ARTICLE IN PRESS

Respiratory Medicine (2005) 99, 718–725

KEYWORDSleep;ICU;Apnea;HypoventilPolysomnoObesity;Failure

0954-6111/$ - sdoi:10.1016/j.r

�CorrespondiE-mail addr

Sleep-related breathing disorders in obese patientspresenting with acute respiratory failure

Ahmed BaHammam�, Suhail Syed, Abdelrahman Al-Mughairy

Sleep Disorders Center, Respiratory Unit, Department of Medicine, College of Medicine, King SaudUniversity, PO Box 2925, Riyadh 11461, Saudi Arabia

Received 3 July 2004

S

ation;graphy;

ee front matter & 2004med.2004.10.021

ng author. Tel.: +966 1ess: ashammam2@yaho

Summary Introduction: The study was conducted to assess the clinical andpolysomnographic characteristics of patients with sleep-related breathing disorderswho presented to the intensive care unit (ICU) with acute respiratory failure and thepracticability of performing polysomnography for such patients.

Material and methods: We analyzed clinical presentation, cause of admission tothe ICU, ICU course and outcome of 11 subjects with acute respiratory failure whowere diagnosed to have sleep disordered breathing based on polysomnographybetween October 1999 and January 2003. Subjects were compared to 11 patientswith obstructive sleep apnea syndrome matched to each subject using body massindex, age and apnea hypopnea index measured at the time of diagnosis (matchedcomparison group). Repeated arterial blood gases and polysomnography were donefor 8 subjects compliant to treatment 6–8 months after discharge from ICU.

Results: The reason for ICU admission for all subjects was hypercapnic respiratoryfailure. pH and daytime PaO2 were significantly lower in studied subjects comparedto the matched comparison group while awake daytime PaCO2 was significantlyhigher. Subjects had frequent episodes of hypoventilation. Follow up arterial bloodgases and polysomnography 6–8 months after treatment (non-invasive ventilation) incompliant subjects showed significant improvement in all blood gases para-meters.

Conclusions: Early polysomnography (or portable cardio-respiratory monitoring)allows accurate diagnosis and institution of the appropriate ventilation methodFurther studies should assess the evolution of respiratory drive in patients with sleepdisordered breathing and hypercapnia under therapy (non-invasive ventilation).& 2004 Elsevier Ltd. All rights reserved.

Elsevier Ltd. All rights reserved.

467 0897; fax: +966 1 467 2558.o.com (A. BaHammam).

ARTICLE IN PRESS

Respiratory failure in sleep disordered breathing 719

Introduction

Sleep disordered breathing is a relatively commondisorder.1 Serious complications like hypertension,2

ischemic heart disease,3 stroke4 and pulmonaryhypertension5 have been reported secondary tosleep disordered breathing. Patients with obstruc-tive sleep apnea syndrome have been shown to haveincreased mortality,6 which has been attributed tocomorbid conditions and complications of obstruc-tive sleep apnea syndrome like obesity, hypertensionand ischemic heart disease. The association be-tween sudden death and sleep disordered breathinghas been reported earlier in patients labeled withthe diagnosis of ‘‘Pickwickian syndrome’’.7 Therelationship between sleep disordered breathingand acute respiratory failure is not well established.Few papers in the literature addressed the issue ofacute respiratory failure in patients with sleepdisordered breathing.8–10 Patients with sleep dis-ordered breathing and respiratory failure may bemisdiagnosed as having chronic obstructive pulmon-ary disease (COPD) or chronic heart disease resultingin delay in receiving the appropriate treatment.

This study was conducted to describe the clinicaland polysomnographic characteristics of patientswith sleep-related breathing disorders who presentto the ICU with acute respiratory failure and toassess the practicability of performing polysomno-graphy in this group of patients.

Material and methods

The study was conducted at King Khalid UniversityHospital (KKUH), which is a 650 beds tertiary carehospital that accepts referrals from differentregions of Saudi Arabia. Around 60–70 patients ayear are admitted to the medical intensive careunit (ICU) for respiratory failure. Data from elevensubjects with acute respiratory failure diagnosed tohave sleep-related breathing disorders (based onpolysomnography) between October 1999 andJanuary 2003 during ICU admission were analyzedfor clinical presentation, reason for ICU admission,management, response to treatment and outcome.The diagnosis of obstructive sleep apnea wasalready established in one subject and continuouspositive airway pressure (CPAP) was prescribed,however, he was not compliant to treatment. Theremaining ten subjects were not known to havesleep-related breathing disorders before.

