ORIGINAL ARTICLE

Is portable monitoring for diagnosing obstructive sleep apneasyndrome suitable in elderly population?

Jéssica Fabia Polese & Rogerio Santos-Silva &

Paula Martins de Oliveira Ferrari &Denis Eduardo Sartori & Sergio Tufik & Lia Bittencourt

Received: 21 March 2012 /Revised: 11 May 2012 /Accepted: 19 June 2012 /Published online: 4 July 2012# Springer-Verlag 2012

AbstractPurpose Obstructive sleep apnea syndrome (OSAS) is high-ly prevalent in the elderly. Unattended, at-home portablemonitoring (PM) is a diagnostic alternative to polysomnog-raphy in adults with high clinical probability of OSAS.However, no studies have evaluated the diagnostic accuracyof PM in elderly population. The aim of our study was toevaluate the effectiveness of PM in elderly patients.Methods We selected patients aged over 65 years with sus-pected OSAS. Two-order randomized evaluations were per-formed: one night of at-home PM (PMhome) and one nightof simultaneous PM and polysomnography (PSG) in thesleep lab (PSG+PM). We obtained three different apnea–hypopnea index (AHI): AHI from PSG (AHI PSG), AHIfrom at-home PM (AHI PMhome), and AHI from PM+PSG(AHI PM+PSG). Two technicians, blinded to the recordingorder, scored each sleep study.Results We studied a total of 43 patients. No differencebetween the AHI values for each of the different recordingswas found (p>0.05). There was good correlation betweenAHI PSG and AHI PMhome (r00.67) and AHI PSG+PM(r00.84). The area under the receiver operator curve wasabove 0.83, indicating good sensitivity and a positive pre-dictive value for AHI with cutoffs of 5, 15, and 30 and goodspecificity and negative predictive value for AHI valuesabove 15. Correlation, accuracy, and agreement were greaterwhen the recordings were made simultaneously.

Conclusions PM was effective for diagnosing OSAS in theelderly and can be used as an alternative to PSG in elderlypatients with a high clinical probability of OSAS.

Keywords Ambulatory monitoring . Aging .

Polysomnography . Obstructive sleep apnea

Introduction

During healthy aging process, physiological changes occur,including weakening of the pharyngeal muscles, reducedvital capacity, decreased thyroid function, weight gain, anddecreased respiratory control. These changes lead to a great-er likelihood of obstructive sleep apnea syndrome (OSAS)in older than younger adult [1–3]. The prevalence of OSASin the elderly ranges from 7 to 87 %, depending on thepopulation, inclusion criteria, and OSAS definition used[1–16]. Early diagnosis and intervention of OSAS, in theelderly population, provide a better quality of life, as well ascontrol of comorbidities.

Full-night in-lab polysomnography (PSG) is recognizedas the gold standard for diagnosing OSAS, but it is anexpensive and complex procedure [17]. In 2007, the Amer-ican Academy of Sleep Medicine (AASM) reviewed studiesevaluating the use of portable monitoring (PM) equipmentto diagnose OSAS, particularly unattended type 3 PM [18].The AASM recommends PM to diagnose OSAS in patientswith a high pretest probability of OSAS and without comor-bidities. Furthermore, the AASM taskforce recommenda-tions specifically apply to adult populations due to limiteddata on the use of PM in the pediatric and older populations[18]. In this context, the objective of this study was toevaluate the effectiveness of a type 3 PM (Stardust II®) as

J. F. Polese : R. Santos-Silva : P. M. de Oliveira Ferrari :D. E. Sartori : S. Tufik : L. Bittencourt (*)Disciplina de Medicina e Biologia do Sono, Departamento dePsicobiologia, Universidade Federal de São Paulo—UNIFESP,Rua Napoleao de Barros, 925,04024-002, São Paulo, Brazile-mail: lia.rita@unifesp.br

Sleep Breath (2013) 17:679–686DOI 10.1007/s11325-012-0742-y

an alternative for the diagnosis of OSAS in a randomized,prospective trial of patients over 65 years old with a highclinical probability.

