| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From St. Mary’s Hospital (Dr. D. Patz), Grand Junction, CO; Mobile Sleep Diagnostics (Mr. Spoon and Mr. Corbin), Grand Junction, CO; Department of Biology (Mr. M. Patz), University of Colorado, Boulder, CO; Department of Biostatistics (Mr. Swihart), University of Colorado Health Sciences Center, Denver, CO; and Sleep Disorders Laboratory (Ms. Dover) and Department of Medicine (Dr. White), Brigham and Women’s Hospital, Boston, MA.
Correspondence to: David S. Patz, MD, FCCP, St. Mary’s Hospital, Box 1628, Grand Junction, CO 81502; e-mail: npatz{at}bresnan.net
Abstract
Background: The present requirement for "at facility" polysomnograms requires many residents in mountain communities to descend in elevation for sleep testing, which may cause misleading results regarding the severity of obstructive sleep apnea (OSA).
Design: Eleven patients with previously undiagnosed sleep apnea living at an altitude > 2,400 m (7,900 feet) in Colorado underwent diagnostic sleep studies at their home elevation and at 1,370 m (4,500 feet), and 5 of the 11 patients were also studied at sea level.
Results: The mean (SE) apnea-hypopnea index (AHI) fell from 49.1 (10.5)/h to 37.0 (11.2)/h on descent to 1,370 m (p = 0.022). In the five patients who traveled to sea level, the AHI dropped from 53.8 (13.2)/h at home elevation to 47.1 (14.8)/h at 1,370 m, and to 33.1 (12.6)/h at sea level (p = 0.018). The reduction in AHI was predominantly a reduction in hypopneas and central apneas, with little change in the frequency of obstructive apneas. Duration of the obstructive apneas lengthened with descent. Of eight patients with an AHI < 50/h at their home elevation, two patients had their AHI fall to < 5/h at 1,370 m, and a third patient dropped to < 5/h at sea level, ie, below many physicians’ threshold for providing therapy. Patients with the most severe OSA had the least improvement with descent.
Conclusions: Because AHI decreases significantly with descent in altitude, polysomnography is most accurately done at the home elevation of the patient. Descent to a sleep laboratory at a lower elevation may yield false-negative results in patients with mild or moderate sleep apnea.
Key Words: altitude • evaluation • obstructive sleep apnea
In mountainous areas such as Colorado, the nearest sleep laboratory for a resident may often be in an urban center at a substantially lower elevation. Presently, patients who live 1,000 to 1,500 m in elevation above sleep laboratories in Grand Junction, CO, at 1,370 m (4,500 feet), or Denver, CO, at 1,585 m (5,200 feet) may need to utilize services in these cities because Medicare and most insurance companies will not authorize "at-home" polysomnography. Three atmospheric variables change with descent: (1) air density and thus viscosity through a critically narrowed upper airway, (2) oxygen content, and (3) barometric pressure or external compressive effect. Changes in these variables with descent raise the concern that we may be misestimating the severity of sleep apnea and depriving some patients of necessary therapy if they descend to a sleep laboratory for evaluation.
In this study, we evaluated patients with symptoms of obstructive sleep apnea (OSA) residing at high elevation and assessed the change in sleep apnea severity during polysomnography, comparing home elevation findings with sleep study results at one or two lower elevations. Our goal was to understand the consequences of bringing these patients down from their home elevation to metropolitan centers in Colorado, or even further down to tertiary referral centers near sea level for sleep evaluation.
Materials and Methods
Study Design and Enrollment
Eleven individuals with suspected sleep apnea living at an altitude > 2,400 m in western Colorado underwent attended full-night diagnostic polysomnography at two elevations, one in the home and one in a clinical sleep laboratory in Grand Junction at 1,370 m. Five of the 11 patients also underwent a third polysomnogram in a clinical sleep laboratory at sea level. Each patient was referred from a primary practitioner who was aware of the study protocol and was concerned about the possibility of OSA. All patients provided informed consent for the protocol, which was approved by the St. Mary’s Hospital Investigational Review Board. The Grand Junction studies and the sea level studies were performed in clinical sleep laboratories. Alice IV sleep recording equipment (Respironics; Murrysville, PA) was used in 25 of 27 studies; the two other studies utilized Sandman (Pleasanton, CA) and Grass-Telefactor (West Warwick, RI) equipment. The median time between the two Colorado studies was 3 days (range, 2 to 34 days). Six of the 11 patients underwent the home study first, with the other 5 patients being first studied in Grand Junction. In all five of the sea level travelers, the sea level study was the third study. It followed the last Colorado study by a median of 7 days (range, 4 to 61 days).
