Hyperbaric oxygen therapy works through one core mechanism. Pressure forces more oxygen into the blood than breathing can deliver at sea level.
That extra oxygen reaches tissue that normal circulation cannot. It drives changes in gene expression. Over weeks, it rebuilds the blood supply in damaged areas.
This guide walks through the science across five threads:
- Pressure physics
- Oxygen transport
- Cellular response
- Where trial data backs the mechanism
- Where the evidence remains thin
The goal is to help patients and clinicians see what HBOT actually does at the cellular level. And to separate that from the marketing layered on top.
The pressure mechanism
Normal sea-level air is 21% oxygen at 1 atmosphere absolute (ATA). At rest, an adult's arterial oxygen tension sits around 95 to 100 mmHg. Tissue gets oxygen mostly through hemoglobin in red blood cells, with a small fraction dissolved in plasma.
Hyperbaric therapy changes this. At 2.4 ATA breathing 100% oxygen — the most common clinical pressure — arterial oxygen tension rises to roughly 1,500 mmHg, per Tibbles & Edelsberg 1996 in NEJM. That is a fifteenfold increase over sea level.
Most of that extra oxygen does not ride hemoglobin. Hemoglobin saturates near 100% at sea level — there is no room to load more.
The extra oxygen dissolves into plasma. Plasma can carry much more oxygen under pressure. It also reaches tissue that low-flow circulation can't always reach.
At 2.4 ATA, plasma can dissolve enough oxygen to meet basic tissue demand without hemoglobin at all. This is why HBOT can keep patients alive who have lost most red cell function — including severe CO poisoning where hemoglobin is bound to CO and useless.
Why pressure matters more than oxygen alone
The pressure piece is what makes HBOT different from supplemental oxygen.
Breathing 100% oxygen at 1 ATA — what hospitals call standard high-flow O2 — raises arterial oxygen to about 600 mmHg. That helps. But it does not produce the same tissue effects as HBOT because the gas is not under additional pressure to dissolve it deeper into plasma.
Adding pressure follows Henry's Law: the amount of gas that dissolves in a liquid is proportional to the partial pressure of the gas above the liquid. Going from 1 ATA to 2.4 ATA roughly multiplies the oxygen that can stay in solution.
This is why soft-shell chambers at 1.3 ATA do not produce hospital-grade HBOT physiology.
At 1.3 ATA breathing room air, arterial oxygen tension barely moves above sea-level baseline. At 1.3 ATA breathing oxygen through a concentrator, the rise is modest. Soft chambers operate in a different pressure regime entirely.
What happens at the tissue level
Once tissue oxygen rises, several downstream effects follow. These are the mechanisms that drive most of the FDA-approved indications.
Blood vessels narrow with oxygen still going up. High oxygen makes blood vessels constrict. But tissue still gets more oxygen because the gas is now dissolved in plasma at high levels.
This cuts swelling in conditions like crush injury and compartment syndrome, per Strauss 2012 in Undersea & Hyperbaric Medicine. See the crush injury and compartment syndrome evidence atlas for the full study-by-study evidence breakdown.
New blood vessel growth. Sustained high oxygen signals damaged tissue to build new capillaries. This effect builds over many sessions and shows up around session 15 to 20.
The vessel-growth response is the central mechanism behind HBOT's use in chronic wound healing and radiation tissue damage.
Anti-bug effect. Many oxygen-hating bacteria — including the Clostridium species behind gas gangrene — cannot survive high tissue oxygen. HBOT kills these germs directly.
For oxygen-tolerant bacteria, HBOT boosts immune cell killing power. This helps restore immune function in tissue that was too low on oxygen to fight infection.
Stem cell release. HBOT boosts bone marrow stem cell release into the blood, per Thom 2006 in The American Journal of Physiology.
What this means for patients is still being worked out in the literature.
Gene expression changes. Studies have shown HBOT shifts how key genes turn on. The list includes HIF (a low-oxygen sensor), VEGF (which drives blood vessel growth), and several antioxidant genes.
These changes stick around after the dive ends. They build up over a full multi-session course.
The 14 FDA-approved indications
The above mechanisms explain why HBOT works for the conditions on the UHMS-recognized list. The 14 uses fall into three buckets:
Oxygen displacement and emergency uses:
- Decompression sickness
- Air or gas embolism
- Carbon monoxide poisoning
- Severe anemia (where transfusion is impossible)
Tissue oxygenation and wound healing:
- Chronic diabetic foot ulcers
- Delayed radiation injury
- Compromised skin grafts and flaps
- Acute thermal burn injury
- Crush injury and acute traumatic ischemia
- Chronic refractory osteomyelitis
Infection control and immune support:
- Gas gangrene (clostridial myositis)
- Necrotizing soft tissue infections
- Intracranial abscess
- Idiopathic sudden sensorineural hearing loss
Each of these indications has clinical trial support at 2.0 to 3.0 ATA. The 2015 Cochrane review on diabetic wounds and the 2018 Eskes review on osteomyelitis summarize the wound and bone evidence.
Clinical protocol basics
Most FDA-approved indications use similar protocols. Forty sessions of 90 to 120 minutes each, at 2.0 to 2.4 ATA, with 100% oxygen breathing during the dive and short air breaks every 20 to 30 minutes to reduce oxygen toxicity risk.
