multiple system atrophy. Mov Disord

2000;15:699-704.2. Iranzo A, Santamaria J, Tolosa E. Continuous positive air pres sure eliminates

nocturnal stridor in multiple system atrophy. Barcelona Multiple System Atrophy Study Group. Lancet 2000;356:1329-30.

3. Iranzo A, Santamaria J, Tolosa E, Vilaseca I, Valldeoriola F, Martí MJ. Long-term effect of CPAP in the treatment of nocturnal stridor in multiple system atrophy. Neurology 2004;63:930-2.

4. Isozaki E, Hayashi M, Hayashida T. Vocal cord abductor paralysis in multiple system atrophy—Paradoxical movement of vocal cords during sleep. Clin Neu-rol 1996;36:529-33.

5. Shimohata T, Shinoda H, Nakayama H, Ozawa T, Terajima K, Yoshizawa H. Day-time hypoxemia, sleep-disordered breathing, and laryngopharyngeal findings in multiple system atrophy. Arch Neurol 2007;64:856-61.

6. Kenyon GS, Apps MC, Traub M. Stridor and obstructive sleep apnea in Shy-Drager syndrome treated by laryngofissure and cord lateralization. Laryngo-scope 1984;94:1106-8.

7. Sato K, Umeno H, Nakashima T. Laser arytenoidectomy for bilateral median vocal fold fixation. Laryngoscope 2001;111:168-71.

8. Umeno H, Ueda Y, Mori K, Chijiwa K, Nakashima T, Kotby NM. Management of impaired vocal fold movement during sleep in a patient with Shy-Drager syn-drome. Am J Otolaryngol 2000;21:344-8.

9. Hirano M, Kurita S, Kiyokawa K, Sato K. Posterior glottis. Morphological study in excised human larynges. Ann Otol Rhinol Laryngol 1986;95:576-81.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication July, 2011Submitted in final revised form April, 2012Accepted for publication May, 2012Address correspondence to: Shun-ichi Chitose, Department of Otolaryngology-Head and Neck Surgery, Kurume University School of Medicine, 67 Asahi-machi, Kurume 830-0011, Japan; Tel: +81-942-31-7575; Fax: +81-942-37-1200; E-mail: [email protected]




This is the case report of a 32-year-old obese male with a history of agitation, hallucinations, and delirium, recently diagnosed with primary hypothyroidism; he gave a several month history of fatigue with nocturnal snoring and frequent awakening. Polysomnogram revealed severe OSA; initia-tion of CPAP and levothyroxine resulted in immediate im-provement. The lack of a previous psychiatric history and acuteness of presentation was consistent with hypothyroid psychosis complicated by sleep deprivation cause by un-treated OSA. Primary hypothyroidism is a common disorder often associated with depression. It is rarely associated with

psychosis and was fi rst described as “myxoedematous mad-ness” in 1949. It has not been previously reported to cause psychosis when associated with obstructive sleep apnea. This case illustrates the need for examination of potential multiple organic causes in a patient who presents with psy-chosis in the critical care setting.Keywords: Hypothyroidism, myxedema madness, psychosis, sleep apneaCitation: Neal JM; Yuhico RJO. “Myxedema madness” asso-ciated with newly diagnosed hypothyroidism and obstructive sleep apnea. J Clin Sleep Med 2012;8(6):717-718.


CA

SE

RE

PO

RTS

Hypothyroidism is uncommonly associated with psycho-sis, and initially described as “myxoedematous madness”

in 1949 by Asher.1 It has not been previously associated with another common medical condition also resulting in cognitive dysfunction—obstructive sleep apnea (OSA).

We report the case of a 32-year-old obese male, diagnosed with primary hypothyroidism (TSH 92.0 mIU/mL, normal 0.40-4.65 mIU/mL) admitted to the critical care unit with a sev-eral day history of agitation, hallucinations, and delirium.

REPORT OF CASE

A 32-year-old Caucasian male was brought to the emergency department with a several day history of headache, fatigue, and hallucinations (being attacked by bats). He had not slept in 6 days, and family members related a history of poor sleep, fre-quent awakenings, and snoring for the last few months. He had been found to have primary hypothyroidism (TSH 92.5 mIU/mL, normal 0.42-4.82 mIU/mL) several days earlier and had not yet begun replacement therapy. He had a history of non-Hodgkin lymphoma, treated 4 years earlier with chemotherapy and mantle radiation, in remission; prior thyroid function stud-ies were not available. Medications included cetirizine, fexof-enadine, and alprazolam; the last medication had been started due to diffi culty sleeping.

Physical examination revealed an obese Caucasian male (BMI 35.0 kg/m2). Vital signs were remarkable for tachycar-dia (heart rate 106). Examination was unremarkable except for obesity, large neck circumference (19 inches), and agitated, belligerent behavior. Laboratory studies demonstrated elevated TSH (98.7 mIU/mL, normal 0.40-4.65) and creatine kinase (> 14,000 U/L, normal 30-170 U/L). Serum toxicology screen revealed no alcohol or illicit substances.

“Myxedema Madness” Associated with Newly Diagnosed Hypothyroidism and Obstructive Sleep Apnea

J. Matthew Neal, M.D., M.B.A.; Rodney Joe O. Yuhico, M.D.Department of Medicine, Indiana University Health Ball Memorial Hospital, Muncie IN

He was admitted to the critical care unit; because of severe agitation, he received intravenous sedation. Daily levothyrox-ine, 150 µg, was initiated. Because of possible rhabdomyolysis (presumably due to agitation and muscle trauma), he received intravenous hydration with sodium bicarbonate.

After 72 h, his symptoms had improved markedly, and serum creatine kinase returned to normal after several days. Given his obesity, poor sleep quality, and snoring, a polysomnogram was ordered at that time; this was performed and revealed an apnea-hypopnea index (AHI) of 61.5 (normal: < 5) and respiratory distress index (RDI) of 96.5 (normal: < 5), indicating severe obstructive sleep apnea. Continuous positive airway pressure (CPAP) was initiated.

He was discharged in stable condition 2 days later and con-tinued to improve as an outpatient after continuing CPAP and thyroid replacement therapy, and has had no recurrent neuro-psychiatric events.

DISCUSSION

This patient’s clinical presentation was consistent with hypo-thyroid psychosis (“myxedema madness”), likely exacerbated by coexistent severe OSA. The psychological disturbances associated with hypothyroidism were recognized in early lit-erature; hypothyroidism was fi rst reported as being associated with psychosis in 1888, when the Committee on Myxedema of the Clinical Society of London fi rst postulated a link between the two.2 The term “myxedema madness” was later coined by Asher in 1949,1 and this entity has been increasingly realized as an uncommon potential cause of psychosis.

While typically associated with lassitude and depression, as many as 15% of patients can exhibit psychosis with hypothy-roidism.3 The clinical presentation of psychosis is not uniform,


JM Neal and RJO Yuhicoand no specific group of findings is typical.4 Hallucinations have frequently been reported. While most studies show a slow reversal of psychosis (within weeks or months), rapid improve-ment (within one week) has rarely been described.2

Bahammam et al. reported the incidence of clinical hypo-thyroidism in patients with newly diagnosed OSA to be only 0.4%, although the incidence of subclinical hypothyroidism was much higher (11.1%).5 Sleep disorders such as obstruc-tive sleep apnea may be associated with psychosis as well as significant cognitive and behavioral dysfunction, although the most common psychiatric disturbance associated with OSA is depression.

OSA symptoms in hypothyroid patients typically abate after replacement therapy. Rajagopal et al. noted that apneic events in hypothyroid OSA patients decreased from 71.8 to 12.7 per hour after three to twelve months of thyroid replacement.6

Several cognitive and psychiatric alterations associated with OSA have also been described. Lee et al. identified a young person presenting with recurrent psychosis refractory to an an-tipsychotic.7 He was later diagnosed with OSA by polysomnog-raphy and underwent tonsillectomy with subsequent remission of psychotic attacks.

The clinician must be aware of the variability in psychiat-ric symptoms of patients with hypothyroidism and OSA. The constellation of neuropsychiatric symptoms is variable, and no specific set of findings can exclude the disorder; this should be done by biochemical evaluation. Since hypothyroidism and obstructive sleep apnea are both common medical conditions, both should be considered in any patient presenting with such

symptoms. It demonstrates the importance of investigating or-ganic causes of psychosis and other causes (e.g., sleep apnea), as the initiation of OSA treatment likely hastened the recovery.

REFERENCES1. Asher R. Myxoedematous madness. Br Med J 1949;2:555-62.2. Heinrich TW, Grahm G. Hypothyroidism presenting as psychosis: myxedema

madness revisited. Prim Care Companion J Clin Psychiatry 2003;5:260-6.3. Lehrmann JA, Jain S. Myxedema psychosis with grade II hypothyroidism. Gen

Hosp Psychiatry 2002;24:275-7.4. Hall RC. Psychiatric effects of thyroid hormone disturbance. Psychosomatics

1983;24:7-11,15-18.5. Bahammam SA, Sharif MM, Jammah AA, Bahammam AS. Prevalence of thyroid

disease in patients with obstructive sleep apnea. Respir Med 2011;105:1755-60.6. Rajagopal KR, Abbrecht PH, Derderian SS, et al. Obstructive sleep apnea in

hypothyroidism. Ann Intern Med 1984;101:491-4.7. Lee S, Chiu HF, Chen CN. Psychosis in sleep apnoea. Australian N Z J Psychia-

try 1989;23:571-3.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication February, 2012Submitted in final revised form March, 2012Accepted for publication May, 2012Address correspondence to: J. Matthew Neal, M.D., Department of Medicine, Indiana University Health Ball Memorial Hospital, 2401 University Ave., Muncie, IN 47303; Tel: (765) 747-4350; Fax: (765) 751-1451; E-mail: [email protected]




Chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) is a chronic, disabling illness that affects approximately 0.2% of the population. Non-restorative sleep despite suffi -cient or extended total sleep time is one of the major clini-cal diagnostic criteria; however, the underlying cause of this symptom is unknown. This review aims to provide a compre-hensive overview of the literature examining sleep in CFS/ME and the issues surrounding the current research fi ndings. Polysomnographic and other objective measures of sleep have observed few differences in sleep parameters between CFS/ME patients and healthy controls, although some dis-crepancies do exist. This lack of signifi cant objective differ-ences contrasts with the common subjective complaints of disturbed and unrefreshed sleep by CFS/ME patients. The emergence of new, more sensitive techniques that examine

the microstructure of sleep are showing promise for detecting differences in sleep between patients and healthy individu-als. There is preliminary evidence that alterations in sleep stage transitions and sleep instability, and other physiological mechanisms, such as heart rate variability and altered corti-sol profi les, may be evident.Future research investigating the etiology of non-restorative sleep in CFS/ME may also help us to undercover the causes of non-restorative sleep and fatigue in other medical conditions.Keywords: Chronic fatigue syndrome, myalgic encephalomy-elitis, sleep, non-restorative sleep, sleep disorders, fatigue, sleepinessCitation: Jackson ML; Bruck D. Sleep abnormalities in chron-ic fatigue syndrome/myalgic encephalomyelitis: a review. J Clin Sleep Med 2012;8(6):719-728.


