Study Looking at End Expiratory Pressure for Altitude Illness Decrease (SLEEP-AID)

Introduction

Acute mountain sickness (AMS) is a potentially debilitating syndrome of headache, nausea, dizziness, sleeplessness, and fatigue common among unacclimatized travelers to high altitudes. Reported rates of AMS range from 25%–70% in tourists, trekkers, and climbers (Honigman et al., 1993; Wagner et al., 2008; Jackson et al., 2010; Lipman et al., 2012; McDevitt et al., 2014). While the best prevention of AMS is a gradual ascent, rapid ascent may provoke AMS despite the use of acetazolamide (Jackson et al., 2010). Both the civilian recreationalist as well as search and rescue and military operations with “high risk” ascent profiles may find themselves forced to ascend to high altitude, and could benefit from an inexpensive and lightweight nonpharmacologic option to protect against AMS.

The low barometric pressure at high altitude results in a lower partial pressure of oxygen leading to lower oxygen saturations. There appears to be a central role of hypoxemia causing AMS (Sutton et al., 1979; Bartsch et al.,1990; Roach et al., 1998; West, 2003), where the ability to compensate for hypoxia dictates subsequent altitude illness as well as mental and physical performance. Oxyhemoglobin saturation decreases during sleep at high altitude (Hultgren, 1997), and individuals who develop AMS consistently have lower nighttime oxygen saturations (Burgess et al., 2004; Erba et al., 2004). Sleep at high altitude has been found to be a more powerful instigator of desaturations then ambulation (Burgess et al., 2004), and nocturnal oxygen supplementation has been found to attenuate AMS and improve sleep quality (Luks et al., 1998; Barash et al., 2001).

Applied positive end-expiratory pressure (PEEP) increases airway and transpulmonary pressures, functional residual capacity, and arterial oxygenation (Duncan et al., 1986). Expiratory positive airway pressure (EPAP) applies constant positive pressure throughout the exhalation phase and is similar to PEEP, in that expiratory resistance is generated. Acute mountain sickness incidence and severity have been shown to decrease with both PEEP machine (Koch et al., 2008; Johnson et al., 2010; 2013) and EPAP-producing face mask with an expiratory valve (Savourey et al., 1998; Launay et al., 2004). However, current systems with face masks and valves are unwieldy, uncomfortable, and due to weight and space constraints are unlikely to be utilized by an expedition doctor, medic, or within the self-supported climber's first aid kit. The objective of this study was to examine whether a novel nasal EPAP device could decrease the incidence AMS, as well as improve AMS symptoms and sleep physiology metrics.

Materials and Methods

Ascent to study enrollment sites

The approach to the study enrollment sites was a gradual ascent, with the majority of trekkers beginning after flying to the town of Lukla at 2860 m (9380 ft). Most trekkers took 5 days to ascend the 1670 m (5479 ft) to Dingboche 4530 m (14,800 ft) or Pheriche 4371 m (14,340 ft) (including an acclimatization day half way up in Namche Bazaar). The average daily altitude gain while hiking is 417 m (1368 ft), with the final 488 m (1601 ft) in the last day, reaching the enrollment site unacclimatized (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Trekking ascent profile. *Study enrollment sites.

Intervention

Participants were randomized using computer-generated random sequence and given sealed numbered packages with their interventional device, Watch-PAT200^™ continuous sleep monitor (Itamar Medical, Caesarea, Israel), questionnaires, and instructions. The randomization code was unavailable to administrators or participants. The sham device was prepared by manual removal of the device's interior gill-like strips that deliver 5 cm H₂O of resistance to expired nasal airflow (Doshi, 2012; personal communication), maintaining an externally identical appearance to the functional EPAP device. Theravent™ (Fig. 2) is FDA cleared for nocturnal use for snoring prevention (Doshi and Westbrook, 2012) but not AMS prevention, thus an investigational device waiver was obtained.

OPEN IN VIEWER

FIG. 2.

Above: Theravent™ nasal EPAP device. Arrows denote directional air-flow on expiration and inspiration, respectively, EPAP, expiratory positive airway pressure. Below: Watch-PAT200™

Results

A total 223 participants were enrolled (4 were excluded by criteria) and randomized (Fig. 3), with similar baseline demographics in both groups (Table 1). We performed a modified intention-to-treat approach, where all randomly assigned study participants were analyzed except those not fulfilling inclusion criteria and those who never received the intervention (Fergusson et al., 2002). Additional analyses were performed in the proper use population, defined as those who wore the nasal strip the entire night as a dichotomous outcome—as there were a substantial portion of participants who removed, lost, or had failure of the nasal strip during the night.

