CPAP Pressure Requirements for Obstructive Sleep Apnea Patients at Varying Altitudes

Abstract

Study Objectives:

This study was performed to determine whether the obstructive sleep apnea patient who requires a certain CPAP pressure at one altitude requires the same or a different CPAP pressure at a different altitude.

Participants:

7 mountain residents with OSA, comfortably using CPAP and planning to travel to lower altitude.

Design and Setting:

Participants used a ResMed Autoset autotitrating CPAP unit for at least 3 days at their home and at each of their lower travel destinations.

Measurements:

Nightly CPAP 95th percentile pressure, median CPAP pressure and AHI were recorded.

Results:

At high altitude, 7400 ft. to 10,100 ft (2255 m to 3080 m), average CPAP 95th percentile pressure was 9.43 (± 0.12) cm H2O. At low altitude, sea level to 2800 ft. (0 – 853m), average CPAP 95^th percentile pressure was 9.54 (± 0.13) cm H2O, P = 0.18. At high altitude, median CPAP pressure was 7.00 (± 0.10) cm H2O vs. 7.21 (± 0.14) cm. H2O at low altitude, P = 0.54. Three patients had a slight decrease in required pressure at low elevation, 4 had a slight increase.

Interventions:

N/A

Conclusions:

For obstructive sleep apnea patients living at altitude, changes in elevation between 10,100 ft (3075 m) and sea level do not significantly alter absolute CPAP pressure requirements.

Citation:

Patz DS; Swihart B; White DP. CPAP pressure requirements for obstructive sleep apnea patients at varying altitudes. SLEEP 2010;33(5):715-718.

CURRENTLY IT IS UNKNOWN WHETHER CPAP PRESSURE REQUIREMENT VARIES AS OBSTRUCTIVE SLEEP APNEA (OSA) PATIENTS TRAVEL FROM ONE elevation to another. From studies done with a non-breathing mannequin with a CPAP machine in a simulated altitude chamber, Fromm et al. determined that in order to maintain consistent pressure, and keep the mannequin's airway splinted opened, at increasing altitudes with thinner air, the fan speed of the CPAP machine needed to be increased.¹ Most CPAP machines on the market include the option of altitude adjustment based on the findings of their study. The machines can be adjusted to increase fan speed so that a patient who lives at sea level using 10 cm H₂O of CPAP, will continue to receive 10 cm H₂O of CPAP when s/he sleeps at altitude, such as 8000 feet (2438 m) in the mountains.

However the mannequin model does not take into account airflow dynamics or ventilatory drive in the breathing patient. What is unknown is whether the OSA patient who needs 10 cm H₂O of CPAP at his home near sea level needs 10 cm H₂O or a higher pressure or a lower pressure in the mountains at 8000 ft (2438 m). Likewise, should the high mountain resident who needs 10 cm H₂O of CPAP at home expect to need the same pressure or a different pressure at sea level? If a Fairplay, Colorado (CO) resident living at 10,200 ft (3109 m) needs 10 cm for an optimal titration while in a sleep lab in Denver, CO at 5200 ft (1585 m), what pressure should his/her machine be set to deliver at his/her home at 10,200 ft (3109 m)?

Environmental factors of high altitude that might affect CPAP pressure requirement for OSA patients include: (1) hypoxia – which tends to increase the proportion of central apneas^2,3 that may not require as much pressure; (2) lower air density/ air viscosity, which may lead to less upper airway resistance and less tendency for airway collapse and thus a lower CPAP pressure requirement; and (3) decreased atmospheric compressive force on the pharynx, which might reduce CPAP pressure requirement. On the other hand, severity of OSA has been shown to increase at altitude,⁴ likely an effect of the hypoxic environment. This might lead to an increased CPAP pressure requirement.

The current study was performed to evaluate whether OSA patients require the same CPAP pressure, or a different pressure, with altitude change.