1.0 INTRODUCTION

Chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) is a medically unexplained disabling illness, with preva-lence estimates of between 0.007% and 2.8% of the general adult population.1-3 It is primarily characterized by persistent relapsing fatigue of at least 6 months in duration that reduces activity levels. Other key features of the disorder include post-exertion malaise of either physical or mental exertion, memory and concentration diffi culties, muscle pain, headaches, tender lymph nodes, sore throat, and non-restorative sleep. Of direct relevance to the current review, patients often report daytime sleepiness, feeling unrefreshed upon waking despite suffi cient or extended total sleep time, extended sleep including daytime napping, and other sleep-related symptoms, such as diffi culty falling asleep and disturbed sleep.2,4,5 Since the late 1980s, re-search in the area of CFS/ME has begun to focus on examining and determining the cause of non-restorative sleep and sleep disturbance in CFS/ME, and how these symptoms may relate or contribute to patients’ daytime fatigue. The current review aims to provide a comprehensive overview of these studies, describe where the fi eld currently stands on this issue, and outline poten-tial directions for future research.

2.0 HISTORICAL OVERVIEW OF DIAGNOSTIC CRITERIA

Clusters of symptoms including chronic fatigue, cogni-tive and mood impairments, sleep diffi culties and pain have

Sleep Abnormalities in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis: A Review

Melinda L. Jackson, Ph.D.; Dorothy Bruck, Ph.D.School of Social Sciences and Psychology, Victoria University, Victoria, Australia

been observed in clinical practice for centuries. A collection of symptoms similar to that of CFS/ME, including fatigue, depression, headache, impotence and neuralgia, were fi rst reported in the 19th century.6 The diagnosis “neurasthenia” was popularized by American neurologist George Beard to describe this collection of symptoms.7 With increasing so-phistication of psychiatric diagnoses over time, diagnosis of neurasthenia has declined, however, the symptoms described in these early patients closely match those we now recognize in CFS/ME today.

The name “chronic fatigue syndrome” was fi rst suggested in 1988, and most commonly used by the medical and scientifi c community to describe the illness. However, the name CFS has been rejected by many patients, patient advocacy groups and some doctors as it undermines and trivializes the illness. Since 1969, myalgic encephalomyelitis (ME) has been included in the World Health Organization’s International Classifi cation of Diseases (ICD-10), and although is clinically distinct from CFS, the two terms are used interchangeably. Due to the stigma of the name “chronic fatigue syndrome,” the illness is often re-ferred to as CFS/ME.

It is important to make a distinction between CFS/ME and fi bromyalgia (FM). FM is a chronic medical disorder character-ized by widespread pain, a heighten pain response, and sleep disturbance,8 and is commonly comorbid with CFS/ME.9 Given the overlapping symptoms and lack of a diagnostic test, it is important for research studies to have clear criteria for distin-guishing these two conditions. This review will focus on re-search that has specifi cally examined sleep in CFS/ME.


ML Jackson and D Bruck2.1 Research Definition

The first working case definition of chronic fatigue syn-drome was introduced in 1988 by the United States Centers for Disease Control and Prevention (CDC).10 The development of a case definition allowed for a systematic and comprehensive approach to defining the etiology and pathophysiology of CFS/ME. These definitions, along with the 198811 and 199012 Aus-tralian definitions, and the 1991 Oxford, UK definition,13 have played an essential role in orienting clinical research, and facili-tating consistency and homogeneity of samples across research studies. Although these initial criteria were considered quite restrictive, the diagnostic criteria did not specifically exclude a sleep disorder. As a result, early studies that examined the association between sleep complaints and functional disability in CFS/ME suggested that a primary sleep disorder may be the cause of unrefreshed sleep in some patients.

In 1994, a revision to the case definition and a set of research guidelines for use in studies of CFS patients was proposed by the Centers for Disease Control and Prevention.14 These guide-lines (known as the Fukuda criteria) used more relaxed defini-tions than the 1988 criteria, requiring only four criteria beyond fatigue for diagnosis, and not excluding non-psychotic psychi-atric disorders. Given that there is no laboratory test for diag-nosing CFS/ME, and the etiology is typically unknown, CFS/ME was seen as a diagnosis of exclusion. This was reflected in the new criteria by specifically excluding comorbid condi-tions such as a treatable sleep disorder (e.g., obstructive sleep apnea [OSA] and narcolepsy) and other potential causes of fa-tigue (e.g., substance abuse, psychiatric disorder). Unrefreshing sleep was the only criteria relating to sleep, with other sleep disturbances not a criterion. Although alternative definitions have been proposed,15 the 1994 Fukuda criteria are considered the international accepted research definition. However, it was recognized that the Fukuda criteria and other broadly inclu-sive criteria15 do not adequately discriminate CFS/ME patients from those with other conditions such as major depressive disorder. Additionally, since the Fukuda criteria were primar-ily developed to inform clinical research excluding comorbid conditions, such as treatable sleep disorders, they may not be appropriate to use exclusively for clinical diagnoses, which are often more broadly defining.

2.2 Clinical GuidelinesIn 2003, an Expert Medical Consensus Panel with extensive

experience in the research and clinical management of CFS/ME developed the Canadian Clinical Case Definition.16 This docu-ment was created specifically to inform healthcare profession-als. The Canadian Criteria has incorporated a larger spectrum of potential symptoms, aimed to assist recognition of the “interre-lationships of each patient’s symptoms and their coherence into a syndrome of related symptoms.”16 This updated clinical defi-nition captured, in addition to chronic fatigue, the issue of post-exertion malaise. These criteria also highlight mental fatigue (loss of cognitive function and alertness) as well as physical fatigue (lack of energy and strength). Sleep disturbance is rec-ognized as a major feature of CFS/ME in the Canadian Criteria. Specifically, sleep and circadian rhythms disturbances are list-ed, including early, middle, and late insomnia, and reversed or abnormal diurnal and sleep rhythms. Further, periodic limb and

restless legs syndrome are reported to accompany these other changes in sleep in many cases. These criteria also recognize the importance of ruling out a treatable sleep disorder, such as upper airway resistance syndrome and OSA.

Most recently, International Consensus Criteria (ICC) have been developed.17 The ICC build from the Canadian Criteria to identify the distinct characteristic patterns of symptom clusters of CFS/ME. These new definitions relaxed the requirement for symptoms to have persisted for at least 6 months, giving the physician more temporal control of when the diagnosis of CFS/ME can be made. Sleep disturbances were divided into two categories: disturbed sleep patterns, including insomnia, pro-longed sleep including naps, frequent awakenings, and vivid dreams/nightmares; and unrefreshed sleep, including excessive daytime sleepiness.17 Importantly, these updated criteria serve to not only diagnose patients in the clinical setting, but to also assist in identifying patients for research studies.

3.0 DIFFERENTIAL DIAGNOSES

The diagnostic overlap between CFS/ME and primary sleep disorders has been documented in a number of studies. Prior to the revision of the diagnostic criteria for CFS/ME in 1994, two studies reported that over half of their CFS/ME patients had a sleep disorder as assessed by overnight polysomnogra-phy (PSG).5,18 These sleep disorders include hypersomnia, sleep maintenance and sleep initiation insomnia, OSA, narcolepsy, and periodic limb movement disorder, as well as inadequate sleep hygiene. CFS/ME patients were found to have higher lev-els of fatigue and sleep disturbance than patients with multiple sclerosis5; those with a comorbid sleep disorder reported greater functional impairment.18 Since these studies included CFS/ME patients with a suspected sleep disorder who were attending the sleep clinic, rather than randomly selected or consecutive pa-tients, these figures are mostly likely inflated.

Subsequent to the 1994 guideline revisions, Fossey et al. (2004) observed prevalence rates > 50% of ICSD-classified pri-mary sleep disorders (OSA and movement disorders) in CFS/ME patients.19 Similarly, Le Bon and colleagues examined the prevalence of primary sleep disorders and objective sleepiness in CFS/ME and found that 46% of the 46 unselected patients who met Fukuda criteria for CFS/ME also presented with OSA (using a criterion of AHI > 5), and a further 5% presented with periodic limb movement disorder.20 On multiple sleep latency testing (MSLT), an objective measure of sleep propensity, 30% of the patients were classified as clinically sleepy. Objective (MSLT) and subjective (Stanford Sleepiness Scale) measures of sleepiness, however, were not associated with a subjective fa-tigue measure, suggesting that the fatigue experienced by CFS/ME patients is separate from the expression of sleepiness. Im-portantly, this study compared CFS/ME patients with and with-out a primary sleep disorder and found that they could not be separated on clinical presentation. The authors concluded that the symptoms of CFS/ME are clearly distinct from those of pri-mary sleep disorders, and the illness is more than simply a so-matic expression of an underlying sleep disorder or sleepiness.20

Larger population-based studies of CFS/ME patients have also provided some insight into the rates of sleep disorders in CFS/ME patients over time.15,21 Approximately 20% of CFS/


Review ArticleME patients in these studies were found to have either OSA or narcolepsy, with one study reporting that 20% of patients were given an alternative diagnosis of sleep disorder at 3-year fol-low-up.21 Subclinical levels of sleep disordered breathing have also been reported in some studies.15,22

From a clinical standpoint, early studies that examined prev-alence of comorbid sleep disorders provide new insights into the syndrome and potential differential diagnoses and highlight the importance of considering a potential sleep disorder as a cause of unexplained fatigue.23 While there is overwhelming evidence of the distinction between CFS/ME and sleep disor-ders,24 some researchers argue that a diagnosis of OSA should not be an exclusion criterion for CFS/ME. It is argued that pri-mary sleep disorders do not influence the core symptoms of CFS/ME and therefore should be considered a comorbid con-dition.25 Indeed, in clinical practice many physicians treat pri-mary sleep disorders concurrently with CFS/ME rather than exclude these patients. CPAP therapy in CFS/ME patients with comorbid OSA has been found to improve some daytime fea-tures, such as cognitive and daytime sleepiness; however, the underlying fatigue state remains.25 The magnitude of OSA in CFS/ME also depends on what diagnostic threshold is used for OSA. Le Bon and colleagues acknowledge that if an AHI cut-off of 20/h was used in their study (rather than AHI < 5), OSA would only be prevalent in 11% of their CFS/ME sample (as opposed to nearly 50%).20 These figures are comparable to the prevalence of sleep disordered breathing in the general popula-tion.26 Thus it could be argued that primary sleep disorders are a comorbid condition that occur at a similar frequency in CFS/ME to the general population and are therefore not reflective of the disorder itself.