Polysomnography
Sleep was recorded with four EEG leads, chin electromyogram, and bilateral electrooculogram. Bilateral leg electromyogram position and ECG were also recorded. Respiration was monitored with chest and abdominal belts with piezo strain gauges, pulse oximetry, snoring, and nasal and oral airflow monitors in each patient. In 22 of the 27 studies, the patients had nasal/oral thermistors; but in 5 of the Colorado studies, the thermistor was only an oral thermistor, recording in parallel with the nasal pressure airflow transducer. In all the Colorado studies, a nasal pressure airflow transducer (Pro-Tech; Mukilteo, WA) was used. This was available in only two of the five sea level studies.
Sleep was staged according to Rechtschaffen and Kales.1 Apnea required 10-s duration, a flat nasal pressure signal (when available), and at least 80% reduction in the thermistor signal. Hypopneas required a detectable reduction in either a flow signal or respiratory belt for 10 s, associated with 4% oxygen desaturation. The 4% desaturation criteria for hypopneas was utilized because Medicare and the largest private insurance provider in western Colorado require this definition to decide who qualifies for insurance payment for OSA therapy. Respiratory effort-related arousals (RERAs) were defined as snoring periods or hypopneas lasting 
10 s, with < 4% oxygen desaturation but leading to EEG arousal. The apnea-hypopnea index (AHI) was defined as the number of apneas plus hypopneas per hour of sleep. The respiratory disturbance index (RDI) was defined as the number of apneas plus hypopneas plus RERAs per hour of sleep.
The 25 studies using Alice IV equipment were scored by a research sleep technician in Boston who was blinded to the study design. They were edited by an American Board of Sleep Medicine diplomat who also was blinded to their identity. The studies using Grass-Telefactor and Sandman equipment were not accessible to blinding.
Statistical Analysis
Each variable was analyzed using a repeated-measure analysis of variance (SAS PROC MIXED; SAS Institute; Cary, NC), with CONTRAST statements facilitating pairwise comparisons in the post hoc analysis. The data were analyzed in two ways. All 11 patients were analyzed at two elevations, and the 5 patients who traveled to sea level were analyzed at three elevations. Data are expressed as mean value (SE). Demographics are expressed as mean ± 1 SD.
Results
Table 1 summarizes the individual patient data, and Table 2 summarizes how the respiratory indexes (mean [SE]) changed with altitude descent. The patients were 46 to 70 years old (mean 55 ± 7.3 years). Three of the 11 patients were women. Body mass index was 20 to 58 kg/m2 (mean, 33.9 ± 11.3 kg/m2). The home elevations ranged from 2,417 to 3,139 m (7,930 to 10,300 feet) [mean, 2,701 ± 214 m]. Home elevation severity of sleep apnea in these patients ranged from mild to severe, with AHIs from 13.5 to 125.6/h. In these 11 patients, there was a trend of reduction in AHI with descent, with AHI of 49.1 (10.5)/h at home elevations > 2,400 m, and an AHI of 37.0 (11.2)/h in Grand Junction at 1,370 m (p = 0.022). In the 5 of 11 patients who also traveled to sea level, AHI decreased from 53.8 (13.2)/h above 2,400 m to 47.1 (14.8)/h at 1,370 m, and 33.1 (12.6)/h at sea level (p = 0.018) [Fig 1 , left].
|
|
|
The values for the indexes for individual respiratory event types as altitude changed from home elevation to Grand Junction in all 11 patients are shown in Figure 2 , left. The obstructive apneas and mixed apneas are grouped together and did not decrease. The central apneas decreased by 70% (p = 0.06), and the hypopneas decreased by 49% (p = 0.008). Focusing exclusively on the five patients who also traveled to sea level (Fig 2, right), we see similar trends. In both groups, there was no remarkable change in RERAs.
|
|
|
Discussion
The AHI decreased with descent to lower elevations, with the reduction being primarily secondary to fewer hypopneas and central apneas. This finding suggests that the increased oxygen content in the air at lower elevation is the more important of the three atmospheric factors affecting respiratory events with elevation change, within the altitude range studied.
