At 4,300 Metres, 'Hyperbaric' Often Means Sea-Level Pressure

Layered illustration of a mountain ridge line descending into a translucent soft-shell oxygen chamber, with a pressure-gauge arc reading a value barely above ambient and a scatter of oxygen molecules thickening toward the chamber's interior

Sixty healthy young men, all of whom had lived for at least six months at 4,300 metres, were randomised to five days of mild hyperbaric oxygen or to a control condition. The group that got the chamber slept better: their Pittsburgh Sleep Quality Index came out at 3.17 against 4.43 in controls, an adjusted mean difference of −1.08 points (95% CI −1.76 to −0.46, p=0.003). They were also more accurate on a working-memory task, and the functional near-infrared spectroscopy recorded a drop in oxygenated haemoglobin in the superior temporal gyrus during that task — reported by the authors as a shift toward more efficient processing rather than a deficit. The paper, from the rehabilitation medicine and high-altitude medicine centres at West China Hospital in Chengdu, appeared in Travel Medicine and Infectious Disease and calls mild hyperbaric oxygen “promising” for sleep and cognition on the plateau.

That result will be quoted a great deal. It is worth reading the fine print first — not because the trial is weak, but because of what “hyperbaric” turns out to mean when you trace these studies back to their methods sections.

What the five days actually contained

The Chengdu trial is the most recent of a run of papers on the same question, and the cleanest in design: randomised, 1:1, with a prespecified primary outcome (insomnia severity on the PSQI) and an imaging secondary. What the indexed abstract does not give is the pressure, the oxygen fraction, or the comparator’s mechanics — the authors describe the intervention as “mild hyperbaric oxygen” and leave the dose to the full text, which sits behind a subscription. The design is also labelled exploratory, enrolment was healthy young men rather than patients, the intervention lasted five days with no follow-up reported in the abstract, and 4,300 metres is a specific and unusual environment in which the ambient partial pressure of oxygen is roughly sixty per cent of what it is at sea level.

None of that is disqualifying. All of it matters when the finding is moved from Lhasa or Chengdu to a wellness clinic in Ohio.

Look at the plateau literature as a set and a pattern appears. In Lhasa, a team at Tibet University’s plateau brain-science laboratory put 56 newcomers who had lived at altitude for two years into a randomised comparison of fifteen hyperbaric sessions against no intervention, and found cognitive control capacity significantly higher afterwards in the treated group (p=0.005) while controls did not move (p=0.972). Their methods are unusually explicit: the device was a soft oxygen chamber, model NBYY-HYDT-003, intervention pressure 0.11 MPa, oxygen concentration 25 per cent, flow 10 litres a minute, about ninety minutes a session with a forty-minute ramp, thirty minutes at pressure and twenty minutes of decompression, delivered every other day for a month. Controls sat in the same chamber without the intervention, “unbeknownst to them” — a partial blind that most of this literature lacks.

Also in Lhasa, a second group ran 62 dyads through a randomised crossover, testing whether hyperbaric oxygen improved cooperative performance, with fNIRS hyperscanning across both participants’ medial prefrontal cortices. The answer was conditional: gains appeared in high-trust pairs and not in low-trust ones, and interbrain synchrony tracked the effect. Their methods section contains the detail that reorganises the whole field. The chamber was a plateau-type soft unit from Ningbo Yangyu Health Technology, holding two to six people, with oxygen purity inside kept below 95 per cent and dispersed oxygen concentration not exceeding 25 per cent. Pressure was raised over about fifteen minutes to 101.3 kPa — 1.00 ATA, which the authors describe, accurately, as “atmospheric pressure at sea level (0 m altitude)”. Participants sat in it for two hours.

At 1.00 ATA there is no hyperbaric component. The exposure is a two-hour stay at sea-level pressure in air modestly enriched with oxygen: a simulated descent from the plateau, not a pressurised oxygen treatment. That is a perfectly reasonable thing to study in people living at 4,300 metres — and it is not the intervention that clinical hyperbaric medicine means by the term, which starts at 2.0 ATA and above with close to 100 per cent oxygen.

In Xinjiang, an army hospital group randomised 80 patients with chronic insomnia at altitude to ten days of hyperbaric oxygen or control, and got larger numbers than anyone: PSQI 4.6 against 9.1 (adjusted mean difference −3.96, 95% CI −5.26 to −2.66), Insomnia Severity Index 5.0 against 9.8, with fatigue, anxiety and depression also improving and no serious adverse events. It was open-label, median age 25, and the conclusion — “efficacious and safe” — is stated more confidently than an unblinded ten-day trial in 80 young adults can support. The pressure used is again not in the abstract.

Four groups, four chambers, three provinces, one shared property: every one of them compares oxygen and pressure against ambient high-altitude hypoxia, not against a sham chamber delivering the same gas mixture at ambient pressure. That is the comparison that would separate “oxygen” from “pressure”, and the plateau is one of the few settings where it is both ethically simple and scientifically decisive.