In addition to generalized joint and muscle myalgia, which are integral features of the diagnostic criteria, CFS/ME is also associated with comorbid pain conditions including irritable bowel syndrome and migraine headaches. Pain experienced by CFS/ME patients may also play a critical role in sleep dis-turbance. Studies of FM have found that both subjective sleep quality measures and phasic alpha sleep are associated with pain sensitivity.27 There is a potential bidirectional relationship be-tween sleep and pain—pain disrupts sleep, and sleep disruption enhances pain. Firstly, pain causes disruption to sleeping pattern and increases sleep onset latency. Experimental manipulations of pain stimuli to the muscles during sleep have revealed de-creased delta and increased alpha activity of the sleep EEG, and impaired sleep quality.28,29 CFS/ME patients are found to have more self-reported awakenings during sleep due to pain com-pared to depressed patients and healthy controls.32 On the other hand, sleep disturbance also leads to reduced pain thresholds during waking.30 Pain and fatigue symptoms, similar to those reported in CFS/ME and FM, have been induced in healthy in-dividuals by disrupting SWS.30 Thus, it has been posited that the physiological arousals that are observed during sleep reflect a vigilant nocturnal state that contributes to daytime fatigue, pain, and hypersensitivity, and subjective feelings of non-restorative sleep. The influence of myalgic symptoms on sleep disturbance in CFS/ME has received little research attention to date.

Some patients with CFS/ME have comorbid psychiatric or somatoform illness, such as depressive disorder or fibromyal-gia, which do not rule out a diagnosis of CFS/ME. Given that

sleep disturbance is a common symptom of psychiatric condi-tions,31 it could be argued that sleep disturbance in CFS/ME is a secondary consequence of comorbid depression. Studies that have explored this potential link by comparing sleep study results of CFS/ME patients with and without a psychiatric co-morbidity suggest that sleep disturbances are common in both subtypes, and therefore do not appear to be solely the result of underlying depression.4,32 All of these studies mentioned above have been critical for highlighting CFS/ME as an autonomous syndrome.

4.0 MACROSTRUCTURE MEASURES OF SLEEP IN CFS/ME

table 1 presents studies that have compared sleep param-eters of CFS/ME patients to healthy controls using polysom-nography. Of 24 papers reviewed, only 15 used recognized diagnostic criteria for patient recruitment, and only these 15 are included in table 1. Of these 15, only 10 reported that they excluded patients who were on medication or asked patients to withdraw from their medication for 2 weeks prior to the study.

Studies examining traditional sleep parameters as measured by PSG have reported variable and nonspecific differences in sleep parameters between CFS/ME patients and controls. The reason for the discrepancy may be due in part to differences in selection criteria (e.g., medication status), the type of con-trol group used, or characteristics of the recorded night (e.g., first night of recording only, home recordings), which also make comparisons between studies difficult. First night effects are apparent in CFS/ME patients who do not have a primary sleep disorder,33 and therefore studies that have only used one night of recording may not provide an accurate picture of the patient’s typical sleep.

Reflecting the disturbed sleep reported by many CFS/ME patients, an increased number18,34,35 and duration18,34,36 of inter-mittent awakenings have been reported. However, using an ad-aptation night prior to the main sleep recording night, Reeves et al. found no difference in sleep efficiency between CFS/ME patients and controls.15 This study, however, enforced bed and wake times for both groups, and as such may not have captured true sleep efficiencies that the participants would typically ex-perience. When patients and controls have been allowed to go to bed at their usual bed time, reduced total sleep time (TST) and sleep efficiency are observed in the CFS/ME group, both in single-night PSG recording in the laboratory,22,37,38 and in some,18,35 but not all36 at-home studies. Togo et al. (2008) ex-amined this observation further by stratifying their patients into those who reported sleepiness upon waking, and those who re-ported more sleepiness in the evening. Using this distinction, they found that those who reported morning sleepiness had lower sleep efficiencies and more periods of interrupted sleep.38

Sleep onset latencies are often longer in CFS/ME patients compared to healthy control subjects, suggesting that some pa-tients may have difficulties initiating sleep.4,37 It has been report-ed that CFS/ME patients may have differential parasympathetic activity at sleep onset, which may contribute to delayed sleep latency.39 This finding could also reflect poor sleep hygiene or extended sleep periods and napping, which may reduce sleep drive at night.


ML Jackson and D Bruck

Table 1—Summary of literature investigating sleep, measured polysomnographically, in CFS/ME patients compared to healthy controls, using standard diagnostic criteria

Author, yearCriteria Used Subjects

Medication Status Outcome Measures Findings

Sharpley 199736

Fukuda 20 CFS/ME (psychiatric illness excluded), 20 controls

Psychotropic medication users excluded.

Subjective reports, sleep diary, 1 night home-based PSG

CFS/ME reported poor quality, unrefreshing sleep, and daytime napping. CFS/ME had ↑ TIB and time awake after sleep onset, and ↓ SE. No difference between CFS/ME and controls on PSG-measured TST. N = 7 CFS/ME patients had abnormal PSG (5 with sleep initiation and maintenance problems, 1 with early awakening).

Watson 200344,†

Fukuda 22 twin pairs discordant for CFS/ME

No medication for > 2 weeks

2 nights of PSG, SDQ (insomnia symptoms)

CFS/ME twin reported ↑ insomnia symptoms and poor sleep ratings. No difference in PSG measures of insomnia. ↑ % REM in CFS/ME twin. Suggests sleep state misperception

Ball 200445,† Fukuda 22 twin pairs discordant for CFS/ME


2 nights of PSG CFS/ME twin displayed ↑ %SWS and REM sleep. No other objective differences in sleep architecture in CFS/ME twin compared to healthy twin.

Reeves 200615,^

Fukuda 43 CFS, 43 controls

Medication continued*

2 nights PSG, MSLT No difference in PSG sleep architecture or MSLT

Majer 200746,^ Fukuda 35 CFS, 40 controls


2 nights PSG, MSLT, subjective sleep quality

CFS/ME report poorer sleep quality, no difference on MSLT or PSG. CFS/ME patients perceive poor sleep in the absence of objective sleep problems.

Neu 200747,# Fukuda 28 “pure” CFS, age and gender match controls


PSQI, fatigue severity scale, PSG

CFS/ME reported poorer sleep on PSQI. No difference in PSG sleep parameters (e.g., SWS). Sleep quality misperception may exist in CFS/ME

LeBon 200748,# Fukuda 28 “pure” CFS, 27 OSA, 27 healthy controls


2 nights PSG, distribution of NREM sleep

CFS/ME display ↑ NREM sleep, and ↑ ratios of SWS-to-light sleep than controls and OSA patients

Armitage 200749,†


No medication for > 2weeks

3 nights in lab: adaptation, baseline, sleep delay. Power spectral analysis

CFS/ME exhibited ↓ SWA power in first NREM period after delayed sleep; baseline SWA similar between CFS/ME and controls. CFS/ME blunted SWA response to sleep challenge

Kishi 200850 Fukuda 22 CFS, 22 control all female, no MDD same menstrual phase

Not reported Overnight PSG, duration and transition statistics

↓ relative frequency of REM to NREM transition in CFS/ME, causing significantly ↑ transitions from REM and S1 to awake. Normal continuation of sleep in S1 and REM is disrupted in CFS/ME

Togo 200838 Fukuda 12 CFS/ME+ FM, 14 CFS26 controls. No MDD

Medication use excluded

Overnight PSG ↑SWS in CFS/ME who reported feeling less sleepy upon waking. CFS/ME display ↓ sleep efficiency, TST, REM. Stratified group based on sleepiness (AM vs PM). AM sleepy group had more periods of interrupted sleep during night.

Armitage 200951,†



2 nights PSG, power spectral analysis

CFS/ME display no sleep micro or macro architectural changes, thus sleep measures cannot explain the fatigue.

Decker 200952 Fukuda 35 CFS, 40 controls


2 nights PSG, FFT, power spectral analysis

CFS/ME displayed ↓ delta power during SWS, but ↑ in S1 and REM compared to controls. Alpha, theta, sigma, beta ↓ during SWS, REM, S2

Neu 200924 (replication of Le Bon 2007)

Fukuda 32 CFS/ME, 30 OSA, 14 controls


2 nights PSG: NREM proportions and ratios

CFS/ME displayed ↑ NREM sleep, and ↑ ratios of SWS-to-light sleep compared to controls and OSA patients

Creti 201053 Reeves, 2005

47 CFS/ME Medication use NOT excluded

Overnight PSG, actigraphy, self-report

Sleep diary most related to PSG and actigraphy. TST and SE differentiated insomnia symptoms. EDS, fatigue, unrefreshed sleep not related to qualitative or quantitative methods

Kishi 201154 Fukuda 14 CFS/ME without FM, 12 CFS/ME with FM, 26 controls, all female

Medication use excluded

Overnight PSG, transition probabilities and rates between sleep stages, and duration in each stage

Probability of transitions from REM to waking significantly higher in CFS/ME than controls. Differences in transition probabilities between CFS/ME + FM and CFS/ME alone, suggesting that they are separate illnesses with distinct problems of sleep regulation.

#Same subjects. †Same subjects. ̂ Same subjects. *Sleep-affecting medications statistically controlled for in sleep analysis. EDS, excessive daytime sleepiness; FFT, fast Fourier transform; FM, fibromyalgia; MDD, Major Depressive Disorder; MSLT, multiple sleep latency test; OSA, obstructive sleep apnea; PSG, polysomnography; PSQI, Pittsburgh Sleep Quality Index; REM, rapid eye movement sleep; S1, stage 1 sleep; S2, stage 2 sleep; SDQ, sleep disorders questionnaire; SE, sleep efficiency; SL, sleep latency; SSS, Stanford Sleepiness Scale; SWA, slow wave activity; SWS, slow wave sleep; TIB, time in bed; TST, total sleep time.


Review ArticleActigraphy studies potentially provide a more ecologically

valid assessment of TST, sleep efficiency, and sleep onset la-tency parameters, as measurements are made in a naturalistic setting where patients can follow their usual sleep routines. A study in children with CFS/ME reported that continuous sleep of > 10 h, measured with actigraphy, was not uncommon.40 Im-paired daily sleep/wake rhythms and disturbed sleep were ob-served in those children who displayed an irregular sleep type. These studies can also examine whether changes in the circa-dian timing of sleep are evident, such as a delayed sleep phase. Most actigraphy studies to date have examined diurnal activity patterns in CFS/ME patients,41 and there are currently mixed findings relating to circadian rhythm disturbances.42,43

The sleep architecture of CFS/ME patients may differ from that of healthy individuals. Stage 3 sleep or slow wave sleep (SWS) is typically observed for approximately a fifth of the sleep period in young healthy individuals.55 A number of stud-ies have reported reduced time in SWS in CFS/ME patients relative to controls4,38 and between monozygotic twins discor-dant for CFS/ME.41 These effects are independent of depres-sion and FM.4,24

Alpha-delta sleep is an atypical electroencephalographic (EEG) pattern recorded during NREM sleep. In normal sleep, alpha activity is characteristic of drowsy wakefulness, and delta activity indicates restorative NREM sleep. Alpha-delta sleep was first observed during studies of sleep EEG of patients with psy-chiatric illness who presented with fatigue. The appearance of alpha-delta sleep, or alpha intrusions, during SWS has been re-ported in some early studies of CFS/ME patients.37,56 The appear-ance of alpha-delta sleep in these patients was initially thought to be the cause of non-restorative sleep. However, whether the patients in these studies had a “pure” diagnosis of CFS/ME or potentially some feature of FM is unclear. Later studies failed to find this phenomenon,51,57 and the role of alpha-delta sleep in the pathophysiology of CFS/ME has since been questioned.