METHODS

Seven OSA patients living between 7400 ft (2255 m) and 10,100 ft (3080 m), on fixed CPAP (or bilevel) therapy participated in the study. None required additional oxygen administration. They ranged in age from 35 to 73 years (mean 56 y). Their apnea-hypopnea indices on diagnostic attended polysomnograms performed months to years prior to this research study, ranged from 22 to 122/h (mean 50/h). Lowest SaO₂s ranged from 59% to 86% (mean 74%). Their CPAP pressures determined during attended titrations had ranged from 8 cm H₂O to 12 cm H₂O (patient 6 was using bilevel therapy at 12/8 cm H₂O). Body mass index ranged from 26-36 kg/m² (mean 31 kg/m²).

They had planned trips to destinations over 5500 feet (1675 m) lower in elevation for ≥ 3 days. Average descent was 8050 feet (range 5600–9500 feet) or 2453 m (range 1707–2895 m). Two of the 7 patients also stopped at mid-level altitudes for ≥ 3 days (one for 13 days, and one patient made 2 mid-altitude stops, 5 days and 3 days, respectively). In most patients, direction of travel was descent, except for patients 4 and 7 who continued to use the autotitrating CPAP after return to altitude after each of their descents. Each patient used a ResMed (Sydney, Australia) CPAP Autoset Spirit Advantage unit at their home, and at their lower travel destinations, for ≥ 3 nights at each location. The effectiveness of the Autoset, compared to sleep lab titrated fixed pressure, has been demonstrated by Stammnitz et al.⁵ The Autoset was set to allow for the full range of pressure from 5 to 20 cm H₂O, as needed. For each night, the 95th percentile CPAP pressure (the pressure at or below which the patient spent 95% of the time that night) was recorded, as well as the median CPAP pressure requirement, the machine generated AHI (apnea-hypopnea index), and hours of use. Nights in which less than 4.5 h of use were recorded were excluded. The Resmed Autoset adjusts automatically to altitude. Thus the CPAP pressure readings, in terms of absolute cm of water pressure at a particular altitude, are correct. The company guarantees validity to 8000 ft (2438 m). Since some of our patients live above that altitude, we evaluated its altitude adjustment at higher levels. With the unit set at a fixed pressure of 10 cm H₂O, with the mask occluded, the output was checked at 4 elevations during a single drive, 4,600 ft (1400 m), 6200 ft (1890 m), 8700 ft (2650 m) and 10,200 ft (3115 m) by water manometry. The Autoset produced precisely 10.0 cm H₂O at each elevation.

The protocol was approved and supervised by the St. Mary's Hospital Investigational Review Board in Grand Junction, Colorado. Each patient signed informed consent.

Statistical Analysis

The relationships between altitude and the 95th percentile CPAP pressures, median CPAP pressures and AHI on Autoset were analyzed. The means and standard errors of the 3 outcome measures for each individual within each high and low elevation were pooled together with equal weight. For statistical inference, the 3 outcomes were regressed on elevation in feet with univariate unadjusted linear mixed models with a random intercept for each patient. The data for the model are the repeated measurements on each individual, with multiple nights at different elevations. The random intercept model allows for this clustered data. This model accounts for the repeated measurements on each individual being correlated, yielding patient-specific estimated effects and P-values associating altitude change to required CPAP pressures and to AHI on auto-adjusting CPAP. Data are expressed as mean values with standard errors in parentheses.

RESULTS

In total, these 7 patients spent 96 nights using the Autoset > 4.5 h at high altitude, 21 nights at mid elevation, and 52 nights at low elevation (Table 1). Among these 7 patients, there were 11 nights with < 4.5 h of machine use and data from these nights were not included. Only 1 patient (patient 6) had a change of > 1 cm H₂O in either 95th percentile pressure or median CPAP pressure between the highest and lowest elevations. At high elevation, average 95th percentile CPAP pressure was 9.43 cm H₂O (± 0.12) vs. 9.54 cm H₂O (± 0.13) at low elevation (P = 0.18 for altitude affecting the 95th percentile pressure requirement over the altitude range of the study [Figure 1A]). At high elevation, the average median CPAP pressure required was 7.00 cm H₂O (± 0.10) vs. 7.21 cm H₂O (± 0.14) at low elevation (P = 0.54 for altitude affecting median CPAP pressure requirement over the altitude range of the study [Figure 1B.]). Thus neither the CPAP 95th percentile pressure nor the median CPAP pressure changed significantly with altitude change. Three of the patients had a slight decrease in pressure requirement at low elevation. Four had a slight increase.