The diagnosis of sleep disordered breathing wasentertained based on clinical suspicion with atypical history obtained from relatives (snoring,

witnessed apnea by family members or excessivedaytime sleepiness), body habitus associated withwitnessed apneas by the ICU staff or repetitivedesaturation using pulse oximeter with fast sam-pling rate (3–10 s) and negative history for alcohol,hypnotic or psychotropic drug dependence. Apartfrom occasional linear atelectasis, chest radio-graphs showed no other abnormalities. Polysomno-graphy was performed before discharge from ICU.For one subject, polysomnography was performedin the ICU. The remaining subjects were trans-ferred to the nearby sleep disorders centeraccompanied by an ICU nurse. Polysomnographywas not performed for subject number 5 as he wasdiagnosed before in another hospital to haveobstructive sleep apnea syndrome. Critical caremonitoring was continued while in the sleepdisorders center. No polysomnography was per-formed for patients during mechanical ventilation.None of the other subjects was known to havechronic lung diseases nor history of skeletal orneuromuscular diseases. Ventilated patients re-ceived sedation during intubation only. After that,no sedation was given to them. For ventilatedpatients, polysomnography was performed 3–4 daysafter extubation. While in the ICU before poly-somnography was performed, subjects were startedon Bi-level positive airway pressure (BilevelPAP)(IPAP 10 cm H2O and EPAP 6 cm H2O). Treatmentmodality and pressures were adjusted later basedon the polysomnographic findings.

Subjects were compared to 11 patients withobstructive sleep apnea syndrome matched to eachsubject using body mass index, age and apneahypopnea index measured at the time of diagnosis(matched comparison group). The comparison groupcomprised of consecutive patients referred electivelyto the sleep disorders center during the study periodfor obstructive sleep apnea syndrome diagnosis.

A data entry sheet was completed for eachsubject assessing; demographic data, reason forICU admission, duration of ICU stay, co-morbidconditions, medications, investigations, and me-chanical ventilation methods. Repeated arterialblood gas analysis and polysomnography were donefor eight subjects compliant to treatment (CPAP orbilevel positive airway pressure (BilevelPAP)) 6–8months after discharge from ICU. Pulmonary func-tion testing (FEV1, FVC) was performed for subjectsbefore discharge from the hospital.

Polysomnography

Alice-4 diagnostic equipment from Respironics, Inc,Murrysville, Pennsylvania, USA was used in data

ARTICLE IN PRESS

A. BaHammam et al.720

acquisition. Data were then downloaded in to IBM-PC. The patient was connected to the monitoringmachine by a trained technician. The technicianstayed with the patient throughout the study.

During the study, the following parameters weremonitored: four electro-encephalography (EEG)(C1-A4, C2-A3, O1-A4, O2-A3), chin electromyogra-phy (EMG), electro-oculography (EOG) (eye move-ment), leg EMG, electro-cardiography (ECG),oxygen saturation, chest and abdominal wall move-ments, two air flow signals (nasal prong pressureand thermistor), sleep position, and microphone forsnoring.

During polysomnography the following protocolwas used. During the initial 1–3 h, the patient wasstudied without intervention to identify the natureand severity of the underlying sleep disorderedbreathing. After that, treatment was started. CPAP(Respironics Inc, Muraysville, PA) was attemptedinitially. BilevelPAP (Respironics Inc, Muraysville,PA) was started if the patient could not tolerateCPAP due to persistent massive mask air leakage ordiscomfort exhaling against positive pressure, orthe patient had frequent episodes of hypoventila-tion without airway obstruction (revealed as aplateau on the inspiratory flow signal using nasalprong pressure without a thoracoabdominal para-dox). Pressure titration was performed according toa described protocol.11 Oxygen supplementationwas added if the patient continued to havehypoxemia despite complete elimination of theobstructive respiratory events. Once the mode ofsupport, pressure level and level of oxygen supple-mentation have been determined, the patient wastransferred back to the ICU. The needed supportmeasures were instituted in the ICU using mechan-ical ventilators in the ICU with internal alarms andmonitors.

All sleep studies were analyzed by the principalauthor to maximize the inter-scorer and intra-scorer reliability.