The amount of SWS and slow wave activity (SWA; power density in the delta frequency) are determined by prior wakeful-ness. For example, SWS and SWA are found to increase follow-ing periods of sleep deprivation or restriction, when there is a build-up of homeostatic sleep pressure.58 As such, SWA during NREM sleep is often used as a marker of sleep homeostasis. A study by Armitage and colleagues exploited this phenomenon by exposing 13 monozygotic twins discordant for CFS/ME to a sleep delay schedule.49 After 2 baseline nights in the laboratory, a mild sleep challenge was imposed, involving a sleep delay of 4 hours, followed by a regular sleep length “recovery” period. Although no differences in SWA during baseline sleep were found, the CFS/ME twin expressed significantly less SWA in the first NREM period after the sleep delay. Additionally, the time course of dissipation of SWA across the night was altered in the CFS/ME twin. The authors concluded that CFS/ME is as-sociated with impaired sleep homeostasis and basic sleep drive. Interestingly, the cytokine systems are intimately involved with sleep regulation; increased SWA also occurs in response to acute infection, with proinflammatory cytokines increasing SWS.59 Alteration in SWA may therefore be associated with the systemic inflammation found in CFS/ME.60 Further research examining SWA and immune dysfunction in CFS/ME would be valuable for understanding this potential link.

The distribution and amount of REM sleep in CFS/ME is not as clearly defined as that of SWS. Reduced REM sleep in CFS/ME patients is reported in some studies,35,38 whereas oth-ers have observed a higher percentage of REM sleep in CFS/ME relative to controls.37,45,61 When statistically controlling for medication use, no difference in REM sleep latency was ob-served between controls and CFS/ME patients,15 suggesting that medications used by these patients may play a role in some sleep architecture differences previously reported. Perhaps the clearest findings have been derived from twin studies, as de-scribed earlier. In these studies, no differences in REM sleep are observed between the CFS/ME twin and the healthy twin.51

To date, approximately 273 clinically diagnosed CFS/ME patients have been assessed using PSG, with less than half of these patients studied not using medication. Based on these limited data, there appear to be very few differences in sleep architecture or TST between CFS/ME patients and healthy in-dividuals, with mixed findings for SWA and SWS. The distribu-tion of SWS and the frequency of sleep stage transitions appear to differ between CFS/ME patients, healthy controls, and other sleep disorders. Of the 4 twin studies published, increased SWS and REM sleep are typically reported in the CFS twin, with evi-dence for an impaired sleep homeostatic response, but no dif-ferences on power spectral analysis were observed. Studies that have had utilized co-twin control methodology have the added benefit of controlling for many genetic and environmental fac-tors that are typically not accounted for in either CFS/ME or sleep research, making for a more powerful and robust design.44

Based on these data, it appears that CFS/ME does not have a characteristic objective sleep disturbance found across all patients. As a result, some researchers have concluded that CFS/ME patients do not have abnormal sleep, and objective sleep measures do not account for subjective reports of non-restorative sleep. Studies using traditional PSG measures have been unable to shed light on a cause for the experience of non-restorative sleep.

5.0 DISCREPANCY BETWEEN OBJECTIVE AND SUBJECTIVE REPORTS OF SLEEPINESS

While CFS/ME patients present with fatigue as their prima-ry symptom, whether they also experience excessive daytime sleepiness is less clear cut. The consensus from a number of studies is that pathological sleepiness objectively measured us-ing multiple sleep latency tests (MSLT) is not observed in CFS/ME patients.15,46,62 This is despite CFS/ME patients reporting higher levels of subjective sleepiness and poorer sleep quality than healthy controls.24,53,61

There are a few possible explanations for this discrepancy between objective and subjective sleepiness measures. Firstly, it has been suggested by some authors that CFS/ME patients have sleep quality misperception.15,61 Interestingly, poor self-rated health and depressive symptoms have been found to be associated with over reporting of sleep difficulties and under-estimation of sleep efficiency.63 However, it is unlikely that this discord is truly a global phenomenon across all CFS/ME pa-tients, and perhaps reflects more of a generalized issue among all patients experiencing sleep disturbance, such as insomnia.64 It may be that sleep disturbed individuals more closely monitor


ML Jackson and D Brucktheir sleep habits and patterns and are more in tune to changes in their sleep.

A second possibility is that the discrepancy reflects issues with definitions and measurement tools used to determine fa-tigue and sleepiness. Outside of the sleep arena, fatigue and sleepiness are ill-defined concepts with overlap in their defi-nitions and are often used interchangeably. This can make it difficult for both patients and clinicians to correctly distinguish between the two states.65 Further, currently there is a lack of a clear and reliable subjective measure that differentiates the two states. One study examined whether current measures cor-rectly distinguish between sleepiness and fatigue, by comparing CFS/ME and OSA patients whose primary symptom is daytime sleepiness.66 In this study, a clear distinction between subjec-tive measures between the 2 patient groups was observed, with CFS/ME reporting the most fatigue and OSA patients reporting higher levels of sleepiness. However, there was some overlap in the levels of subjective sleepiness between the 2 groups, combined with the well-recognized discordance between ob-jective and subjective sleepiness in the CFS/ME group. This study highlights the need for more precise tools and analyses for distinguishing these 2 states.

Another explanation for differences in objective and subjec-tive measures of sleep in CFS/ME is that potential nocturnal neurophysiological disturbances that result in the non-restor-ative sensation following sleep in CFS/ME patients are not de-tected by traditional sleep variables or sleep stage distributions measures. More sensitive micro-analyses of the sleep EEG and other nocturnal parameters are currently being explored in this population.

6.0 MICROSTRUCTURE MEASURES OF SLEEP IN CFS/ME

While objective sleep parameters do not clearly predict subjective reports of sleep disturbance, other physiological measures may have more promise for detection of alterations in CFS/ME patients’ sleep. Stage shifts4,34 and dynamic stage transitions50 have been shown to discriminate CFS/ME patients and healthy controls. The relative frequency of REM to NREM transition is lower in CFS/ME, while there are significantly more transitions from REM and stage 1 sleep to wakefulness.50 Normal continuation of sleep in stage 1 and REM is disrupted in CFS/ME, which may contribute to feelings of non-restor-ative sleep upon waking.

More recently, studies have utilized alternative methods for quantifying EEG in CFS/ME patients, which provide a more sensitive method for evaluation of sleep abnormalities. Power spectral analysis using fast Fourier transform (FFT) has been used in a few recent studies,51,52 with conflicting results. Decker reported differences in delta power, with reduced delta power observed in SWS and increased delta power during stage 1 and REM sleep in 35 CFS/ME patients relative to 40 healthy con-trols.52 Reductions in alpha power were observed most strongly during REM but were also seen in SWS and stage 2 in the CFS/ME subjects. The finding of attenuated delta power comple-ments the reduced SWA reported previously following sleep delay,49 providing further evidence of an altered sleep homeo-stat in CFS/ME. In contrast, Armitage found no difference in

any frequency band between twins discordant for CFS/ME.51 There is a clear need for future research utilizing this method of sleep EEG analysis to clarify these findings.

It is also plausible that CFS/ME patients may experience arousals during sleep that are not detected using traditional scoring methods.67 Supporting this idea, more microarous-als have been observed in the sleep of CFS/ME patients than healthy control subjects.24,66 Recently, other methods of arous-ability or sleep instability have been developed and used in this population. Cyclic alternating pattern (CAP) is an EEG-derived measure of sleep instability, which is reflected as periodic EEG activity during NREM sleep.68,69 In contrast, periods of Non-CAP are indicative of consolidated sleep. CAP is somewhat distinct from typically measured arousals from sleep, both as a phenomenon and in terms of how they are scored.69 Clinical studies of insomniacs have found strong correlations between CAP rate (ratio of total CAP time to total NREM sleep time) and subjective reports of sleep quality.70 This measure may therefore provide a more specific marker of sleep quality that the arousal index derived from PSG. Only one study to date has examined CAP rate in CFS/ME patients.22 Despite having similar NREM sleep times, CFS/ME patients had higher CAP rates than matched controls, indicating higher NREM sleep in-stability in their CFS/ME patients. Abnormal CAP rate was also accompanied by an increase in slow wave delta power. Interest-ingly, there was no difference in arousal indices between the pa-tient and control groups. A number of CFS/ME patients in this study, however, were found to have nasal cannula flow limita-tion, indicative of upper airway resistance syndrome (UARS), which may have influenced the findings. CAP has also been as-sociated with UARS and other sleep related breathing disorders previously.71 Whether the higher CAP rate observed in CFS/ME patients was solely a result of UARS is unclear. Importantly, CAP was associated with both subjective fatigue and sleepiness ratings, as has been found previously in other sleep disordered populations.70

Taken together, these studies indicate that conventional sleep stage scoring methods use may not be sensitive enough to detect microstructural changes in CFS/ME patients. Consideration of these microstructure methods for analyzing the sleep EEG may provide a fruitful alternative for uncovering subtle differences during sleep in individuals reporting non-restorative sleep and daytime fatigue.