Methods

From the patients referred to PSG in the Sleep Institute, weincluded those aged ≥65 years (the World Health Organiza-tion's definition of elderly), both genders, and with sus-pected OSAS, i.e., complaints of daytime sleepiness, loudsnoring, and apnea witnessed by a bed partner. We excludedpatients with a suspicion of other sleep disorders, those whohad previously undergone PSG or treated for OSAS,patients with severe or unstable medical illnesses, thosewho are on oxygen therapy, and those who are using hyp-notics, alcohol, or recreational drugs.

The research protocol was approved by the Ethics Com-mittee of the Universidade Federal de São Paulo (CEP 0478/08). All participants read and signed the consent form.

Protocol

All patients were evaluated for body mass index, neckcircumference, and blood pressure. The Berlin questionnaire[19] and the Epworth sleepiness scale [20] were used tocollect information about risk for OSAS and sleepiness,respectively.

The patients underwent two sleep assessments: one wasperformed with PM at home (PMhome), without technicalsupervision and the other was performed with PM simulta-neously with PSG in the Sleep Institute (PM+PSG). Bothevaluations were conducted in the same 1-week period, andthe recording order (PMhome or PM+PSG) was randomlydetermined. We derived three different apnea–hypopneaindex (AHI) values from the recordings: AHI from PSG(AHI PSG), AHI from PMhome (AHI PMhome), and AHIfrom PM used simultaneously with PSG (AHI PM+PSG).

After the patients underwent PMhome monitoring, theywere asked to complete an analog scale assessing the fol-lowing criteria: difficulty with assembling the equipment,operation difficulties, and comfort while using it. The scoresranged from 0 (greatest difficulty and discomfort) to 10(least difficulty and discomfort).

Polysomnography

Full-night PSG (Embla® S7000, Embla Systems, Inc.,Broomfield, CO, USA) was performed by a trained techni-cian. The PSG montage included electroencephalogram,electrooculogram, electromyogram (submental region andbilateral anterior tibialis muscle), airflow (nasal pressureand thermistor), respiratory effort of thorax and abdomen

(inductance plethysmography), oxyhemoglobin saturation(SpO2), snoring, body position, and video monitoring.

Portable monitoring

The type 3 portable device used was the Stardust II® (PhilipsRespironics, Inc., Murrysville, PA, USA), which has beenshown to be capable of diagnosing OSAS in a non-elderlypopulation. The Stardust records include SpO2 (finger sen-sor), heart rate (finger sensor), airflow (nasal pressure),respiratory effort (belt with piezoelectric sensor set at thelower sternum), and body position (device positioned at thelower sternum). Data are collected and stored on internalmemory in the device. The data are then downloaded to acomputer for automated analysis by the host software(Stardust Host Software, Philips Respironics, Inc., USA).

A trained PSG technician applied the PM and sensorsused for the PSG recording during the overnight PSG in thesleep lab. A research assistant instructed the patients how touse PM at home. The explanation included verbal andwritten instructions to illustrate the correct hook-up of thePM and included diagrams and a brief practical demonstra-tion. During training, patients were asked to indicate “lightsout” and “lights on” and any time during the night he/sheremained awake for more than 15 min.

Scoring procedures

All PM and PSG recordings were manually scored by twodifferent PSG technologists who were blinded to the record-ing order and patient medical information. For PSG scoring,sleep stages were determined by standard Rechtschaffen andKales criteria [21]. The respiratory event scoring was basedon AASM criteria [22]. Apnea was defined as a completecessation of airflow during sleep lasting ≥10 s (central,obstructive, or mixed). Hypopnea was defined by (1) a cleardecrease (>50 %) from baseline in the amplitude of the nasalcannula during sleep or (2) a clear amplitude reduction ofthe nasal cannula during sleep (<50 %) but associated witheither an oxygen desaturation of >3 % or an arousal. Thehypopnea event lasts 10 s or longer. Arousals [23] andperiodic limb movement (PLM) [24] were quantifiedaccording to the American Sleep Disorders AssociationTask Force. The following parameters were obtained fromthe PSG: total recording time (TRT), total sleep time (TST),sleep latency, sleep efficiency, sleep stage distribution,arousals, awakenings, AHI (calculated using the TST asdenominator), and SpO2.