Analysis and scoring of the sleep studies

Analysis and scoring of the electronic raw data wasdone manually page by page evaluating time in bed,total sleep time (TST), sleep stages, and arousalindex according to established criteria.12,13 Reportswere generated using the Alice-4 software. Standarddefinitions were used to classify central, obstructiveapneas and hypopneas and central hypoventila-tion.14 Apnea was defined as the cessation of airflowX10 s, and hypopnea as a recognizable, transientreduction of airflow (450%) for X10 s usingnasal prong pressure flow signal. The diagnosis of

obstructive sleep apnea syndrome was based onhistory of snoring and excessive daytime sleepinessand apnea hypopnea index of410h. Hypoventila-tion was defined as sustained oxygen desaturationthat was not associated with obstructive apneas orhypopneas or periodic breathing.15–17 Patients withhypoventilation were labeled as having sleep hypo-ventilation syndrome (SHVS).

During follow up visits, compliance to treatment(CPAP, BilevelPAP and home oxygen) was assessedby asking family members who are living with thepatient. Good compliance was defined as usingCPAP or BilevelPAP more than five nights perweek, poor compliance as less than three nightsper week, and intermediate compliance between3 and 5 nights per week. For one patient, hismachine was provided with a data-recording cardthat allowed objective assessment of complianceduring every visit.

Statistical analysis

Data are expressed in text and tables as mean7-standard error of mean (SEM). Comparisons betweensubjects and their matched comparison group weremade using t-test. For comparing the studiedsubjects before and after treatment, paired t-testwas used. Results were considered statisticallysignificant at the Po0:05:

Results

Eleven subjects were included in the study, 6 malesand 5 females. Table 1 shows the characteristics ofall 11 subjects on admission. All subjects wereadmitted to the ICU due to hypercapnic respiratoryfailure. All studied subjects were discharged to theward, in which ten of them were discharged homelater. Subject number 5 required re-admission toICU two days after transfer to the ward as herefused to use CPAP. On his second admission, hedeveloped severe sepsis and expired after 3 days.Subject number 3, was sent home in a stablecondition on BilevelPAP and home oxygen. How-ever, she was not compliant to treatment andrefused to use BilevelPAP or to have tracheostomy.Three months after discharge, she developedhypercapnic respiratory failure and congestiveheart failure necessitating intubation and mechan-ical ventilation. The patient expired during thatadmission. Unfortunately, the details of her lastadmission were unavailable since she was admittedto another hospital. Subject number 10 refused to

ARTIC

LEIN

PRES

S

Table 1 Characteristics of the studied subjects.

SubjectNo

Sex Reason foradmission

Cor-pulmonale

Apneaknown

BP GCS APACHEscore

Mechanicalventilation

PSG diagnosis ICU stay(days)

Outcome

1 M Hypercapnic RF Yes No 130/70 13 24 No SHVS+OSA 4 Discharged2 M Hypercapnic RF Yes No 145/69 11 17 No SHVS+OSA 7 Discharged3 F Hypercapnic RF Yes No 140/90 7 17 Yes SHVS+OSA 11 Discharged*

4 M Hypercapnic RF Yes No 147/70 15 24 No SHVS+OSA 5 Discharged5 M Hypercapnic RF Yes Yes 136/58 7 10 Yes OSA 7 Discharged**

6 M Hypercapnic RF Yes No 120/80 15 19 No SHVS+OSA 3 Discharged7 M Hypercapnic RF Yes No 160/70 15 32 No OSA 6 Discharged8 F Hypercapnic RF Yes No 160/100 15 25 No SHVS+OSA 14 Discharged9 F Hypercapnic RF Yes No 130/70 15 30 No SHVS+OSA 4 Discharged10 F Hypercapnic RF Yes No 150/70 11 21 Yes SHVS 8 Discharged11 F Hypercapnic RF Yes No 124/60 13 21 No SHVS+OSA 3 DischargedMean(SEM)

140/74(4.1/3.7)

12.5 (0.9) 21.8 (1.9) 6.5 (1)

CHF: congestive heart failure; GCS: Glasgow coma scale (range from 3 (poor)-15 (good)); APACHE: Acute Physiology and Chronic Health Evaluation score (range 0–71, increasing scorerepresents a greater severity of illness); RF: respiratory failure; SHVS: sleep hypoventilation syndrome; OSA: obstructive sleep apnea.

*Discharged home but readmitted again to ICU few months later and expired.**Discharged to the ward and readmitted few days later with septic shock and expired.