OPEN IN VIEWER

FIG. 3.

Participant flow chart. AMS, acute mountain sickness; EPAP, end expiratory airway pressure; ITT, intent-to-treat.

Of the 219 enrolled, 33 participants used the EPAP device without the Watch-PAT200. These individuals completed symptom questionnaires included in primary and secondary outcome analyses, but did not participate in sleep monitoring. Of the 186 sleep studies, 35 (19%) contained some form of unrecoverable device error. These errors stemmed from issues with disposable sensors and their connections, and in some cases may have been exacerbated by cold sleeping conditions. An additional 23 (12%) measurements were excluded due to unrecoverable human error (e.g., sensor misapplication, premature device removal, or device not turned on). In nine of the incomplete studies due to device or human error, the PAT probe failed to collect data, while the Spo₂ probe was functioning. In these cases, we manually computed the various oxygenation variables from extracted raw sensor data. The remaining 128 (69%) studies were processed without any error and relevant sleep metrics reported.

Discussion

This was the first double-blind randomized placebo controlled trial to examine expiratory positive airway pressure for prevention of AMS. The intervention was not found to be protective against the development of acute mountain sickness incidence or overall severity. However, those randomized to the EPAP device found significant decreases in both incidence and severity of the AMS subgroup symptoms of headache and dizziness. Headache is a defining characteristic of AMS and overwhelmingly the most common complaint at high altitude (Burtscher et al., 1998; 2011; Hackett and Roach, 2001). High altitude headache is an important health problem in those who travel to high altitude (Burtscher et al., 2011), and the benefit of its prevention should not be underestimated. This headache has been described not only as the predominant symptom in the context of AMS, but also as an isolated occurrence solely due to hypoxia (Serrano-Duenas, 2007). It is reasonable that prevention of nocturnal hypoxia may have led to the prevention of high altitude headache upon awakening.

The EPAP group demonstrated an improvement in sleeping Spo₂ while limiting the percentage of nocturnal desaturations. It is likely that the improved nocturnal oxygenation was due to the increased positive airway pressures induced by nasal EPAP. This theoretically increases lung volume, improves upper airway patency, and minimizes alveolar collapse during expiration (Heinzer et al., 2005; 2006). The maintenance of alveolar patency and recruitment of micro-atelectatic alveoli may also increase functional residual capacity (Wayne, 1976). Positive pressure has been shown to improve hemoglobin oxygen saturation (Larson et al., 1982; Schoene et al., 1985) and prevent nocturnal desaturations at high altitude (Johnson et al., 2013). This is similar to findings in patients with hypersomnia-sleep apnea syndrome (Mahadevia et al., 1983), in addition to postoperative hypoxemic patients (Rieg et al., 2012).

Sleep at high altitude is affected by periodic breathing. First described in 1898 (Di Giulio and West, 2013), this central form of sleep apnea (CSA) occurs due to a respiratory feedback loop of hypobaric hypoxia induced hyperventilation that alternates with hypocarbia-induced inhibition of the ventilatory drive—and subsequent apneic events. The hypothesis that CSA induced nocturnal hypoxia is a major factor in development of AMS has been debated by some (Eichenberger et al., 1996), but seems biologically intuitive that more frequent hypoxic events and a decreased average nocturnal Spo₂ could compound the hypoxic insult of high altitude and worsen AMS (Burgess et al., 2004; Erba et al., 2004).

The nasal EPAP device was found to significantly improve the average nocturnal peripheral saturation above 80%, as well as substantially decrease the percentage of sleep time below 80% Spo₂. It is reasonable to attribute the observed decrease in subsequent AMS symptoms to this effect. Oxygen saturation of 80% results from a partial pressure of oxygen (po₂) of approximately 45 mmHg, and is situated near the inflection point of the “steep” aspect the oxygen dissociation curve. At this point in the curve, small decreases in po₂ result in significantly larger drops of Spo₂. At high altitude, the body is operating nearer critical thresholds in po₂ and carbon dioxide (White et al., 1987), and similarly small incremental decreases in oxygenation have been significantly associated with mild and severe AMS (Mandolesi et al., 2014).