For PM scoring, apnea events were required to show anairflow cessation ≥10 s (central, obstructive, or mixed).Hypopnea was defined as a decrease (>50 %) from baselinein the amplitude of the nasal cannula during sleep or (2) aclear amplitude reduction of the nasal cannula during sleep

680 Sleep Breath (2013) 17:679–686

(<50 %) but associated with an oxygen desaturation of>3 %. The hypopnea event lasts 10 s or longer. TRT, AHI(calculated using the TRT as denominator), and SpO2 wereanalyzed from PM recordings.

Statistical analysis

The demographic variables and dropout values are pre-sented as descriptive statistics. The variables from the PSGwere compared with the PM+PSG variables obtained duringsimultaneous recording and with the PMhome variables,using the ANOVA test with repeated measures. Post hocanalysis was performed with the Bonferroni comparisontest. To determine significance, p values <0.05 were re-quired. The rates of diagnostic agreement, overestimation,and underestimation were calculated by comparing the AHIfrom PSG to the AHI from PM+PSG and the AHI from PSGto the AHI from PMhome, according to the criteria shown inTable 1. The kappa coefficient was calculated to indicate theagreement between both diagnostic methods. We evaluatedthe Pearson's correlation coefficient between the diagnosticmethods. Sensitivity, specificity, negative predictive value(NPV), and positive predictive value (PPV) at cutoffs of 5,15, and 30 events per hour were calculated using the AHIvalues from PSG vs. PM+PSG and PSG vs. PMhome.Using the same series of comparisons, receiver operatorcurves (ROCs) were constructed to illustrate true andfalse-positive results, with AHI cutoff values of 5, 15, and30 events per hour. Bland–Altman plots were generated toassess the agreement between the PSG and PM results.

Results

We contacted 189 elderly patients referred to perform the PSGin the sleep lab. A total of 44 met the inclusion criteria. Onepatient refused to participate. A total of 43 patients wereevaluated, with a mean age 70±5 years and a similar propor-tion of men (n019) and women (n024). The demographicdata of the evaluated population are presented in Table 2.

From the total PM recordings (PMhome plus PM+PSG;n086), we observed complete loss of data in 10.5 % ofrecordings. In the PM+PSG group, there was complete lossof record in five patients (11.6 %) due to failure to downloadthree recordings, battery problems during one recording,and complete oximetry record loss for one patient. In thePMhome group, there was complete record loss for fourpatients (9.3 %), three because of failure to download, andone because of complete loss of oximetry data.

Partial data losses were observed in 31 % of the record-ings. The data from these recordings were included in theanalysis because more than 75 % of each recording wasacceptable. Partial data losses in the PM+PSG group totaled18 %. Momentary pulse oximetry loss was observed in sixrecordings and airflow loss in two recordings. In thePMhome group, partial data loss was observed in 44 % ofthe recordings: six recordings showed a partial loss of pulseoximetry, airflow signal loss in eight recordings, and chestsignal band loss in five patients. The analysis of the analogscale of discomfort, difficulty with assembly, and difficultyusing the analog scale showed the following averages: 8.25,8.2, and 8.5, respectively.

A total of 93 % of the patients showed an AHI >5, and72 % showed an AHI >30. The mean of the PLM index wasobserved to be >15 in 23.2 % of the patients. Significantdifferences were observed in the TRT, which was higher inPMhome (p00.004), and in the SpO2 nadir, which waslower in PMhome (p00.04). There were no statisticallysignificant differences in other sleep parameters (Table 3).The mean value of AHI did not show difference betweenrecordings in PSG, PM+PSG, and PMhome, respectively(32.7±25.7; 33.7±18.8, and 29.1±16.4; p00.541). Thecomparative diagnostic of the AHI values from both PM+PSG and PMhome showed good agreement (79 and 81 %),overestimation (18 and 14 %), and underestimation (3 and5 %) compared with PSG (Table 4).