Respiratory

failurein

sleepdisord

eredbreathing

721

ARTICLE IN PRESS

A. BaHammam et al.722

use BilevelPAP. She was admitted 4 times that yearfollowing her diagnosis. Subject number 11 did notcome for follow up.

All subjects were diagnosed to have pulmonaryhypertension based on clinical picture and echo-cardiographic findings. Repeated echocardiographyfor subject numbers 1 and 4 one year aftertreatment revealed normalization of the pulmonaryartery pressure (no data was available for othersubjects).

Table 2 summarizes the characteristics of sub-jects and their matched comparison group. Meanage for subjects was 5374.9 years and body massindex was 47.774.1 kg/m2. Mean age of thecomparison group was 54.273.1 years and bodymass index was 44.171.9 kg/m2. Two subjectswere known to have hypothyroidism and were onthyroid hormone replacement at the time ofadmission. Thyroid function was normal for theremaining subjects and the matched comparisongroup. Both subjects and their matched comparisongroup were identical in the apnea hypopnea index.However, compared to the comparison group,subjects spent more time with PaO2o90%, hadhigher desaturation index and their lowest re-corded oxygen saturation was significantly lower.Desaturation index was higher in subjects becausethis group had frequent episodes of sustainedoxygen desaturation that was not associated withobstructive apneas, hypopneas or periodic breath-ing. Those episodes were not scored as apneas orhypopneas and were labeled as episodes of hypo-ventilation. The affected subjects were diagnosedto have sleep hypoventilation syndrome. Ninesubjects were labeled as having sleep hypoventila-tion syndrome.

pH was significantly lower in subjects(7.2970.02) compared to their matched compar-ison group (7.470.01) (Po0:05). Awake daytimePaO2 was also significantly lower in subjects

Table 2 Comparison between subjects and their matfindings upon diagnosis and arterial blood gases paramete

AgeBody mass indexApnea hypopnea indexTime with oxygen saturation o90% (% of total sleep timeDesaturation indexMinimum oxygen saturationpHPaCO2 (mmHg)PaO2 (mmHg)HCO3 (mmol/L)

compared to their matched comparison group(Po0:05). All subjects had daytime hypercapniawith PaCO2 ranging from 52.2 to 116mmHg(77.975.1).

Pulmonary function testing was performed in allsubjects before discharge from the hospital. Itshowed obstructive pattern in subject numbers 3, 8and 11. Pulmonary function measurements werewithin normal limits for the rest of the subjects.

Table 3 summarizes the treatment modalitiesused. Ten of the studied subjects required Bilevel-PAP and supplemental oxygen. One subject wascontrolled by CPAP alone. Eight subjects werecompliant to treatment. Six to eight months aftertreatment, all compliant subjects did not requireoxygen supplementation while awake. However,three of them continued to need nocturnal oxygenhooked to BilevelPAP during sleep.

Table 4 summarizes the arterial blood gasparameters on admission and 6–8 months aftertreatment in compliant subjects. It demonstratessignificant improvement in all parameters.

Discussion

Obesity is known to be a major risk factor for arange of respiratory disorders.18 The present studydescribes an unusual cause of acute respiratoryfailure necessitating ICU admission in obese pa-tients. We described 11 subjects with sleep-relatedbreathing disorders who developed severe hyper-capnic respiratory failure necessitating ICU admis-sion. Early diagnosis allowed early institution ofappropriate treatment, hence, avoiding the needfor invasive mechanical ventilation in most sub-jects. This study stresses the importance of earlysleep study in patients with acute respiratoryfailure suspected of having sleep-related breathing

ched comparison group (reported polysomnographicrs upon admission to the ICU).

Subjects Comparison group P value

53.374.9 54.273.1 NS47.774.1 44.171.9 NS45.479.0 48.875.8 NS

) 76.976.9 33.477.9 o0.0578.5710.3 56.173.8 o0.0563.173.5 76.678.4 o0.057.2970.02 7.470.01 o0.0577.975.1 36.371.7 o0.0548.372.9 76.273.3 o0.0538.872.2 24.373.6 0.05

ARTICLE IN PRESS

Table 3 Treatment modalities for studied subjects on diagnosis and 6-8 months after treatment.