Our findings parallel the findings of other investigators who noted AHI decreased with supplemental oxygen therapy. Smith et al2 found that nocturnal oxygen reduced the number of central apneas more than obstructive apneas in patients with primarily obstructive sleep apnea, but obstructive apneas did decrease. Gold et al3 found that in patients with a greater mixture of central and obstructive apneas, the AHI decreased with oxygen but the number of obstructive apneas actually increased.
Five of our patients (patients 4, 6, 7, 8, and 10) had a mixture of obstructive apneas and central apneas, with > 25% of the home elevation apneas being central. In this group, each had a reduction in central apneas (p = 0.07) and an increase in obstructive apneas with descent (p = 0.007). The pattern followed the changes described by Gold et al,3 when his patients with a mixture of central and obstructive apneas were administered oxygen.
The reduction in hypopneas with descent in our patients may be in part due to another phenomenon described by Gold et al.3 With oxygen administration, they found not only an increase in the baseline SaO2 but also a reduction in the magnitude of the SaO2 drop with each respiratory event.3 This effect becomes quite significant when a 4% SaO2 drop is part of the definition of hypopnea. Patient 5, who had primarily snoring and hypopneas, had his AHI fall from 44.4/h in Breckenridge at 3,139 m, to 13.9/h in Grand Junction at 1,370 m, and to 4.1/h near sea level. However his RERA index only increased from 1.2/h at 3,139 m to 4.1/h traveling to sea level.
Central apneas may be a manifestation of instability in ventilatory control.4 Wellman et al5 found that the OSA patients with the most instability in their ventilatory feedback loop, measured with the proportional assist ventilation technique, had the greatest decrease in their AHI in response to oxygen. Our patient with the greatest reduction in his AHI with descent (patient 9) had exclusively central apneas at his home elevation, perhaps a reflection of his increased loop gain. The AHI decreased from 49.5 to 4.2/h descending from 2,957 to 1,370 m. At 1,370 m, he had predominantly snoring, hypopneas, and RERAs.
The increased duration of the obstructive apneas and mixed apneas at lower elevation may also be an effect of oxygen. The falling PO2, as well as the rising PCO2, provides stimulus for arousal during a respiratory event.6 At lower altitude, the threshold SaO2 to trigger arousal may take longer to be reached, leading to a longer apnea. Motta and Guilleminault7 noted longer duration of sleep-induced apneas in response to oxygen.
Several studies have addressed the effect of increasing altitude on OSA. Netzer et al8 and Zielinski et al9 noted an increase in obstructive events during sleep, as a snoring mountain climber8 and as nine healthy volunteers9 ascended > 3,000 m. Burgess et al,10 however, observed the elimination of all obstructive apneas as an Australian climber, who had moderate OSA at sea level, reached 3,446 m and had prominent periodic breathing with exclusively central apneas.
Burgess et al11 then studied five patients with known moderate OSA at sea level, at 610 m, and at "2,750 m" simulated with isobaric hypoxia. At the simulated 2,750 m, all five OSA patients had exclusively central apnea with periodic breathing. The obstructive RDI fell from 25.5 ± 14.4/h at sea level to 0.5 ± 0.7/h at simulated 2,750 m, with the central RDI increasing from 0.4 ± 0.5 to 78.8 ± 29.7/h.