7.0 OTHER MEASURES OF SLEEP DISTURBANCE

7.1 Autonomic Activity Measures in CFS/MEAutonomic activity alterations, such as hypotension and

reduced heart rate variability (HRV), are a common feature of CFS/ME,72 and are a feature of the diagnostic criteria. In healthy individuals, autonomic nervous system dynamics also have characteristic profiles during sleep onset,73 and different sleep stages and depths of sleep.74 In some CFS/ME patients, the autonomic dysfunction observed during waking also trans-fers into sleep.75 HRV during sleep is consistently found to be significantly lower in CFS/ME patients compared to well-matched controls, reflecting a reduction in nocturnal parasym-pathetic activity.39,43,75,76 Decreased HRV is thought to reflect a


Review Articlepersistent state of autonomic hypervigilance. The influence of daytime physical activity, however, should not be dismissed as a potential confounder. Regression analyses have also demon-strated that HRV is the best predictor of subjective sleep quality in CFS/ME patients in one study.39

Cardiopulmonary coupling (CPC) is another emerging tech-nique, used as a measure of sleep quality and stability based solely on the electrocardiogram (ECG) signal.77 This technique records both heart rate (R-R interval) and respiration dynamics to generate a spectrogram of cardiopulmonary coupling. Based on this output, one can determine periods of high-frequency and low-frequency coupling, indicative of high and low sleep stability. CPC has been used to demonstrate sleep instability in sleep disordered populations,77 major depressive disorder,78 and fibromyalgia.79 There is preliminary evidence that sleep quality and stability measured by CPC is poor in CFS/ME patients, with reduced high-frequency coupling and increased low-frequency coupling.80 Studies utilizing such autonomic ac-tivity techniques indicate that autonomic measures during sleep may be a promising mechanism associated with non-restorative sleep in CFS/ME.43

7.2 Hormone ProfilesSubstantial research examining the pathophysiology of CFS/

ME indicates that the hypothalamic-pituitary axis (HPA) may be implicated in this disorder.81,82 The HPA also plays an im-portant role in sleep regulation.83 Dysregulation of the HPA has been examined using salivary cortisol profiling in a few stud-ies.84-86 Morning cortisol levels upon awakening, recognized as an indicator of the HPA response to stress, were attenuated in CFS/ME patients compared to healthy controls,85 with this difference most evident in females with CFS/ME.84 Given that CFS/ME is 2 to 3 times more prevalent in women, it has been proposed that sex differences in hypocortisolism may explain the increased risk of CFS in women.84

Heightened IL-6 plasma levels have also been reported in CFS/ME patients,87 possibly reflecting the low levels of sys-temic inflammation associated with CFS/ME. However, whether increased levels of IL-6 in this study were directly related to CFS/ME or other confounders such as BMI and un-derlying sleep disorders is unclear. IL-6 has been implicated in the pathogenesis of psychological and physiological fatigue in healthy individuals,88 and has been reported in other patient groups that suffer debilitating fatigue, such as cancer patients.89 Of direct relevance to this review, altered levels of IL-6 are also associated with somnolence and chronic sleep restriction.59 It is plausible that a link between IL-6 or other inflammatory mark-ers and sleep disturbance in CFS/ME exists; this remains an interesting hypothesis that warrants further investigation.

7.3 NeuroimagingNeuroimaging studies have allowed researchers to examine

the sleeping brain in healthy subjects. These techniques have also applied to clinical populations, including primary insom-nia,90 OSA,91,92 and depression.93 Particularly in primary insom-nia, functional neuroimaging studies during sleep have helped understand the neurophysiological underpinnings of sleep dys-regulation in these patients. In line with the hyperarousal theory of primary insomnia, patients were found to have abnormally

high regional brain activity during sleep compared to controls subjects.94 This is proposed as one of the mechanisms contribut-ing to sleep state misperception and sleep disturbance found in such patients. A growing number of studies are using functional neuroimaging to examine CFS/ME patients during wake (see Lange95 for a review). These studies have typically observed altered neural activity during performance of a fatiguing cogni-tive task in CFS/ME patients relative to controls, in the absence of performance impairment, potentially reflecting greater per-ceived cognitive effort.96,97 Future studies may also benefit from examining neural function and cerebral blood flow in CFS/ME patients during sleep.

8.0 DIRECTIONS FOR FUTURE RESEARCH

There are still many unanswered questions regarding the pathogenesis and nature of sleep disturbance and unrefreshed sleep in CFS/ME. Sleep disturbance may precipitate CFS/ME, may alter or complicate its course by worsening fatigue, pain, or mood, or may represent an independent factor unrelated to fatigue itself. With the development of new standardized crite-ria for diagnosing CFS/ME, more homogenous patient samples and comparability across studies will be afforded for future re-search. Standardized research designs with less ridged imposi-tion of sleep-wake times, and control over medication use and symptoms and comorbid conditions (such as pain and depres-sion) will assist in understanding the state of sleep in CFS/ME.

Non-restorative sleep has a fairly distinct subjective defi-nition; however, the physiological markers that underlie this experience and the extent to which alterations in sleep relate to the experience of non-restorative sleep in these patients are relatively unknown. Does non-restorative sleep stem from hy-perarousability, sleep hygiene, or circadian rhythm disturbance, or is it biologically driven in some other way? Is the nature of non-restorative sleep in CFS/ME similar to that experienced in other clinical conditions and by healthy individuals on oc-casion, or is it a heterogeneous phenomenon? There may also be other symptoms that CFS/ME patients experience that could cause non-restorative sleep, such as upper airway resistance syndrome, food intolerance, or immunological and metabolic changes. Emergence of new methods for analyzing the mi-crostructure of sleep have allowed researchers to detect more subtle EEG changes in CFS/ME patients during sleep. Given that very little is known about the nature and cause of non-re-storative sleep, these studies have opened up new avenues for investigating sleep disturbance and non-restorative sleep, not only in CFS/ME but other clinical conditions such as insomnia.

Having a clearer understanding about the pathophysiology of non-restorative sleep in CFS/ME may lead to better treatment options for patients. For example, one theory for the non-re-storative sleep experienced in insomnia suggests that negative cognitions trigger autonomic arousal during wakefulness that transfers into sleep.98 Nocturnal arousal is often treated by reducing hyperarousability and cognitions during the day, us-ing cognitive behavioral therapy99 or mindfulness based tech-niques,100 with the aim of reducing arousal at night. If a similar arousal phenomenon is found during sleep in CFS/ME, then insomnia-based treatments have the potential to be utilized in CFS/ME patients.


ML Jackson and D BruckMore precise tools and analyses for differentiating fatigue

and sleepiness are needed. Distinguishing these two states re-mains a diagnostic challenge for clinicians. While there are validated objective and subjective measures of sleepiness, only subjective measures currently exist for assessing fatigue. De-velopment of an objective measure of fatigue and a question-naire that assesses both fatigue and sleepiness would be useful not only in the context of CFS/ME, but for assessing other dis-orders in which both sleepiness and fatigue are a feature. In ad-dition, further research into the relationship between sleep and fatigue is greatly needed.

Patients commonly report improved sleep following some treatment regimes, such as graded exercise therapy and antibi-otic treatment for imbalances in gut microbial flora. There are avenues for research into objective assessments of sleep before and after treatment, and to follow the time course of changes in sleep as patients’ symptoms improve. Future studies are needed to explore the interaction of sleep homeostasis, SWA, and im-mune system activation in CFS/ME, given the changes in im-mune function that are observed in these patients.6

The issue surrounding possible sleep state misperception in CFS/ME raises the question of whether abnormalities in sleep associated with CFS/ME are similar to those found in other populations that experience sleep misperception, such as insomnia, or whether they are specific to CFS/ME. This also raises the question of whether CFS/ME patients report a big-ger discrepancy between objective and subjective measures of sleep compared to other medical conditions, and is this relative to the experience of non-restorative sleep? Understanding the association between these two conditions may provide insight into the mechanisms of sleep misperception. Future studies will benefit by incorporating subjective, physiological and behav-ioral measures of sleep to gain a broader insight into the nature of sleep disturbance of CFS/ME.

Finally, CFS/ME may provide a unique insight into the link between non-restorative sleep and fatigue. Other clinical condi-tions also experience non-restorative sleep and fatigue, such as FM, narcolepsy, and coronary heart disease,101,102 as well as a high number of otherwise healthy individuals.103 It is plausible that there may be subgroups of such people with fatigue and non-restorative sleep with similar underlying symptoms and etiologies, or, alternatively, there may be a continuum of such symptoms with CFS/ME being at the upper end of the spectrum. Given the constellation of physiological and psychological symptoms characterizing CFS/ME, examining sleep and fatigue in CFS/ME patients may allow us to better understand the neu-robiology and etiology of fatigue in other patient populations.

SUMMARY AND CONCLUSIONS

CFS/ME is a complex and severely debilitating condition. Non-restorative sleep, reduced sleep quality, and extended peri-ods of sleep are commonly reported, however the basis of these symptoms are unclear. The heterogeneities associated with this disorder make it challenging for researchers to study and make cross-study comparisons difficult. While there is little evidence of sleep architecture differences between CFS/ME and healthy individuals, many patients subjectively report sleep disturbance and unrefreshed sleep, highlighting a potential disconnect be-

tween objective and subjective measures of sleep. There is preliminary evidence that alteration in sleep stage transitions and sleep instability, and other physiological mechanisms such as heart rate variability and altered cortisol profiles, may be implicated in the sleep difficulties of this population. Further research is required to investigate the cause of non-restorative sleep and fatigue in CFS/ME, which may aid understanding of this symptom in other medical conditions.

ABBREVIATIONS

CAP, cyclical alternating patternCDC, United States Centers for Disease Control and PreventionCFS, chronic fatigue syndromeCPAP, continuous positive airway pressureCPC, cardiopulmonary couplingECG, electrocardiogramEDS, excessive daytime sleepinessEEG, electroencephalogramFFT, fast Fourier transformFM, fibromyalgiaHPA, hypothalamic-pituitary axisHRV, heart rate variabilityICC, International Consensus CriteriaICD-10 International Classification of DiseasesICSD, International Classification of Sleep DisordersIL-6, interleukin 6MDD, Major Depressive DisorderME, myalgic encephalomyelitisMSLT, multiple sleep latency testOSA, obstructive sleep apneaPI, Primary InsomniaPSG, polysomnographyPSQI, Pittsburgh Sleep Quality IndexS1, stage 1 sleepS2, stage 2 sleepSDQ, Sleep Disorder QuestionnaireSE, sleep efficiencySL, sleep latencySSS, Stanford Sleepiness ScaleSWA, slow wave activitySWS, slow wave sleepTIB, time in bedTST, total sleep time

REFERENCES1. Jason LA, King CP, Frankenberry EL, et al. Chronic fatigue syndrome: assessing

symptoms and activity level. J Clin Psychol 1999;55:411-24.2. Afari N, Buchwald D. Chronic fatigue syndrome: a review. Am J Psychiatry

2003;160:221-36.3. Ranjith G. Epidemiology of chronic fatigue syndrome. Occup Med 2005;55:13-9.4. Fischler B. Review of clinical and psychobiological dimensions of the chronic

fatigue syndrome: differentiation from depression and contribution of sleep dys-functions. Sleep Med Rev 1999;3:131-46.

5. Krupp LB, Jandorf L, Coyle PK, Mendelson WB. Sleep disturbance in chronic fatigue syndrome. J Psychosom Res 1993;37:325-31.

6. Lorusso L, Mikhaylova SV, Capelli E, Ferrari D, Ngonga GK, Ricevuti G. Immu-nological aspects of chronic fatigue syndrome. Autoimmunity Rev 2009;8:287-91.

7. Beard G. Neurasthenia, or nervous exhaustion. Boston Med Surg J 1869;80:217-20.8. Wolfe F, Smythe HA, Yunus MB, et al. The American College of Rheumatology

1990 criteria for the classification of fibromyalgia. Arthritis Rheum 1990;33:160-72.


Review Article9. Friedberg F, Jason LA. Chronic fatigue syndrome and fibromyalgia: Clinical as-

sessment and treatment. J Clin Psychol 2001;57:433-55.10. Holmes GP, Kaplan JE, Gantz NM, et al. Chronic fatigue syndrome: a working

case definition. Ann Intern Med 1988;108:387.11. Lloyd AR, Wakefield D, Boughton CR, Dwyer J. What is myalgic encephalomy-

elitis? Lancet 1988;8597:1286-7.12. Lloyd AR, Hickie I, Boughton CR, Spencer O, Wakefield D. Prevalence of chron-

ic fatigue syndrome in an Australian population. Med J Australia 1990;153:522-8.13. Sharpe MC, Archard LC, Banatvala JE, et al. A report--chronic fatigue syndrome:

guidelines for research. J Royal Soc Med 1991;84:118.14. Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. The chronic

fatigue syndrome: a comprehensive approach to its definition and study. Ann Intern Med 1994;121:953-9.