By evaluating the kappa coefficient of agreement be-tween the AHIs obtained from the three recordings (Table 5),we observed moderate to good agreement (kappa ≥0.48)

Table 1 Diagnostic and agreement/disagreement criteria for AHIobtained from PSG compared with AHI from PM (PMhome and PM+PSG)

AHI PSG AHI PM Diagnostic classification

≥30 ≥30 Agreement

<30 PSG AHI ±10 or less Agreement

<30 PSG AHI +>10 Overestimation

<30 PSG AHI −>10 Underestimation

AHI apnea–hypopnea index, PSG polysomnography, PM portablemonitoring

Table 2 Demographics of the 43 participants evaluated (mean±SD)

Age, years 70.0±5.0

Gender ratio (M/F), % 44:56

BMI, kg/m2 30.3±6.0

Neck circumference, cm 38.9±3.4

Systolic blood pressure 140.9±24.0

Diastolic blood pressure 82.9±11.8

ESS 9.1±6.1

Berlin risk (low/high), % 81:19

M male, F female, BMI body mass index, ESS Epworth sleepinessscale

Sleep Breath (2013) 17:679–686 681

with the PM with PSG for all cutoff values of severity. Thiswas true whether PM was performed simultaneously or athome, except at the cutoff value of AHI <5 (kappa00.36)and the cutoff values between 5 and 15 when the PM wasperformed at home compared with PSG (kappa00.32),which had weak agreement. We observed good correlationsfor the AHI values obtained in all three conditions with thestrongest found for the simultaneous recording: for the AHIfrom PSG vs. AHI from PM+PSG (r00.84); for AHI fromPSG vs. AHI from PMhome (r00.67); and for AHIfrom PM+PSG vs. PMhome (r00.74) (p<0.001)(Fig. 1).

The sensitivity, specificity, PPV, and NPV, for cutoffs of5, 15, and 30, showed good results. However, there wasgood specificity (above 60) only for AHI values greater than15. The NPV for all cutoff levels was zero or very low(Table 6). The area under the ROC curves of PM vs. PSG(cutoffs of 5, 15, and 30) were greater than 0.83 for allcomparisons, with the simultaneous recordings showingthe largest area (Fig. 2).

The Bland–Altman visual analysis is based on the rela-tion between the mean of two individual values and thedifference between these values. By calculating the mean

Table 3 Comparison of polysomnographic parameters observed inPSG, PM+PSG, and PMhome recordings (mean±SD)

PSG PM+PSG PMhome p

TRT, min 430.9±47.3* 454±57.2 491.3±80.4* 0.004

AHI 32.7±25.7 33.7±18.8 29.1±16.4 0.541

Apnea index 15.7±21.2 13.9±15.4 13±12 0.653

Hypopnea index 16.4±11.4 18.3±9.4 15.3±9.8 0.798

Central apnea index 0.5±2.0 1.3±3.3 0.7±3 0.144

Mixed apnea index 0.3±1.4 0.4±0.9 0.16±0.2 0.622

Mean SpO2, % 93.1±2.3 93.6±2.1 93.3±2.5 0.711

Desaturation index 24.0±23.0 25.5±20.6 26.1±20.0 0.247

Time below90 %, min

28.3±39.2 29.4±47.4 33.1±47.4 0.252

SpO2 nadir, % 81.5±7.7** 78.2±9.3 74.0±14.6** 0.001

Desaturation index 0 number of 3 % drops in SpO2 per hour of sleep.ANOVA variance test

TRT total recording time, PSG polysomnography, PM portable monitor-ing, AHI apnea–hypopnea index, SpO2 oxyhemoglobin saturation

*p00.0038, **p00.042

Table 4 Diagnostic agreement, overestimation, and underestimationbetween AHI from PSG vs. AHI from PM+PSG and PMhome

PSG vs. PM+PSG PSG vs. PMhome

Diagnostic agreement (%) 79 81

Overestimation of AHI (%) 18 14

Underestimation of AHI (%) 3 5

AHI apnea–hypopnea index, PSG polysomnography, PM portablemonitoring

Table 5 Kappa coefficient for AHI values from PSG vs. AHI valuesfrom PM+PSG and PMhome, according to severity

PM+PSG PMhome

AHI <5 0.072 0.036

AHI 5–15 0.62 0.32

AHI 15–30 0.58 0.56

AHI >30 0.67 0.48

AHI apnea–hypopnea index, PSG polysomnography, PM portablemonitoring

Fig. 1 Pearson's correlation between apnea–hypopnea index (AHI)from portable monitoring (PM) and polysomnography (PSG) in the threerecordings. a PSG vs. PM+PSG (r00.84; p<0.001); b PSG vs. PMhome(r00.67; p<0.001); c PM+PSG vs. PMhome (r00.74; p<0.001)

682 Sleep Breath (2013) 17:679–686

and the standard deviation of these differences, it is possibleto calculate the confidence limits for these differences.These analyses showed good agreement between the AHIvalues from PM and PSG (Fig. 3).