Our findings are not incompatible with those of Burgess et al.11 Our patient 4 had largely central apneas at altitude and obstructive apneas at sea level. Four other patients with a mixture of central and obstructive events at their home elevation had primarily obstructive events with descent. However, in our patients, obstructive events did not totally disappear at their residences > 2,400 m. This may be due to our patients being acclimated to their home elevation, while the patients of Burgess et al11 spent just 1 night at the simulated 2,750 m. With acclimatization, periodic breathing and central apneas might decrease allowing obstructive events to appear. Several authors1213 have found, below 4,400 m, periodic breathing decreases in some subjects with acclimatization. However, Salvaggio et al14 noted an increase in periodic breathing during 1 month at 5,050 m.
Warner et al15 also exposed snorers and OSA patients to severe hypoxia during non-REM sleep. Most patients had exclusively central apneas with severe hypoxia. However one patient with high baseline upper airway resistance and OSA in a normoxic setting had a pattern of mixed apneas with hypoxia. This may be analogous to our patient 7, whose obstructive/mixed apnea ratios were 358:11 at 1,370 m and 117:587 at 2,815 m. He had predominantly obstructive apneas at 1,370 m, but at his home elevation he had predominantly mixed apneas. The findings of Warner et al15 also parallel our findings that our three patients who maintained a predominance of obstructive and mixed apneas at their home elevation were the three patients who required 17 cm of continuous positive airway pressure, perhaps a reflection of very high upper airway resistance; all of our other patients required 
13 cm of continuous positive airway pressure.
Limitations
In considering our findings several limitations must be considered. First, this was a small population of patients but is likely representative of patients living at altitude. Second, there was some inconsistency in monitoring equipment. In five patients, the thermistor was exclusively oral in one of their Colorado studies, compared to nasal/oral thermistors in all of the other studies. However, in all of these five studies a nasal pressure cannula was also used. Thus, this likely had minimal impact. The oral thermistor was used in two patients at home elevation and three patients at 1,370 m, thus not inducing a bias. Third, there were also three studies done at sea level without the nasal pressure cannula. This may have led to an underscoring of RERAs, thus underestimating the RDI at sea level. However, this would not have affected the AHI since there were few 4% oxygen desaturations that occurred at sea level without some recognizable reduction in thermistor signal or belts. Fourth, 2 of the 27 studies could not be reviewed blinded as to the site of study. However, if we delete those two studies, the decrease of AHI, RDI, and hypopnea index with descent remains significant (p < 0.05 for each). A fifth limitation is the lack of physiologic studies to help characterize the individual patients regarding their hypoxic ventilatory responses and critical closing pressures. These measurements, in retrospect, may have helped explain why different patients responded differently to altitude change.
Conclusion
For patients residing between 2,400 m and 3,139 m, the AHI and RDI decrease with descent. All of the changes we observed regarding the effect of altitude descent paralleled previous observations on the effect of oxygen on OSA patients at sea level. Central apneas and hypopneas decreased in number, and obstructive apneas lengthened. It is important to appreciate that some patients with mild or moderate sleep apnea traveling from their home elevation in the mountains to lower urban sleep centers for sleep studies may fail to qualify for sleep apnea therapy. It therefore appears important for patients to undergo polysomnography at their home elevation if accurate diagnostic and therapeutic decisions are to be made.
Footnotes
Abbreviations: AHI = apnea-hypopnea index; non-REM = non-rapid eye movement; OSA = obstructive sleep apnea; RDI = respiratory disturbance index; RERA = respiratory effort-related arousal; SaO2 = arterial oxygen saturation
Dr. D. Patz directs and owns a private sleep laboratory in Grand Junction, Western Colorado Sleep Institute, which financed the research study. Dr. White receives income from Respironics (Murrysville, PA) for being Chief Medical Officer, but this had nothing to do with the choice of sleep recording systems in this study. Mobile Sleep Diagnostics owned Alice IV (Respironics) recording equipment for years prior to the conception of this research study. The study was also completed before Dr. White began his employment with Respironics. Dr. White is also a consultant for AspireMedical, WideMed, PAVAD, and Itamar Medical. The authors have no conflicts of interest to report.
Received for publication May 25, 2006. Accepted for publication June 27, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