15. Reeves WC, Heim C, Maloney EM, et al. Sleep characteristics of persons with chronic fatigue syndrome and non-fatigued controls: results from a population-based study. BMC Neurol 2006;6:41.

16. Carruthers BM, Jain AK, Meirleir KLD, et al. Myalgic encephalomyelitis/chronic fatigue syndrome: clinical working case definition, diagnostic and treatment pro-tocols. J Chronic Fatigue Syndr 2003;11:7-116.

17. Carruthers BM, van de Sande MI, De Meirleir KL, et al. Myalgic encephalomyeli-tis: International Consensus Criteria. J Intern Med 2011;270:327-38.

18. Morriss RK, Sharpe M, Sharpley AL, Cowen PJ, Hawton K, Morris J. Abnormali-ties of sleep in patients with the chronic fatigue syndrome. BMJ 1993;306:1161-4.

19. Fossey M, Libman E, Bailes S, et al. Sleep quality and psychological adjustment in chronic fatigue syndrome. J Behav Med 2004;27:581-605.

20. Le Bon O, Fischler B, Hoffmann G, et al. How significant are primary sleep dis-orders and sleepiness in the chronic fatigue syndrome? Sleep Res 2000;3:43-8.

21. Nisenbaum R, Jones J, Unger E, Reyes M, Reeves W. A population-based study of the clinical course of chronic fatigue syndrome. Health Qual Life Outcomes 2003;1:49.

22. Guilleminault C, Poyares D, Da Rosa A, Kirisoglu C, Almeida T, Lopes MC. Chronic fatigue, unrefreshing sleep and nocturnal polysomnography. Sleep Med 2006;7:513-20.

23. Buchwald D, Pascualy R, Bombardier C, Kith P. Sleep disorders in patients with chronic fatigue. Clin Infect Dis 1994;18:S68-S72.

24. Neu D, Cappeliez B, Hoffmann G, Verbanck P, Linkowski P, Le Bon O. High slow-wave sleep and low-light sleep: Chronic fatigue syndrome is not likely to be a primary sleep disorder. J Clin Neurophysiol 2009;26:207-12.

25. Libman E, Creti L, Baltzan M, Rizzo D, Fichten CS, Bailes S. Sleep apnea and psychological functioning in chronic fatigue syndrome. J Health Psychol 2009;14:1251-67.

26. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328:1230-5.

27. Osorio CD, Gallinaro AL, Lorenzi-Filho G, et al. Sleep quality in patients with fibro-myalgia using the Pittsburgh Sleep Quality Index. J Rheumatol 2006;33:1863-5.

28. Lavigne G, Brousseau M, Kato T, et al. Experimental pain perception remains equally active over all sleep stages. Pain 2004;110:646-55.

29. Modofsky, H. The significance, assessment, and management of nonrestorative sleep in fibromyalgia syndrome. CNS Spectrums 2008;13:22-6.

30. Older SA, Battafarano DF, Danning CL. The effects of delta wave sleep inter-ruption on pain thresholds and fibromyalgia-like symptoms in healthy subjects; correlations with insulin-like growth factor I. J Rheumatol 1998;25:1180-6.

31. Benca RM, Obermeyer WH, Thisted RA, Gillin JC. Sleep and psychiatric disor-ders: A meta-analysis. Arch Gen Psychiatry1992;49:651-68.

32. Morriss RK, Wearden AJ, Battersby L. The relation of sleep difficulties to fa-tigue, mood and disability in chronic fatigue syndrome. J Psychosom Res 1997;42:597-605.

33. Le Bon O, Minner P, Van Moorsel C, et al. First-night effect in the chronic fatigue syndrome. Psychiatry Res 2003;120:191-9.

34. Fischler B, Le Bon O, Hoffmann G, Cluydts R, Kaufman L, De Meirleir K. Sleep anomalies in the chronic fatigue syndrome. Neuropsychobiology 1997;35:115-22.

35. Stores G, Fry A, Crawford C. Sleep abnormalities demonstrated by home poly-somnography in teenagers with chronic fatigue syndrome. J Psychosom Res 1998;45:85-91.

36. Sharpley A, Clements A, Hawton K, Sharpe M. Do patients with pure chronic fatigue syndrome (neurasthenia) have abnormal sleep? Psychosom Med 1997;59:592-6.

37. Whelton CL, Salit I, Moldofsky H. Sleep, Epstein-Barr virus infection, musculosk-eletal pain, and depressive symptoms in chronic fatigue syndrome. J Rheumatol 1992;19:939.

38. Togo F, Natelson BH, Cherniack N, FitzGibbons J, Garcon C, Rapoport DM. Sleep structure and sleepiness in chronic fatigue syndrome with or without coex-isting fibromyalgia. Arthiritis Res Ther 2008;10:R56.

39. Burton A, Rahman K, Kadota Y, Lloyd A, Vollmer-Conna U. Reduced heart rate variability predicts poor sleep quality in a case–control study of chronic fatigue syndrome. Exp Brain Res 2010;204:71-8.

40. Ohinata J, Suzuki N, Araki A, Takahashi S, Fujieda K, Tanaka H. Actigraphic assessment of sleep disorders in children with chronic fatigue syndrome. Brain Dev 2008;30:329-33.

41. Evering RMH, van Weering MGH, Groothuis-Oudshoorn KCGM, Vollenbroek-Hutten MMR. Daily physical activity of patients with the chronic fatigue syn-drome: a systematic review. Clin Rehabil 2011;25:112-33.

42. Tryon WW, Jason L, Frankenberry E, Torres-Harding S. Chronic fatigue syn-drome impairs circadian rhythm of activity level. Physiol Behav 2004;82:849-53.

43. Rahman K, Burton A, Galbraith S, Lloyd AR, Vollmer-Conna U. Sleep-wake be-havior in chronic fatigue syndrome. Sleep 2011;34:671-8.

44. Watson NF, Kapur V, Arguelles LM, et al. Comparison of subjective and objec-tive measures of insomnia in monozygotic twins discordant for chronic fatigue syndrome. Sleep 2003;26:324-8.

45. Ball N, Buchwald DS, Schmidt D, Goldberg J, Ashton S, Armitage R. Mono-zygotic twins discordant for chronic fatigue syndrome: Objective measures of sleep. J Psychosom Res 2004;56:207-12.

46. Majer M, Jones J, Unger E, et al. Perception versus polysomnographic assess-ment of sleep in CFS and non-fatigued control subjects: results from a popula-tion-based study. BMC Neurol 2007;7:40.

47. Neu D, Mairesse O, Hoffmann G, et al. Sleep quality perception in the chronic fatigue syndrome: correlations with sleep efficiency, affective symptoms and in-tensity of fatigue. Neuropsychobiology 2007;56:40-6.

48. Le Bon O, Neu D, Valente F, Linkowski P. Paradoxical NREMS distribution in “pure” chronic fatigue patients: a comparison with sleep apnea-hypopnea pa-tients and healthy control subjects. J Chronic Fatigue Syndrome 2007;14:2.

49. Armitage R, Landis C, Hoffmann R, et al. The impact of a 4-hour sleep delay on slow wave activity inb twins discordant for chronic fatigue syndrome. Sleep 2007;30:657-62.

50. Kishi A, Struzik ZR, Natelson BH, Togo F, Yamamoto Y. Dynamics of sleep stage transitions in healthy humans and patients with chronic fatigue syndrome. Am J Physiol 2008;294:R1980-R7.

51. Armitage R, Landis C, Hoffmann R, et al. Power spectral analysis of sleep EEG in twins discordant for chronic fatigue syndrome. J Psychosom Res 2009;66:51-7.

52. Decker M, Tabassum H, Lin J-M, Reeves W. Electroencephalographic correlates of chronic fatigue syndrome. Behav Brain Funct 2009;5:1-8.

53. Creti L, Libman E, Baltzan M, Rizzo D, Bailes S, Fichten CS. Impaired sleep in chronic fatigue syndrome. J Health Psychol 2010;15:596-607.

54. Kishi A, Natelson BH, Togo F, Struzik ZR, Rapoport DM, Yamamoto Y. Sleep stage transitions in chronic fatigue syndrome patients with or without fibromy-algia. In: Engineering in Medicine and Biology Society (EMBC). Annual Interna-tional Conference of the IEEE 2010; p. 5391-4.

55. Van Cauter E, Leproult R, Plat L. Age-related changes in slow wave sleep and REM sleep and relationship with growth hormone and cortisol levels in healthy men. JAMA 2000;284:861-8.

56. Manu P, Lane TJ, Matthews DA, Castriotta RJ, Watson RK, Abeles M. Alpha-delta sleep in patients with a chief complaint of chronic fatigue. South Med J 1994;87:465-70.

57. Flanigan MJ, Morehouse RL, Shapiro CM. Determination of observer-rated al-pha activity during sleep. Sleep 1995;18:702-6.

58. Dijk D-J, Brunner DP, Beersma D, Borbely AA. Electroencephalogram power density and slow wave sleep as a function of prior waking and circadian phase. Sleep 1990;13:430-40.

59. Madje JA, Krueger JM. Links between the innate immune system and sleep. J Allergy Clin Immunol 2005;116:1188-98.

60. Majde JA, Krueger JM. Neuroimmunology of sleep. In: D’haenen H, ed. Text-book of biological psychiatry. London: John Wiley & Sons, Ltd, 2002:1247-57.

61. Watson NF, Kapur V, Arguelles LM, et al. Comparison of subjective and objec-tive measures of insomnia in monozygotic twins discordant for chronic fatigue syndrome. Sleep 2003;26:324-8.

62. Watson NF, Jacobsen C, Goldberg J, Kapur V, Buchwald DS. Subjective and ob-jective sleepiness in monozygotic twins discordant for chronic fatigue syndrome. Sleep 2004;27:973-7.

63. Jackowska M, Dockray S, Hendrickx H, Steptoe A. Psychosocial factors and sleep efficiency: discrepancies between subjective and objective evaluations of sleep. Psychosom Med 2011;73:810-6.

64. Edinger JD, Krystal AD. Subtyping primary insomnia: is sleep state mispercep-tion a distinct clinical entity? Sleep Med Rev 2003;7:203-14.


ML Jackson and D Bruck65. Pigeon WR, Sateia MJ, Ferguson RJ. Distinguishing between excessive day-

time sleepiness and fatigue: Toward improved detection and treatment. J Psy-chosom Res 2003;54:61-9.

66. Neu D, Hoffmann G, Moutrier R, Verbanck P, Linkowski P, Le Bon O. Are pa-tients with chronic fatigue syndrome just ‘tired’ or also ‘sleepy’? J Sleep Res 2008;17:427-31.

67. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Public Health Service, U.S. Government Printing Office 1968.

68. Terzano MG, Mancia D, Salati MR, Costani G. The cyclic alternating pattern as a physiologic component of normal NREM sleep. Sleep 1985;8:137-45.

69. Terzano MG, Parrino L. Origin and significance of the cyclic alternating pattern (CAP). Sleep Med Rev 2000;4:101-23.

70. Parrino L, Milioli G, De Paolis F, Grassi A, Terzano MG. Paradoxical insomnia: The role of CAP and arousals in sleep misperception. Sleep Med 2009;10:1139-45.

71. Guilleminault C, Lopes MC, Hagen CC, Da Rosa A. The cyclic alternating pattern demonstrates increased sleep instability and correlates with fatigue and sleepi-ness in adults with upper airway resistance syndrome. Sleep 2007;30:641-7.

72. Yamamoto Y, LaManca JJ, Natelson BH. A measure of heart rate variability is sensitive to orthostatic challenge in women with chronic fatigue syndrome. Exp Biol Med 2003;228:167-74.

73. Shinar Z, Akselrod S, Dagan Y, Baharav A. Autonomic changes during wake–sleep transition: A heart rate variability based approach. Auton Neurosci 2006;130:17-27.

74. Trinder J, Kleiman J, Carrington M, et al. Autonomic activity during human sleep as a function of time and sleep stage. J Sleep Res 2001;10:253-64.

75. Boneva RS, Decker MJ, Maloney EM, et al. Higher heart rate and reduced heart rate variability persist during sleep in chronic fatigue syndrome: A population-based study. Auton Neurosci 2007;137:94-101.

76. Tajima S, Kuratsune H, Yamaguti K, et al. Estimation of fatigue state in patient with CFS using actigraph and R-R interval power spectrum analysis. Jpn J Clin Med 2007;65:1057-64.

77. Thomas RJ, Mietus JE, Peng CK, Goldberger AL. An electrocardiogram-based technique to assess cardiopulmonary coupling during sleep. Sleep 2005;28:1151-61.

78. Yang AC, Yang C-H, Hong C-J, et al. Sleep state instabilities in major depressive disorder: Detection and quantification with electrocardiogram-based cardiopul-monary coupling analysis. Psychophysiol 2011;48:285-91.

79. Thomas RJ, Mietus JE, Peng C-K, Goldberger AL, Crofford LJ, Chervin RD. Impaired sleep quality in fibromyalgia: Detection and quantification with ECG-based cardiopulmonary coupling spectrograms. Sleep Med 2010;11:497-8.

80. Cunnington D, Buccella D, Bastiampillai S, Swieca J. Sleep architecture and sleep stability in chronic fatigue syndrome. J Sleep Res 2011;20:25.

81. Demitrack M, Dale J, Straus S, et al. Evidence for impaired activation of the hypothalamic-pituitary-adrenal axis in patients with chronic fatigue syndrome. J Clin Endocrinol Metab 1991;73:1224-34.

82. Di Giorgio A, Hudson M, Jerjes W, Cleare AJ. 24-Hour pituitary and adrenal hor-mone profiles in Chronic Fatigue Syndrome. Psychosom Med 2005;67:433-40.

83. Vgontzas A, Chrousos G. Sleep, the hypothalamic-pituitary-adrenal axis, and cytokines: multiple interactions and disturbances in sleep disorders. Endocrinol Metab Clin North Am 2002;31:15-36.

84. Nater UM, Maloney E, Boneva RS, et al. Attenuated morning salivary cortisol concentrations in a population-based study of persons with chronic fatigue syn-drome and well controls. J Clin Endocrinol Metab 2008;93:703-9.

85. Roberts ADL, Wessely S, Chalder T, Papadopoulos A, Cleare AJ. Salivary cortisol response to awakening in chronic fatigue syndrome. Br J Psychiatry 2004;184:136-41.

86. Jerjes WK, Cleare AJ, Wessely S, Wood PJ, Taylor NF. Diurnal patterns of sali-vary cortisol and cortisone output in chronic fatigue syndrome. J Affect Disord 2005;87:299-304.

87. Nater UM, Youngblood LS, Jones JF, et al. Alterations in diurnal salivary cortisol rhythm in a population-based sample of cases with chronic fatigue syndrome. Psychosom Med 2008;70:298-305.

88. Singh A, Petrides JS, Gold PW, Chrousos GP, Deuster PA. Differential hypotha-lamic-pituitary-adrenal axis reactivity to psychological and physical stress. J Clin Endocrinol Metab 1999;84:1944-8.

89. Schubert C, Hong S, Natarajan L, Mills PJ, Dimsdale JE. The association be-tween fatigue and inflammatory marker levels in cancer patients: a quantitative review. Brain Behav Immunol 2007;21:413-27.

90. Smith MT, Perlis ML, Chengazi VU, et al. Neuroimaging of NREM sleep in pri-mary insomnia: a Tc-99-HMPAO single photon emission computed tomography study. Sleep 2002;25:325-35.

91. Morrell MJ, Jackson ML, Twigg GL, et al. Changes in brain morphology in pa-tients with obstructive sleep apnoea. Thorax 2010;65:908-14.

92. O’Donoghue FJ, Wellard RM, Rochford PD, et al. Magnetic resonance spectros-copy and neurocognitive dysfunction in obstructive sleep apnea before and after CPAP treatment. Sleep 2012;35:41-8.

93. Nofzinger EA, Price JC, Meltzer CC, Buysee DJ, Villemagne VL, Miewald JM. Towards a neurobiology of dysfunctional arousal in depression: the relationship between beta EEG power and regional cerebral glucose metabolism during NREM sleep. Psychiatry Res 2000;98:71-91.

94. Nofzinger EA, Buysee DJ, Germaine A, Price JC, Miewald JM, Kupfer DJ. Func-tional neuroimaging evidence for hyperarousal in insomnia. Am J Psychiatry 2004;161:2126-8.

95. Lange G, Wang S, DeLuca J, Natelson BH. Neuroimaging in chronic fatigue syndrome. Am J Med 1998;105:50S-3S.

96. Cook DB, O’Connor PJ, Lange G, Steffener J. Functional neuroimaging cor-relates of mental fatigue induced by cognition among chronic fatigue syndrome patients and controls. Neuroimage 2007;36:108-22.

97. Schmaling KB, Lewis DH, Fiedelak JI, Mahurin R, Buchwald DS. Single-photon emission computerized tomography and neurocognitive function in patients with chronic fatigue syndrome. Psychosom Med 2003;65:129-36.

98. Harvey AG. A cognitive model of insomnia. Behav Res Ther 2002;40:869-93.99. Harvey AG. A cognitive theory and therapy for chronic insomnia. J Cogn Psy-

chother 2005;19:41-59.100. Lundh LG. The role of acceptance and mindfulness in the treatment of insomnia.

J Cogn Psychother 2005;19:29-39.101. Modofsky H. The significance, assessment, and management of nonrestorative

sleep in fibromyalgia syndrome. Sleep Med 2008;13:22-6.102. Droogleever Fortuyn HA, Fronczek R, Smitshoek M, et al. Severe fatigue in

narcolepsy with cataplexy. J Sleep Res 2012;21:163-9. 103. Ohayon MM. Prevalence and correlates of nonrestorative sleep complaints.

Arch Intern Med 2005;165:35-41.

ACKNOWLEDGMENTSWork for this study was performed at Victoria University, Melbourne, Australia.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication April, 2012Submitted in final revised form June, 2012Accepted for publication June, 2012Address correspondence to: Melinda L. Jackson, School of Social Sciences and Psychology, Victoria University, PO Box 14428, Melbourne, Australia, 8001; Tel: +613 9919 9582; Fax: +613 9919 2218; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. Dr. Bruck received industry support from

Bioscreen Inc. over three years 2011-2013 for research involving Chronic Fatigue Syndrome patients. The other authors have indicated no financial conflicts of interest.



SLE

EP

ME

DIC

INE

PE

AR

LS

“Why Did My CPAP Beat Me Up?” Bilateral Periorbital Ecchymosis Associated with Continuous Positive Airway

Pressure TherapyLourdes DelRosso, M.D.; David E. McCarty, M.D.; Romy Hoque, M.D.

Louisiana State University School of Medicine, Department of Neurology, Division of Sleep Medicine, Shreveport, LA

The patient is a 63-year-old man with past medical history of hypertension, coronary artery disease,

diabetes mellitus, hyperlipidemia, gastroesophageal refl ux disease, and obstructive sleep apnea (OSA) diagnosed at the age of 58. Medications include as-pirin 325 mg daily and clopidogrel 75 mg daily for 2 years without complication; clonazepam 1 mg daily, gemfi brozil 600 mg twice daily, metoprolol 12.5 mg twice daily, niacin 500 mg daily, ranitidine 150 mg twice daily, rosuvastatin 10 mg daily, and valsartan 160 mg daily. Two months ago he started using vi-tamin E 1000 international units per day and fi sh oil supplements. For 5 years he has been compliant with continuous positive airway pressure (CPAP) of 10 cm through a soft nasal pillow interface and soft head

gear without complication. CPAP compliance card re-port indicates 100% CPAP use > 4 h/day, average use 9 h/day, and no signifi cant air leaks.

He now presents with 2 episodes of bruising around his eyes upon awakening following nocturnal CPAP use. He initially noticed bruising around the left eye which resolved over the course of one month. He did not pursue evaluation or treatment. Then he noticed new bruising around the right eye. He denied histo-ry of trauma, headaches, eye dryness or pain, vision changes, cough, sinus pain, or nasal discharge. He de-nied strenuous exercise or weight lifting. Signifi cant air leak from his CPAP mask was not detected. He denied any other bleeding, including gums, urinary tract, or rectum.

Figure 1—Periorbital ecchymosis associated with continuous positive airway pressure therapy for obstructive sleep apnea

The photograph on the left shows the patient with periorbital ecchymosis around the left eye, which resolved. The photograph on the right was taken one month later and showed a new right periorbital ecchymosis.


Sleep Medicine PearlsPhysical examination revealed a large periorbital ecchy-

mosis around the right eye (Figure 1). No scleral, skin, or nail bed petechial hemorrhages were noted. Pupils were round and reacted to light bilaterally. Visual fields were intact bilaterally. Fundoscopic examination revealed sharp discs bilaterally, with no evidence of intraocular or retinal hemorrhage.

Laboratory evaluation revealed no evidence of coagulopa-thy, thrombocytopenia, or Vitamin C deficiency (table 1). Aspirin and clopidogrel platelet inhibition assays showed normal platelet inhibitory responses to these medications. Computerized tomography of the head was unremarkable, with no evidence of intracranial bleeding or basilar skull fracture.

QUESTIONWhat is the most likely mechanism for this patient’s bilateral periorbital ecchymosis? What is your treatment recommendation?