Discussion

Our results demonstrated that the Stardust II® PMshowed good agreement, correlation, and accuracy com-pared with PSG for diagnosing OSAS in the elderly. Asfar as we know, this is the first study which attemptedto investigate the use of a PM in a special population:the elderly.

Stardust II® PM technology meets AASM recommen-dations [18], which state that a type 3 PM must recordat least airflow, respiratory effort, heart rate, and bloodoxygenation. The validation of the specific use of theStardust II® PM in adults with suspected OSAS was

performed by our group in a previous study where weobserved good agreement, correlation, and acceptablesensitivity and specificity [25]. The Stardust II® PMhas been also used in other studies. Yin and colleaguesevaluated the Stardust II® PM evaluating only the dis-comfort, difficulty of applying the various sensors, andoperation of the device [26]. In that study, the authorsdid not compare PM results with PSG, but they didevaluate the automated and manual scoring of respirato-ry events. They suggested that the scoring should beperformed manually and that the discomfort score waslow. Our study showed that elderly patients felt com-fortable using the device and found it easy to assemble.In another study, Yin and colleagues evaluated the reli-ability of the Stardust II® PM in OSAS patients com-pared with PSG [27]. The sensitivity, specificity, NPV,and PPV were similar to our findings. Nevertheless,they performed PSG for only a portion of their patientpopulation and they did not compare the results withsimultaneous PSG. Other studies using different type 3PM in adults without comorbidities have demonstratedeffectiveness in diagnosing OSAS compared to PSG[28–32].

The loss of data in our study (10.5 %) was consistent withthe previously described (3 to 18 %) [33]. In another studyconducted by our group that also used the Stardust II® PM,we reported data loss of only 2 % due to inaccessible datadownloads from at-home PM [25]. We reported that partialdata loss caused by oximeter problems was 10 % lower thanin other studies. The present study showed a greater numberof partial data losses in the PMhome group than the PM+PSG group. This may be due to challenges in assemblingthe device in this population. However, our findings suggestthat PM may be comfortably and safely used by elderly

Table 6 Sensitivity, specificity, and positive (PPV) and negative(NPV) predictive value for different AHI cutoffs from PMhome andPM+PSG vs. AHI from PSG

Sensitivity Specificity PPV NPV

AHI ≥5 PMhome 90 30 90 60

PM+PSG 100 0 1 0

AHI ≥15 PMhome 80 60 88 45

PM+PSG 100 70 90 0

AHI ≥30 PMhome 80 80 70 15

PM+PSG 90 68 71 0.1

AHI apnea–hypopnea index, PSG polysomnography, PM portablemonitoring

Fig. 2 ROC of different apnea–hypopnea index (AHI) cutoffs forpolysomnography (PSG) and portable monitoring (PM), consideringthe three recordings. a AHI cutoff 5: PSG vs. PM+PSG (AUC00.93)and PSG vs. PMhome (AUC00.83); b AHI cutoff 15 and AHI cutoff

5: PSG vs. PM+PSG (AUC00.99) and PSG vs. PMhome (AUC00.85); c AHI cutoff 30: PSG vs. PM+PSG (AUC00.90) and PSG vs.PMhome (AUC00.85)

Sleep Breath (2013) 17:679–686 683

patients. Moreover, the partial losses, although significant,do not invalidate the entire test. More detailed guidelines onthe installation and use of the equipment could reduce theselosses.