The mouse experiment that points the same way

A mechanism paper from the hyperbaric oxygen department at the Chinese PLA General Hospital in Beijing supplies the same hint from the bench. Mice were kept in a hypobaric chamber simulating 7,000 metres — about 9.8 per cent oxygen at 0.47 ATA — for three consecutive days, then split into four groups: no treatment, no oxygen, normobaric oxygen at 1.0 ATA for an hour a day for three days, or hyperbaric oxygen at 2.0 ATA on the same schedule. Both oxygen arms reduced the cerebral damage seen on histology. Transcriptomic profiling showed partial restoration of gene expression, with enrichment in NF-κB, cytokine–cytokine receptor, IL-17 and PI3K-AKT signalling; targeted validation found antioxidant markers up (SOD, GSH), oxidative-stress markers down (MDA, NO), PI3K-AKT activation increased, and reduced microglial and astrocytic activation with lower IL-1β, IL-6 and TNF-α through the TLR4–NF-κB axis.

Hyperbaric oxygen modulated more pathways than normobaric oxygen did — but normobaric oxygen worked. The authors are careful about the rest: “the causal relationships between these pathways and their interactions require further validation.”

Where the pressure really is the medicine

There is a version of this therapy where the pressure is not decoration, and it runs in the opposite direction: at altitude, reduced barometric pressure is a cause of decompression sickness, not a target for treating it.

The Israeli Air Force has the numbers. Between 2015 and 2022 it put 2,279 altitude-chamber trainees and aviation physiology instructors through hypobaric exposures; 11 developed altitude decompression sickness, a risk of 0.5 per cent, with four further cases reported after combat flights. Musculoskeletal involvement was the commonest presentation, as it is in divers, and recompression remained the treatment — though a shorter hyperbaric protocol was used in 53 per cent of the altitude cases against 30 per cent of diving cases.

A case report from Lake Van in eastern Türkiye shows how unforgiving the arithmetic gets when the mountain and the water compound. A 54-year-old on his first-ever dive descended to about thirty metres in a lake whose surface lies well above 1,500 metres, stayed several minutes, then ascended rapidly without a decompression stop. He surfaced with bilateral leg weakness and breathlessness, aspirated water, lost awareness, and was towed roughly 300 metres to shore, where his dive team found no pulse and started CPR. Emergency services found spontaneous respiration and a pulse; he reached a hospital sixteen kilometres away within half an hour on high-flow oxygen and was referred onward for recompression. A single session on a US Navy Table 5 protocol produced marked resolution. He was discharged after declining further sessions.

Both of these are ordinary hyperbaric medicine, and both are indications the US Food and Drug Administration clears: decompression sickness sits on the FDA’s 510(k) Product Code CBF list, and it is indication 05 in the Undersea and Hyperbaric Medical Society’s fifteenth-edition list of fifteen.

What is cleared, and what the label is doing

Nothing on that UHMS list of fifteen covers altitude exposure, sleep, cognition, cooperation or working memory. Not one entry. Those trials are legitimate research, and their subjects were healthy volunteers rather than patients — but there is no cleared indication behind a chamber session for a fatigued newcomer on the plateau, and no insurance pathway either.

The terminology problem has now been named in the literature. A narrative review published in Medical Sciences in June by two clinicians from a Michigan hyperbaric practice argues that hyperbaric oxygen therapy and mild hyperbaric oxygen therapy “are not equivalent medical interventions”, and that calling low-pressure soft-chamber systems “HBOT” is a serious problem precisely because those systems have different mechanisms of action and different regulatory requirements. The authors — who disclose that they practise in a clinic using hyperbaric interventions — call for classification criteria covering pressure, oxygen fraction and chamber type, and point out that the overlap muddles research interpretation, patient expectations, reimbursement and trial design.

Read against the four plateau trials, that is not a semantic quibble. A study run at 1.00 ATA with dispersed oxygen below 25 per cent, a study at 0.11 MPa with 25 per cent oxygen, and a study at 2.0 ATA with 100 per cent oxygen are three different interventions, and only the abstracts that specify them can be compared. The wellness market has already blurred the line: soft chambers rated to 1.3 to 1.5 ATA are sold to consumers as hyperbaric therapy, and 1.3 ATA is closer to the exposures used in the Lhasa studies than to anything in a hospital wound-care unit.

The FDA’s position has not moved. Its August 2025 letter to health care providers reminds clinicians that HBOT devices are Class II, cleared through 510(k) under Product Code CBF, and that the agency is aware of fires “that resulted in serious injuries and deaths”, with the root cause not yet known. The recommendations are procedural and apply whatever the indication: follow the instructions for use, maintain fire prevention, train staff, monitor the patient for the whole treatment, keep to cleaning and maintenance intervals, avoid prohibited electrical or static devices, and dress patients in hyperbaric-compatible fabric such as cotton. None of that is waived by a wellness framing, and a soft chamber in a hotel room is not exempt because its pressure is low.

What would settle the plateau question is not another open-label course. It is a sham-controlled trial — same chamber, same gas, ambient pressure — that reports the pressure and oxygen fraction in the abstract, follows participants past the last session, and tests people with a clinical problem rather than healthy young men. Until then, the honest summary of this literature is narrower than its headlines: restoring sea-level pressure and somewhat more oxygen to people living at 4,300 metres appears to improve how they sleep and how they perform on cognitive tasks for as long as they are in the chamber. Whether that requires pressure at all is a question the field has so far declined to ask directly.

Sources

← Back to blog