The sensation of poor sleep at altitude was frequently observed in both groups, with similar numbers of nocturnal awakenings. The observed 66% of overall sleep impairment was similar to prior high altitude sleep studies, where 32%–60% of subjects reported disturbed sleep (Jafarian et al., 2008). Sleep disturbances at high altitude are frequent, and sleep quality has been shown to have the weakest correlation with other AMS symptoms (Macinnis et al., 2013). While sleep time was significantly decreased in the EPAP group, it was difficult to determine if this was due to the discomfort of increased work of breathing with expiratory resistance. Altitude-associated changes in sleep architecture include decreased time spent in the slow wave/deep stages (III and IV) with an increase in the lighter (I and II) sleep stages (Goldenberg et al., 1988; Anholm et al., 1992; Johnson et al., 2010) as well as decreased duration of rapid eye movement (REM) sleep by as much as 50% (Anholm et al., 1992; Mizuno et al., 1993; Wickramasinghe and Anholm, 1999). The EPAP group showed a significant decreases in sleep latency. The improved oxygen saturation may have contributed to a more rapid time to sleep. We found a positive correlation between the percentage of time in deep sleep and average Spo₂ with nasal EPAP, indicating that while participants had decreased total amount of sleep, the decrease in hypoxia may have contributed to subtle sleep architecture changes.

The studied EPAP device is inexpensive, portable, and convenient—a marked departure from the cumbersome EPAP and electricity dependent PEEP devices that have been examined as nonpharmacologic alternatives for AMS prevention. Nasal EPAP is also more logistically feasible for prevention of nocturnal hypoxia than dead space masks, which have shown to decrease central apnea but not AMS (Patz et al., 2013). The challenge of implementation of the studied device was the level of user discomfort. For some participants, sleeping in the cold environment of high altitude led to an uncomfortable amount of condensation within the nasal cavities. While it may be implied that the expiratory resistance of the EPAP device may have led to greater discomfort, subsequent noncompliance, there was no statistically significant difference in the rate of noncompliance or sleep quality between the two groups.

The observed low incidence of acute mountain sickness was approximately half the rate as previously reported at the same locations (Fagenholz et al., 2009; Cushing et al., 2013). The decrease of AMS at the study enrollment sights followed a similar historical trend recently reported from Manang, Nepal (McDevitt et al., 2014). Re-calculating a power analysis with an effect size of 10% reduction of disease using our observed incidence of AMS would have result in a study enrollment goal of 368 participants, which was potentially unfeasible. Although we exceeded our power analysis of 124 by 95 randomized participants, we may not have been able to find efficacy of the intervention due to the previously undocumented low incidence of disease at our enrollment sites. This decrease in AMS incidence could have been attributed to increased educational efforts, more conservative ascent profiles, increased use of acclimatization medications, or better recognition and treatment of AMS symptoms prior to further ascent.

Although there appeared to be substantial benefits to the nasal EPAP device, there was some difficulty in correct usage, as only 54% indicated they used the strip continuously throughout the night. The most common reported reasons for noncompliance were: discomfort with the strip over the nares, feeling of breathlessness, condensation buildup inside the nasal strip, feeling of dry mouth or mouth-breathing, or failure of the strip adhesive. Implementation of the EPAP device to the high altitude milieu may require low altitude practice and familiarity prior to adoption and utilization that may potentially improve compliance and benefit.

On account of the high numbers of partial or noncompliance, we analyzed the data in two ways: proper use and intent-to-treat analysis. Thus we could address any differences in trends that were noted between groups and be as concise as possible with data interpretation. We considered noncompliance as a dichotomous outcome, as we were unable to quantify the amount of partial use. Therefore, the intent-to-treat analysis is likely most reflective of the actual affect of the nasal EPAP device, as it included a fair amount of partial implementation that may have conferred physiologic benefit.

While some study participants may have been able to determine their randomization group by inspecting the device closely, this bias is unlikely to have occurred as 54% of participants correctly guessed their intervention group the following morning.

Conclusion

This randomized placebo controlled trial found that an inexpensive and disposable nasal EPAP device did not prevent acute mountain sickness, but did decrease the incidence and severity of headache and dizziness while improving nocturnal oxygen saturation and limiting the percentage of desaturations. Further studies of nocturnal EPAP devices are warranted to further explore this adjunct for potential AMS prevention.