As recommended by the AASM, our study includedpatients with a high clinical probability of OSAS. We ob-served that 93 % of the elderly patients had an AHI >5 andthat more than 70 % of these patients had severe OSASwhich corroborated the importance of the diagnostic inves-tigation in this population. In the elderly, structural abnor-malities and chest wall rigidity, weight gain, and reducedventilatory control [34–39] are associated with a higher

prevalence of OSAS (7 to 87 %) [1–16], but the clinicalsignificance of OSAS in this population remains uncertain.Studies suggest that OSAS in the elderly, compared withmiddle-aged adults, may have a distinct impact on clinicaloutcomes and mortality [9, 40]. Considering the observa-tional nature of these studies without randomized interven-tions, future studies need to evaluate the exact impact ofOSAS in the elderly for early diagnosis and treatment of thedisease.

The comparison of three different recordings showedthat patients had higher TRT in PMhome compared withPSG. This could be explained by the patients' autonomyin turning on the recording device. Moreover, the SpO2

nadir was lower in the PMhome group which may beexplained either by the use of different devices andoximeter technology (Philips Respironics vs. Medcare)or by the artifact impacting on the SpO2 values forPMhome.

The reliability of PM compared with PSG was good, andthe kappa coefficient showed a better correlation when therecordings were performed simultaneously. We also ob-served that correlation, sensitivity, PPV, and area under theROC curve were better when PM and PSG were conductedsimultaneously. This could be explained by the variability ofsleep parameters, including AHI, on different nights [41,42]. However, we found that the NPVs were zero or verylow at all threshold levels for both PM recordings. This canbe explained by the fact that our study included elderlypatients with a high clinical probability of OSAS. Therefore,when a small number of individuals without OSAS wereevaluated, it resulted in rare cases in which the PM record-ings presented false-negative results.

One limitation of the use of PM in elderly patientsinvolves the fact that this type of assessment does notallow the identification of PLM. Our study showed ahigh rate of PLM assessed by PSG, in agreement withthe prevalence described in the literature for the olderaged group [14, 43].

In our study, although we contacted 189 elderly patientswho were referred to perform the PSG in the sleep lab, onlya relatively small number of patients (N043) were analyzed.Many were excluded mainly due to the presence of severe orunstable medical illnesses. In addition, some patients re-fused to participate in the study because of the difficulty inattending to the sleep laboratory and the need for an accom-panying person.

In summary, OSAS is prevalent in the elderly popu-lation. Early detection could improve quality of life andreduce morbidity in this group. The Stardust II, a type 3PM, provided effective and good diagnostic agreementwith attended PSG in elderly population and couldprovide an affordable and comfortable alternative toin-laboratory PSG.

Fig. 3 Bland–Altman analysis of AHIs from PM and PSG, consider-ing the three recordings. a PSG vs. PM+PSG; b PSG vs. PMhome; cPM+PSG vs. PMhome

684 Sleep Breath (2013) 17:679–686

Acknowledgments This work was supported by grants from theAssociaçao Fundo de Incentivo a Pesquisa and Fundaçao de Amparoa Pesquisa do Estado de Sao Paulo (# 07/50525-1 to RS-S and CEPIDno. 98/14303-3 to ST). ST and LRAB received the Conselho Nacionalde Desenvolvimento Científico e Tecnológico (CNPq) fellowship. Wewould like to thank Philips, who provided the Stardust II® and trainingon the operation of the devices and the host software.

References

1. Bliwise DL (2009) Epidemiology of age-dependence in sleepdisordered breathing (SDB) in old age: the bay area sleep cohort(basc). Sleep Med Clin 4:57–64

2. Phillips BA, Berry DT, Lipke-Molby TC (1996) Sleep-disorderedbreathing in healthy aged persons. Fifth and final year follow-up.Chest 110:654–658

3. Phillips BA, Ancoli-Israel S (2001) Sleep disorders in the elderly.Sleep Med 2:99–114

4. Ancoli-Israel S, Kripke DF, Klauber MR, Mason WJ, Fell R,Kaplan O (1991) Sleep-disordered breathing in community-dwelling elderly. Sleep 14:486–495

5. Bixler EO, Vgontzas AN, Ten Have T, Tyson K, Kales A (1998)Effects of age on sleep apnea in men: I. Prevalence and severity.Am J Respir Crit Care Med 157:144–148

6. Bixler EO, Vgontzas AN, Lin HM, Ten Have T, Rein J, Vela-BuenoA, Kales A (2001) Prevalence of sleep-disordered breathing in wom-en: effects of gender. Am J Respir Crit Care Med 163:608–613