Treatment modality Follow up after 6–8 months

Subjectno

Treatmentmodality

IPAPcmH2O

EPAPcmH2O

Need fordaytime O2

Compliancewith treatment

Need forPPV

Need fordaytime O2

Need fornocturnal O2

1 BiPAP+O2 8 4 Yes Good Yes No Yes2 BiPAP+O2 8 4 Yes Good Yes No No3 BiPAP+O2 10 6 Yes Poor Yes Yes Yes4 BiPAP+O2 20 16 Yes Good Yes No Yes5 BiPAP+O2 10 6 Yes Poor No data No data No data6 BiPAP+O2 18 14 Yes Good Yes No No7 CPAP — 12 No Good Yes No No8 BiPAP+O2 8 4 Yes Good Yes No No9 BiPAP+O2 12 8 No Good Yes No No10 BiPAP+O2 10 6 Yes Poor Yes Yes Yes11 BiPAP+O2 10 6 Yes No data No data No data No data

IPAP: inspiratory positive airway pressure, EPAP: expiratory positive airway pressure; BiPAP: bilevel positive airway pressure, O2:Oxygen.

Table 4 Arterial blood gases parameters in compliant subjects, before treatment with non-invasive ventilationand 6–8 months after treatment.

Before treatment After treatment P value

pH 7.370.01 7.470.004 o0.05PaCO2 (mmHg) 79.478.1 49.174.6 0.05PaO2 (mmHg) 45.773.7 71.173.2 o0.05HCO3 (mmol/L) 40.572.9 27.170.8 o0.05

Respiratory failure in sleep disordered breathing 723

disorders. In the present study, we used fullpolysomnography due to the vicinity of the sleepdisorders center to the ICU. However, portablesleep diagnostic system may suffice. Sleep disor-dered breathing is overlooked as a cause of acuterespiratory failure.9,10 In obese patients with un-explained hypercapnic respiratory failure, sleepdisordered breathing should be considered as apossible cause. Ventilating this group of patientsvia non-invasive ventilation is safe and veryeffective in this setting.

Although the apnea hypopnea index, body massindex and age were comparable in studied subjectsand their matched comparison group, all but twosubjects had sleep hypoventilation syndrome andhypercapnia compared to none in the comparisongroup. Our findings contrast with Eugene et al.8

who reported obstructive sleep apnea syndromeonly in their group of apnea patients with acuterespiratory failure. However, later reports9,10 con-curred with our findings where high prevalence ofsleep hypoventilation syndrome was reported inpatients with sleep disordered breathing admitted

to the ICU with acute respiratory failure. There-fore, it seems that the severity of the apneahypopnea index may not be a good predictorof apnea patients who may develop respiratoryfailure. Based on our data and previous reports(Table 5), it seems that middle-aged obese patientswith sleep disordered breathing, hypercapnia andhypoxemia are at a higher risk of having acuterespiratory failure. Clinicians should be alert inmanaging apneic patients with hypoventilation,daytime hypercapnia and hypoxemia, as thesepatients may have a greater risk tendency todevelop respiratory failure. Our findings demon-strate that CPAP may not be effective in somepatients due to the persistent hypoventilation. Thisstresses the importance of performing polysomno-graphy to document the pattern of breathingand response to treatment. All studied subjectsbut one had a combination of obstructive sleepapnea and sleep hypoventilation syndrome. Thatexplains the high EPAP used in some cases. EPAP isneeded to prevent upper airway closure duringexpiration. Fibro-optic imaging studies in patients

ARTICLE IN PRESS

Table

5Com

parison

ofou

rfind

ings

withpreviou

slyreporteddata.

Num

ber

ofsubjects

Age

( SD)

BMI(SD)

AHI(S

D)

PH(SD)

PaCO2(S

D)

PaO2(S

D)

PSG

diagn

oses

(num

ber

ofsubjects)

Fletch

eret

al.8

857

41.6

9045

8245

AllOSA

SBuc

kleet

al.1

09

51.6

(14.7)

43.6

(14.8)

Nodata

7.3(0.03)

80.8

(8.6)

48.1

(6.8)

OHS+OSA

S(2),

OHS(2),

OSA

S(2),

CSA

(3)

Restaet

al.9

1457

.1(12.9)

42.7

(9.6)

25.9

7.28

(0.04)

65.9

(11.9)

42.1

(5.6)

OHS+OSA

S(5),

OHS(2),

OSA

S(5),

CSA

(2)

Presen

tstud

y11

53.3

(4.9)

47.7

(4.1)

45.4

(9.0)

7.29

(0.02)

77.9

(5.1)

48.3

(2.9)

SHVS+OSA

S(8),

SHVS(1),

OSA

S(2)

OSA

S:ob

structiveslee

pap

neasynd

rome;

OHS:

obesityhy

pov

entilation

synd

rome;

CSA

:ce

ntralslee

pap

nea;

SHVS:

slee

phy

pov

entilation

synd

rome;

BMI:bod

ymassindex

;AHI:ap

nea

hypop

neainde

x.