Table 1.

CPAP pressure requirements at varying altitudes, determined by autoCPAP

pt. #	Elevation (meters)	Elevation (feet)	Number of Nights/Avg. hrs. per Night	Mean CPAP 95th Percentile Pressure, cm. H20	Mean CPAP Median Pressure, cm. H20	Mean AHI On AutoCPAP, #/hr.	Mean Apnea Index on AutoCPAP, #/hr.
1	3075	10,100	9 / 7.7	9.62 (0.29)	7.24 (0.21)	9.81 (1.79)	0.44 (1.01)
	1372	4500	13 / 7.0	9.95 (0.16)	7.14 (0.16)	4.03 (0.27)	0.02 (0.06)
	186	610	5 / 7.3	10.04 (0.52)	7.32 (0.66)	2.36 (0.32	0.16 (0.09)
2	2713	8900	6 / 7.0	9..43 (0.25)	6.92 (0.38)	19.92 (0.08)	6.82 (3.76)
	122	400	4 / 6.4	9.63 (0.29)	7.68 (0.32)	10.73 (0.85)	1.25 (0.94)
3	2713	8900	7 / 8.6	11.34 ( 0.15)	9.63 (0.18)	3.91 (0.40)	0.06 (0.05)
	30	100	5 / 7.6	10.88 (0.16)	9.44 (0.25)	3.94 (0.37)	0.08 (0.08)
4	2256	7400	18 / 7.4	10.26 (0.61)	7.72 (0.48)	4.39 (0.24)	0.37 (0.30)
	61	200	14 / 8.1	9.56 (0.90)	6.95 (0.93)	4.26 (0.33)	0.91 (0.26)
5	2621	8600	8 / 6.7	7.55 (0.29)	5.44 (0.14)	7.01 (1.68)	0.00 (0.15)
	2	5	7 / 6.4	7.93 (0.34)	5.40 (0.12)	4.56 (0.41)	0.59 (0.52)
6	2530	8300	3 / 8.0	8.23 (0.62)	5.83 (0.52)	3.13 (0.32)	0.00 (0.00)
	853	2800	4 / 8.6	9.85 (0.50)	7.48 (0.47)	2.48 (0.46)	0.13 (0.25)
7	2377	7800	45 / 6.6	9.57 (0.15)	6.19(0.08)	7.27 (0.46)	0.76 (0.86)
	1737	5700	3 / 6.7	9.87 (0.52)	6.20 (0.25)	5.23 (0.85)	0.53 (0.32)
	1280	4200	5 / 6.4	9.30 (0.60)	6.26 (0.27)	6.14 (1.55)	0.90 (0.88)
	213	700	13 / 6.4	8.90 (0.28)	6.19 (0.11)	5.12 (0.56)	0.45 (0.55)
Average of all 7 pts.	2612	8570	96 / 7.0	9.43 (0.12)	7.00 (0.10)	7.93 (0.71)	1.22 (0.31)
	209	687	52 / 7.2	9.54 (0.13)	7.21 (0.14)	4.74 (0.46)	0.51 (0.16)

This is the average data for each individual patient (pt.), at each elevation, including CPAP pressure requirements determined, and machine determined respiratory indices on autoCPAP at that elevation, and the number of nights that included over 4 ½ hours of machine use, and the average hours of use. Means are displayed with 1 standard error in parenthesis.

Figure 1.

Effect of altitude change on Autoset CPAP pressures delivered and machine recorded AHI for seven patients.

Figure 1A, and 1B show the values of CPAP 95th percentile pressure, and median CPAP pressure respectively for the individual patients at various elevation. 1C displays the machine reported AHI data at the various altitudes. Each data point in 1A, B, and C is a mean of 3 or more nights for that patient, at the given elevation. Bars represent 2 standard errors.