7. Durán J, Esnaola S, Rubio R, Iztueta A (2001) Obstructive sleepapnea-hypopnea and related clinical features in a population-basedsample of subjects aged 30 to 70 year. Am J Respir Crit Care Med163:685–689

8. Young T, Shahar E, Nieto FJ, Redline S, Newman AB, GottliebDJ, Walsleben JA, Finn L, Enright P, Samet JM, Sleep HeartHealth Study Research Group (2002) Predictors of sleep-disordered breathing in community-dwelling adults: the SleepHeart Health Study. Arch Intern Med 162:893–900

9. Punjabi NM, Caffo BS, Goodwin JL, Gottlieb DJ, Newman AB,O’Connor GT, Rapoport DM, Redline S, Resnick HE, Robbins JA,Shahar E, Unruh ML, Samet JM (2009) Sleep-disordered breath-ing and mortality: a prospective cohort study. PLoS Med 6:e1000132

10. Carskadon MA, Dement WC (1981) Respiration during sleep inthe aged human. J Gerontol 36:420–423

11. Coleman RM, Miles LE, Guilleminault CC, Zarcone VP Jr, vanden Hoed J, Dement WC (1981) Sleep-wake disorders in theelderly: polysomnographic analysis. J Am Geriatr Soc 29:289–296

12. Roehrs T, Zorick F, Sicklesteel J, Wittig R, Roth T (1983) Age-related sleep-wake disorders at a sleep disorder center. J AmGeriatr Soc 31:364–370

13. Yesavage J, Bliwise D, Guilleminault C, Carskadon M, Dement W(1985) Preliminary communication: intellectual deficit and sleep-related respiratory disturbance in the elderly. Sleep 8:30–33

14. Ancoli-Israel S, Kripke DF, Mason W, Kaplan OJ (1985) Sleepapnea and periodic movements in an aging sample. J Gerontol40:419–425

15. Mosko SS, Dickel MJ, Paul T, LaTour T, Dhillon S, Ghanim A,Sassin JF (1988) Sleep apnea and sleep-related periodic leg move-ments in community resident seniors. J Am Geriatr Soc 36:502–508

16. Tufik S, Santos-Silva R, Taddei J, Bittencourt LA (2010) Obstruc-tive sleep apnea syndrome in the Sao Paulo Epidemiologic SleepStudy. Sleep Med 11:441–446

17. American Sleep Disorders Association (1994) Practice parametersfor the use of portable recording in the assessment of obstructive

sleep apnea. Standards of Practice Committee of the AmericanSleep Disorders Association. Sleep 17:372–377

18. Collop NA, Anderson WM, Boehlecke B, Claman D, Goldberg R,Gottlieb DJ, Hudgel D, Sateia M, Schwab R, Portable MonitoringTask Force of the American Academy of Sleep Medicine (2007)Clinical guidelines for the use of unattended portable monitors inthe diagnosis of obstructive sleep apnea in adult patients. J ClinSleep Med 3:737–747

19. Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP (1999)Using the Berlin Questionnaire to identify patients at risk for thesleep apnea syndrome. Ann Intern Med 131:485–491

20. Johns MW (1991) A new method for measuring daytime sleepi-ness: the Epworth sleepiness scale. Sleep 14:540–545

21. Rechtschaffen A, Kales A (1968) A manual of standardized termi-nology, techniques, and scoring systems for sleep stages of humansubjects. Brain Information/Brain Research Institute UCLA, LosAngeles

22. American Academy of Sleep Medicine (1999) Sleep-relatedbreathing disorders in adults: recommendations for syndrome def-inition and measurement techniques in clinical research. The reportof an American Academy of Sleep Medicine Task Force. Sleep22:667–689

23. No authors listed (1992) EEG arousals: scoring rules and exam-ples: a preliminary report from the Sleep Disorders Atlas TaskForce of the American Sleep Disorders Association. Sleep15:174–182

24. American Sleep Disorders Association (1993) Recording and scor-ing leg movements—Atlas Task Force. Sleep 16:748–759

25. Santos-Silva R, Sartori DE, Truksinas V, Truksinas E, Alonso FF,Tufik S, Bittencourt LRA (2009) Validation of a portable monitor-ing system for the diagnosis of obstructive sleep apnea syndrome.Sleep 32:629–636