A. BaHammam et al.724

with obstructive sleep apnea have demonstratedthat upper airway narrowing begins prior toinspiration in sleeping humans.19

Subjects compliant to therapy had significantimprovement in their daytime PaCO2. Six of themdid not require daytime oxygen supplementation6–8 months after institution of nocturnal non-invasive ventilation treatment. Earlier studies usingtracheostomy and then CPAP led to the clearrecognition that treating obstructive sleep apneasyndrome may lead to improvement in hypercap-nia.20,21 Nocturnal non-invasive ventilation mayhelp normalize PaCO2 due to changes in respiratorydrive that occurs as a consequence of improvedupper airway mechanics.22 Sullivan et al., havepreviously reported that patients with obstructivesleep apnea and daytime hypercapnia have low orabsent hypoxic responsiveness.23 Moreover, in agroup of 13 obese patients with obstructive sleepapnea (half of them were hypercapnic), Garay etal.24 have observed a diminished ventilatory re-sponse to hypoxia and hypercapnia in patients whowere hypoxemic and hypercapnic, respectively.These results suggest that repeated apnea couldfavor a diminished ventilatory response to hypoxiaand hypercapnia.25 Several factors might havecontributed to the improvement in daytime PaO2.Theoretically, positive pressure ventilation mighthave improved lung mechanics by decreasing theincidence of atelectasis. Moreover, the reduction inpulmonary artery pressure and the daytime PaCO2

will result in improvement in daytime PaO2.Our study also demonstrated the safety and

feasibility of performing polysomnography in thesleep disorders center for ICU patients withclinical suspicion of sleep-disordered breathing.We described the protocol used to performpolysomnography in such patients. Performingpolysomnography for ICU patients with acuterespiratory failure and clinical suspicion of sleep-disordered breathing will allow a more accuratediagnosis and appropriate institution of manage-ment protocols. However, if performing full poly-somnography is not feasible, portable sleepdiagnostic system can be used.

This study has some limitations. The number ofstudied subjects is relatively small. However,previous studies reported small number of patientswith the largest reporting 14 patients only8–10

(Table 5). Another limitation is the unavailabilityof follow up data on all subjects as some subjectsdid not come for follow up.

In summary, this study showed that sleepdisordered breathing can be the cause of acuterespiratory failure in obese patients admitted tothe ICU. Intensivists and respirologists should have

ARTICLE IN PRESS

Respiratory failure in sleep disordered breathing 725

a high index of suspicion especially in obesepatients with hypercapnia. Our study also showsthat polysomnography is feasible and efficacious inacutely ill patients. Early polysomnography (orportable cardio-respiratory monitoring) allows ac-curate diagnosis and institution of appropriateventilation method, which may lead to avoidanceof mechanical ventilation and reduction in morbid-ity. Compliance to treatment results in improve-ment of daytime hypercapnia and hypoxemia.Further studies with a larger number of subjectsand sufficient records of follow up visits should bedone to supplement, explore and confirm ourobservations. Furthermore, we need more studieson the evolution of respiratory drive in patientswith sleep disordered breathing and hypercapniaunder therapy (non-invasive ventilation).

References

1. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. Theoccurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328:1230–5.

2. Nieto FJ, Young TB, Lind BK, Shahar E, Samet JM, Redline S,D’Agostino RB, Newman AB, Lebowitz MD, Pickering TG.Association of sleep-disordered breathing, sleep apnea, andhypertension in a large community-based study. Sleep HeartHealth Study. JAMA 2000;283:1829–36.

3. Young T, Peppard P. Sleep-disordered breathing and cardio-vascular disease: epidemiologic evidence for a relationship.Sleep 2000;23:S122–6.

4. Neau JP, Paquereau J, Meurice JC, Chavagnat JJ, Gil R.Stroke and sleep Apnea: cause or consequence? Sleep MedRev 2002;6:457–69.