The machine recorded AHI on the autotitrating CPAP was increased slightly at high altitude. The mean AHI at the home (high altitude) locations was 7.93 /h (± 0.71), and at the lowest travel destination was 4.74 /h (± 0.46) (P < 0.01 for altitude affecting AHI while on the Autoset [Figure 1C.]). Only 2 of the 7 patients had an increase > 2.5/h comparing AHI at high vs. low altitude. It is unknown whether this represented an increase in central apneas. Patient 2 had a higher apnea index at altitude (6.82/h) than at sea level (1.25/h). Machine recorded apnea index for all other patients remained < 1/h at all elevations.

DISCUSSION

Changes in altitude from 10,100 ft (3100 m) to sea level did not significantly alter absolute CPAP pressure requirement in these 7 patients with obstructive sleep apnea. All but 1 of the 7 patients had ≤ 1 cm H₂O change in either CPAP 95th percentile or median CPAP pressure over this range of altitude. Thus it appears that altitude change, at least in the range from 3100 m to sea level, has minimal effect on required CPAP pressure.

Of the environmental factors changing with altitude, the barometric or compressive pressure may have little effect, because at any altitude, there is the same air density within the pharynx as outside of the neck, at least when the airway is patent. Regarding hypoxia at altitude, there may be an increased number of central events but central apneas would not necessarily demand a change in CPAP pressure. Finally, greater air viscosity at lower altitude might increase the CPAP pressures required at lower elevation. However in these 7 patients spending 169 nights at various altitudes, this factor did not cause statistically significant change in CPAP pressure requirements.

In addition to being helpful for OSA patients who live at altitude and travel, the findings in this study may be useful in the management of OSA patients in mountainous communities like Colorado, where much of the sleep lab availability is in urban facilities below the altitude of many residents. Sleep physicians can have some assurance that the CPAP pressure determined optimal in the urban sleep lab will probably still be effective at the home altitude of the patient, provided the CPAP unit is altitude adjusted to provide that precise pressure. However, oximetry or a follow up portable sleep study on CPAP at the home altitude may be necessary to see whether central apneas or hypoxemia are a problem at the home elevation. If so, supplemental oxygen could be added in-line.

Limitations

One limitation of the current autotitration evaluation is its dependence on the black box algorithm of the Autoset system. However, the design of the current study with autotitration was specifically chosen to allow multiple days at each elevation in order to minimize the importance of variations in body position and night to night variability. Also automated titration, despite its drawbacks, eliminated concern regarding human variation in titration style.

A second limitation in this study is the small number of patients analyzed. However, in total, there were 169 sleep nights evaluated with consistent results. Based on variances demonstrated in these 7 patients, a prospective power analysis suggests a power of over 80% for a study of similar design and analysis, with 7 patients, to detect a requirement for > 1 cm H₂O change in either CPAP 95th percentile pressure or change in median CPAP pressure with altitude varying from sea level to 10,000 ft (3048 m).

A third limitation is that the subjects are exclusively mountain residents, and their direction of travel was mostly descent. It is possible that lowlanders on CPAP have additional factors that affect CPAP requirements, traveling up into the mountains that are different than those encountered in these 7 highlanders.

A fourth potential limitation of this study is the slight increase in AHI on the Autoset at altitude. One might interpret this as indicating that the Autoset does not respond as well to obstructive events at altitude, and that in fact, greater pressure is required at altitude, but the machine simply does not respond to this requirement. On the other hand, this may represent an increase in central events, primarily in 2 of the patients at altitude, and it may have been appropriate that the machine not respond to them.

CONCLUSION

For patients with OSA, descent in altitude from 10,000 ft (3048 m) to sea level does not significantly alter absolute CPAP pressure requirements. The change in fan speed currently utilized by many CPAP machines to adjust for barometric pressure change and keep the delivered CPAP pressure consistent at different altitudes is all that is necessary for obstructive sleep apnea patients traveling to lower elevation.

DISCLOSURE STATEMENT

This was not an industry supported study. Dr. Patz owns and is medical director of a private, for-profit sleep lab. Dr. White is Medical Director of Phillips Respironics which markets CPAP machines, and he consults for Itamar Medical and PAVAD. Mr. Swihart has indicated no financial conflicts of interest.