26. Yin M, Miyazaki S, Itasaka Y, Shibata Y, Abe T, Miyoshi A,Ishikawa K, Togawa K (2005) A preliminary study on applicationof portable monitoring for diagnosis of obstructive sleep apnea.Auris Nasus Larynx 32:151–156

27. Yin M, Miyazaki S, Ishikawa K (2006) Evaluation of type 3portable monitoring in unattended home setting for suspected sleepapnea: factors that may affect its accuracy. Otolaryngol Head NeckSurg 134:204–209

28. White D, Gibb T, Wall J, Westbrook PR (1995) Assessment ofaccuracy and analysis time of a novel device to monitor sleep andbreathing in the home. Sleep 18:115–126

29. Tonelli de Oliveira AC, Martinez D, Vasconcelos LF, Gonçalves SC,Lenz MC, Fuchs SC, Gus M, Abreu-Silva EO, Moreira LB, FuchsFD (2009) Diagnosis of obstructive sleep apnea syndrome and itsoutcomes with home portable monitoring. Chest 135:330–336

30. To KW, Chan WC, Chan TO, Tung A, Ngai J, Ng S, Choo KL, HuiDS (2009) Validation study of a portable monitoring device foridentifying OSA in a symptomatic patient population. Respirology14:270–275

31. Dingli K, Coleman EL, Vennelle M, Finch SP, Wraith PK, MackayTW, Douglas NJ (2003) Evaluation of a portable device for diag-nosing the sleep apnoea/hypopnoea syndrome. Eur Respir J21:253–259

32. Nigro CA, Dibur E, Malnis S, Grandval S, Nogueira F (2012)Validation of ApneaLink Ox™ for the diagnosis of obstructivesleep apnea. Sleep Breath. doi:10.1007/s11325-012-0684-4

33. Flemons WW, Littner MR, Rowley JA, Gay P, Anderson WM,Hudgel DW, McEvoy RD, Loube DI (2003) Home diagnosis ofsleep apnea: a systematic review of the literature. An evidencereview cosponsored by the American Academy of Sleep Medicine,the American College of Chest Physicians, and the AmericanThoracic Society. Chest 124:1543–1579

34. Chan ED, Welsh CH (1998) Geriatric respiratory medicine. Chest114:1704–1733

Sleep Breath (2013) 17:679–686 685

35. Kronenberg RS, Drage CW, Ponto RA, Williams LE (1973) Theeffect of age on the distribution of ventilation and perfusion in thelung. Am Rev Respir Dis 108:576–586

36. Peterson DD, Pack AI, Silage DA, Fishman AP (1981) Effects ofaging on ventilatory and occlusion pressure responses to hypoxiaand hypercapnia. Am Rev Respir Dis 124:387–391

37. Brischetto MJ, Millman RP, Peterson DD, Silage DA, Pack AI(1984) Effect of aging on ventilatory response to exercise andCO2. J Appl Physiol 56:1143–1150

38. Janssens JP, Pautex S, Hilleret H, Michel JP (2000) Sleep disor-dered breathing in the elderly. Aging 12:417–429

39. Tack M, Altose MD, Cherniack NS (1981) Effect of aging onrespiratory sensations produced by elastic loads. J Appl Physiol50:844–850

40. Lavie P, Lavie L (2009) Unexpected survival advantage inelderly people with moderate sleep apnoea. J Sleep Res18:397–403

41. Bittencourt LR, Suchecki D, Tufik S, Peres C, Togeiro SM, BagnatoMC, Nery LE (2001) The variability of the apnoea-hypopnoea index.J Sleep Res 10:245–251

42. Maestri R, La Rovere MT, Robbi E, Pinna GD (2001) Night-to-night repeatability of measurements of nocturnal breathing disor-ders in clinically stable chronic heart failure patients. Sleep Breath15(4):673–678

43. Hirshkowitz M, Moore CA, Hamilton CR 3rd, Rando KC,Karacan I (1992) Polysomnography of adults and elderly:sleep architecture, respiration, and leg movement. J Clin Neu-rophysiol 9:56–62

686 Sleep Breath (2013) 17:679–686