5. Weitzenblum E, Krieger J, Apprill M, Vallee E, Ehrhart M,Ratomaharo J, Oewald M, Kurtz D. Daytime pulmonaryhypertension in patients with obstructive sleep apneasyndrome. Am Rev Respire Dis 1988;138:345–9.

6. He J, Kryger MH, Zorick FJ, Conway W, Roth T. Mortality andapnea index in obstructive sleep apnea: Experience in 385male patients. Chest 1988;94:9–14.

7. Miller A, Granada M. In-hospital mortality in the Pickwickiansyndrome. Am J Med 1975;56:144.

8. Fletcher EC, Shah A, Qian W, Miller CC. Near miss death inobstructive sleep apnea: a critical care syndrome. Crit CareMed 1991;19:1158–64.

9. Resta O, Guido P, Foschino Barbaro MP, Picca V, Talamo S,Lamorgese V. Sleep-related breathing disorders in acuterespiratory failure assisted by non-invasive ventilationtreatment: utility of portable polysomnographic system.Resp Med 2000;94:128–34.

10. Buckle P, Pouliot Z, Millar T, Kerr P, Kryger MH. Polysomno-graphy in acutely ill intensive care unit patients. Chest1992;102:288–91.

11. Loube DI, Gay PC, Strohl KP, Pack AI, White DP, Collop NA.Indications for positive airway pressure treatment of adultobstructive sleep apnea patients. A consensus Statement.Chest 1999;115:863–6.

12. Rechtschaffen A, Kales A, editors. A Manual of StandardizedTerminology, Techniques and Scoring System for SleepStages of Human Subjects. NIH Publication number 204.Washington: US Government Printing Office; 1968.

13. American Sleep Disorders Association. Atlas task Force, EEGarousals: Scoring rules and examples. Sleep 1992;15:174–84.

14. American Sleep Disorders Association. The Internationalclassification of sleep disorders, revised: Diagnostic andcoding manual. Rochester, Minnesota: American SleepDisorders Association; 1997.

15. Berger KI, Ayappa I, Chatr-amontri B, Marfatia A, Sorkin B,Rapoprt DM, Goldring RM. Obesity hypoventilation syndromeas a spectrum of respiratory disturbances during sleep.Chest 2001;120:1231–8.

16. American Academy of Sleep Medicine. Sleep-related breath-ing disorders in adults: recommendations for syndromedefinition and measurement techniques in clinical research.Sleep 1999;22:667–89.

17. Cirignotta F. Classification and definition of respiratorydisorders during sleep. Minerva Med 2004;95:177–85.

18. Jubber AS. Respiratory complications of obesity. Int J ClinPract 2004;58:573–80.

19. Borowiecki B, Pollak CP, Weitzman ED, Rakoff S, Imperato J.Fibro-optic study of pharyngeal airway during sleepin patients with hypersomnia obstructive sleep-apneasyndrome. Laryngoscope 1978;88:1310–3.

20. Rapport DM, Garay SM, Epstein S, Goldring RM. Hypercapniain the obstructive sleep apnea syndrome: a reevaluation ofthe Pickwickian syndrome. Chest 1986;89:627–35.

21. Cistulli PA, Sullivan CE. Pathophysiology of sleep apnea. In:Saunders NA, Sullivan C, editors. Sleep and Breathing. 2nded. New York: Marcel Dekker Inc; 1994. p. 405–48.

22. Berthon-Jones M, Sullivan CE. Time course of change inventilatory response to CO with long-term CPAP therapyfor obstructive sleep apnea. Am Rev Respir Dis 1987;135:144–7.

23. Sullivan CE, Parker S, Grunstein RR, Ho KY, Scheibner T.Ventilatory control in sleep apnea: a search for brainneurochemical defects. In: Issa FG, Suratt PM, RemmersJE, editors. Sleep and Respiration. New York: Wiley-Liss;1990. p. 325–36.

24. Garay SM, Rapoport D, Sorkin B, Epstein H, Feinberg I,Goldring RM. Regulation of ventilation in the obstructivesleep apnea syndrome. Am Rev Respir Dis 1981;124:451–7.

25. Weitzenblum E, Chaouat A, Kessler R, Oswald M, Apprill M,Krieger J. Daytime hypoventilation in obstructive sleepapnea syndrome. Sleep Med Rev 1999;3:79–93.