Table 1—Laboratory ResultsCoagulation Labs Result Normal Range

Platelet count 227 150-450Prothrombin time (PT) 13.4 seconds 12-14.7 secondsPartial thromboplastin time (PTT) 29 seconds 24-37 secondsInternational normalized ratio 1Antithrombin III 126% 80-123%Factor VIII level 122% 61-147%Von Willebrand 98% 50-140%Platelet inhibition for aspirin 399 ARU 0-550 ARUPlatelet inhibition for clopidogrel 76% 40-60%

Other LabsAspartate transaminase (AST) 28 U/L 15-40 U/LAlbumin 3.9 g/dL 3.4-5.0 g/dLTotal bilirubin 0.5 mg/dL 0.02-1.2 mg/dLBlood urea nitrogen (BUN) 16.9 mg/dL 6-20 mg/dLCreatinine 0.8 mg/dL 0.6-1.3 mg/dLVitamin C level 0.9 mg/dL 0.4-9 mg/dL


L DelRosso, DE McCarty and R Hoque

ANSWERInitiation of Vitamin E and fish oil supplementation in combination with aspirin and clopidogrel use may have led to ecchymosis. We recommended discontinuation of Vitamin E and fish oil supplements.

Ecchymosis is an area of blood extravasation larger than 5 mm in size and is usually secondary to small or medium sized blood vessel trauma.3 In patients with either congenital or acquired co-agulation defects, even trivial trauma may result in ecchymosis. Petechial hemorrhages are smaller (< 2 mm) and typically due to damage to dermal capillaries. Causes of petechial hemorrhages in-clude increased hydrostatic pressure in the capillary system (e.g., post-tussive or post-strangulation); infections (e.g., Rocky Moun-tain Spotted Fever); small vessel vasculitis (e.g., Henoch Schön-lein Purpura); and poor capillary structural integrity (e.g., scurvy or amyloidosis). Confluent periorbital petechial hemorrhages may morphologically resemble ecchymosis as can be seen in amyloi-dosis.4 Our patient had no clinical features, or laboratory results suggestive of congenital or acquired coagulation defects; increased hydrostatic pressure; infection; vasculitis; or amyloidosis.

Complications associated with nasal CPAP therapy for OSA include nasal congestion; rhinorrhea; sneezing; mild to moder-ate epistaxis; and rarely severe epistaxis.1,2 To our knowledge, periorbital ecchymosis has not previously been reported as a complication of CPAP therapy.

We postulated that this patient’s recurrent periorbital ecchymo-sis was due to a combination of factors. Vitamin E has anticoagu-lant properties, possibly due to an inhibition of collagen-induced platelet activation and protein kinase C-dependent aggregation.5 Omega-3 fatty acids are also natural anticoagulants, and case reports indicate an increased bleeding risk when combined with other forms of anticoagulation.6 The combined antiplatelet activ-ity of aspirin, clopidogrel, Vitamin E, and fish oil supplements, along with a CPAP induced elevation in central venous hydro-static pressure, may have led to the development of periorbital capillary damage. The initial left periorbital ecchymosis followed by right periorbital ecchymosis may have been due in part to sleep position and/or positioning of the nasal CPAP nasal mask. Dis-continuation of the Vitamin E and fish oil supplements resulted in complete resolution of the CPAP associated periorbital ecchymo-sis. The patient continued using CPAP with adequate compliance.

CLINICAL PEARLS

1. Bruising associated with CPAP should prompt a search for a bleeding diathesis.

2. Assessment of over-the-counter supplement use and their potential interactions with prescription medications is an important component of the sleep medicine history.

3. Vitamin E and fish oil supplements should be used with caution in patients on antiplatelet medications such as as-pirin or clopidogrel.

4. Even low to moderate continuous positive airway pres-sures (CPAP) can lead to facial ecchymosis in patients on multiple antiplatelet medications or supplements.

CITATIONDelRosso L; McCarty DE; Hoque R. “Why did my CPAP beat me up?” Bilateral periorbital ecchymosis associated with continuous positive airway pressure therapy. J Clin Sleep Med 2012;8(6):730-732.

REFERENCES1. Pepin JL, Leger P, Veale D, Langevin B, Robert D, Levy P. Side effects of nasal

continuous positive airway pressure in sleep apnea syndrome. Study of 193 patients in two French sleep centers. Chest 1995;107:375-81.

2. Strumpf DA, Harrop P, Dobbin J, Millman RP. Massive epistaxis from nasal CPAP therapy. Chest 1989;95:1141.

3. Girolami A, Luzzatto G, Varvarikis C, Pellati D, Sartori R, Girolami B. Main clini-cal manifestations of a bleeding diathesis: an often disregarded aspect of medi-cal and surgical history taking. Haemophilia 2005;11:193-202.

4. Outteryck O, Stojkovic T, Launay D, Meignie-Ramon B, Vermersch P. Periorbital ecchymoses are not pathognomonic of the light-chain type of amyloidosis. Acta Derm Venereol 2007, 87:544-5.

5. Phang M, Lazarus S, Wood LG, Garg M. Diet and thrombosis risk: nutrients for prevention of thrombotic disease. Semin Thromb Hemost 2011;37:199-208.

6. Buckley MS, Goff AD, Knapp WE. Fish oil interaction with warfarin. Ann Phar-macother 2004, 38:50-2.

ACKNOWLEDGMENTThe authors acknowledge Dr. Cesar Liendo for his advice with this manuscript.

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication January, 2012Submitted in final revised form March, 2012Accepted for publication April, 2012Address correspondence to: Lourdes DelRosso, M.D., Assistant Professor of Sleep Medicine, Sleep Medicine Program, Department of Neurology, Louisiana State University School of Medicine, Shreveport, Louisiana; E-mail: [email protected]





LETT

ER

TO

TH

E E

DIT

OR

Gabapentin Effi cacy in Reducing Nighttime Awakenings in Premenopausal Women: A Class Effect of GABAergic

Medications or Unique Property of Gabapentin Only?Gautam Ganguly, M.D.

Neurology Consultants Medical Group, Whittier, CA

The recent scientifi c investigation by Guttuso1 pub-lished in the Journal of Clinical Sleep Medicine

suggested gabapentin to be effective in reduction of a novel sleep disorder unique to women found to have low serum estradiol and nighttime awakenings. This proposed sleep disorder was coined LUNAs. Gaba-pentin has been in use for treatment of neurological, psychiatric, chronic pain syndromes, and even meno-pausal-related hot fl ashes.2-4 The mechanism of action of gabapentin in both LUNAs and hot fl ashes is unclear, however. Gabapentin is a structural analogue of GABA, an inhibitory neurotransmitter. Gabapentin is known to reduce the sleep latency, increase total sleep time, slow wave sleep, and REM sleep. Pharmacologically, it is known to increase GABA in central nervous system.5

The theories behind nighttime awakenings and hot fl ushes in pre menopausal women are hypothesized to be a combination of “narrowing of the thermo neu-tral zone” (regulated by hypothalamus) along with “increase in sympathetic activity.”6 Increased fi ring of GABAergic neurons in the VLPO nucleus of the hypo-thalamus leads to increase in GABA, resulting in acti-vation of the sleep promoting neurons and suppression of the awake aminergic neurons during NREM sleep. Hypothetically, GABA may also suppress the thermo neutral zone of the hypothalamus during NREM sleep, when night sweats and frequent awakenings are most common. Medications like gabapentin with the poten-tial of increasing GABA in the central nervous system during sleep may help in reducing LUNAs.

Is the action of gabapentin in reducing LUNAs, as reported in this scientifi c investigation, unique to gabapentin, or do other medications like pregabalin have the potential of increasing GABA in the central nervous system during sleep? Most LUNAs, with or without night sweats, occur in NREM sleep, usu-ally in the fi rst half of the night. Medications such as gamma hydroxybutyrate or pregabalin that con-solidate NREM sleep by increasing slow wave sleep may be helpful in reducing these symptoms in this particular group of patients. Additionally, further studies with overnight polysomnogram and qualita-tive electroencephalograms may help in objective as-

sessment of these medications in reducing the night sweats and LUNAs.

Premenopausal women with LUNAs and night sweats have few therapeutic options. It has been re-ported that the use of medications like selective sero-tonin receptor agonists such as paroxetine7or hormone replacement therapy (HRT) may be helpful in reduc-ing night sweats in premenopausal women. However, many patients are reluctant to use HRT due to concerns of increased risk of breast cancer and cardiovascular events as alluded to in the Women’s Health Initiative report.8 This scientifi c report by Dr. Guttuso shows that gabapentin may be an alternative with a more acceptable, mild side effect profi le than the present therapeutic options available for this specifi c group of patients. This may result in improvement in quality of life by reducing LUNAs that may cause fragmenta-tion of sleep and sleep maintenance insomnia.

CITATIONGanguly G. Gabapentin effi cacy in reducing nighttime awakenings in pre-menopausal women: a class effect of GABAergic medications or unique property of Gabapentin only? J Clin Sleep Med 2012;8(6):733-734.

REFERENCES1. Guttuso T. Nighttime awakenings responding to gabapentin thera-

py in late premenopausal women: a case series. J Clin Sleep Med 2012;8:187-9.

2. Butt DA, Lock M, Lewis JE, Ross S, Moineddin R. Gabapentin for the treatment of menopausal hot fl ashes: a randomized controlled trial. Menopause 2008;15:310-8.

3. Reddy SY, Warner H, Guttuso T Jr, et al. Gabapentin, estrogen and placebo for treating hot fl ashes: a randomized controlled trial. Obstet Gynecol 2006;108:41-8.

4. Lopizi L, Barton DL, Sloan JA, et al. Pilot evaluation of gabapentin for treating hot fl ashes. Mayo Clinic Proc 2002;77:1159-63.

5. Brunton LL, Chabner BA, Knollman BC, eds. Goodman and Gil-man’s the pharmacological basis of therapeutics. 12th ed. Mc-Graw-Hill, 2010.

6. Kryger M, Roth T, Dement C, eds. Principles and practice of sleep medicine. 5th ed. Saunders, 2010.

7. Stearns V, Slack R, Greep N, et al. Paroxetine is an effective treat-ment for hot fl ashes results from a prospective randomized clinical trial. J Clin Oncol 2005;28:6919-30.

8. Women’s Health Initiative- National Institute of Health, 1991. Wom-en’s Health Initiative Program Offi ce, Bethesda, MD.


G Ganguly

SUBMISSION & CORRESPONDENCE INFORMATIONSubmitted for publication June, 2012Submitted in final revised form September, 2012Accepted for publication September, 2012Address correspondence to: Gautam Ganguly, M.D., Neurological Consultants Medical Group, 12291 E Washington Blvd, Ste#303, Whittier, CA 90606; Tel: (562) 698-6296; Fax: (562) 693-6752; E-mail: [email protected]

DISCLOSURE STATEMENTThis was not an industry supported study. The author has reported no financial


ISSN 1550-9389 iN this issue · ISSN 1550-9389 Volume 8, Number 6 / December 15, 2012 Pages 627-734...

Documents

Transcript of ISSN 1550-9389 iN this issue · ISSN 1550-9389 Volume 8, Number 6 / December 15, 2012 Pages 627-734...