CO poisoning and decompression sickness use higher pressures (2.5 to 2.8 ATA) and shorter sessions. These are often single emergency dives.
The US Navy Treatment Table 6 is the reference protocol for decompression sickness worldwide.
Wound care typically runs 30 to 60 sessions over 6 to 12 weeks. Patients return daily Monday through Friday during active treatment, with weekend breaks.
The protocols matter because they match what was tested. Off-label protocols at lower pressure or different session counts may not produce the outcomes the published trials showed.
The off-label question
Beyond the 14 approved uses, HBOT runs off-label for many other conditions. The list includes TBI, long COVID, autism, anti-aging, multiple sclerosis, and fibromyalgia. See the multiple sclerosis evidence atlas for the full investigational evidence breakdown.
The evidence base varies a lot by condition.
Better-studied off-label uses with mixed results:
- Stroke recovery — preliminary positive data, no consensus
- Mild TBI — studies show modest effects, the BIMA trial sponsored by DOD found no benefit over sham
- Long COVID — Zilberman-Itskovich 2022 in Scientific Reports showed cognitive improvement at 2.0 ATA over 40 sessions
Less well-studied off-label uses:
- Autism — Granpeesheh 2010 found no benefit at 1.3 ATA; reviews remain inconclusive
- Anti-aging — the Hachmo 2020 telomere study at Tel Aviv University had 35 participants without independent replication
- Fibromyalgia, chronic fatigue, lyme disease — small studies, no major reviews
Off-label use is legal but not FDA-approved. Insurance does not cover off-label HBOT. Patients pursuing off-label protocols should understand they are paying for treatments where the evidence ranges from preliminary to absent.
Safety profile of the mechanism
HBOT works through real biology. That same biology creates real risks at high pressure and high oxygen.
Middle ear barotrauma is the most common adverse event. Rates run 2% to 10% per the Camporesi 2014 review in Undersea & Hyperbaric Medicine.
Patients with ear tube issues, recent ear surgery, or a cold should tell the clinic before starting.
Oxygen toxicity seizures occur at roughly 1 to 4 per 10,000 sessions at 2.4 ATA. Risk rises sharply above 2.8 ATA, which is why most non-emergency protocols cap there. Patients on certain medications including ciprofloxacin and steroids may have higher seizure risk.
Lung barotrauma is rare but possible. Risk goes up with untreated pneumothorax, severe COPD, or recent chest surgery. Pre-treatment chest imaging catches most cases.
Fire risk is the most serious chamber-related risk. The 1997 Milan chamber fire killed 11 people inside a multiplace chamber. Reputable facilities follow NFPA 99 standards including no synthetic clothing, no electronics, no oils or alcohols, and exhaust scrubbing protocols.
Chamber types and the mechanism
The mechanism applies to clinical chambers from Sechrist Industries, Perry Baromedical, ETC Biomedical, and Healing Chambers International. All run at 2.0 to 3.0 ATA under FDA Class A clearance.
The mechanism does not transfer to consumer soft chambers from OxyHealth, Newtowne Hyperbarics, or Summit to Sea. These chambers are FDA-cleared only for acute mountain sickness at 1.3 ATA per the 510(k) database. At that pressure, arterial oxygen tension does not approach hospital HBOT levels.
Marketing that claims soft-shell mild HBOT produces the same effects as hospital HBOT contradicts the physiology. The pressure gap is too large for the dissolved-oxygen mechanism to operate the same way.
UHMS accreditation and clinical quality
Roughly 180 of the 1,588 US HBOT centers carry UHMS accreditation — about 11%. Accreditation signals that the site:
- Has a medical director trained in hyperbaric medicine
- Uses CHT or CHRN-certified technicians
- Follows documented emergency and fire protocols
- Submits chambers to annual third-party inspection
For FDA-approved uses under Medicare code G0277, billing requires physician supervision. UHMS accreditation is the best signal that a clinic runs the protocols the approved uses are built on.
Related reading
- HBOT pressure explained: 1.3 vs 2.0 vs 2.4 ATA
- How HBOT increases tissue oxygenation: the physiology
- UHMS approved HBOT indications: the 14 evidence-backed uses
- Hard chamber vs soft chamber HBOT: real clinical differences
Frequently asked questions
How does HBOT differ from breathing oxygen at sea level?
Breathing 100% oxygen at 1 ATA raises arterial oxygen tension to about 600 mmHg. HBOT at 2.4 ATA raises it to about 1,500 mmHg. The added pressure forces oxygen into plasma at levels the gas can't reach at sea level. The plasma-dissolved fraction is what reaches tissue that normal hemoglobin-bound delivery cannot.
How long does the tissue effect last after a session?
Direct oxygen elevation drops within minutes of leaving the chamber. The downstream effects — angiogenesis, gene expression, antimicrobial — persist for hours to days and accumulate across sessions. This is why most protocols use 20 to 60 daily sessions rather than single dives.
Does HBOT work for off-label conditions?
Sometimes. Long COVID and stroke recovery have preliminary positive data. TBI is mixed. Autism trials at 1.3 ATA have not shown clear benefit. Anti-aging rests on a single 35-patient study without replication. Patients pursuing off-label HBOT should match expectations to the published evidence for their specific condition.
Why do soft chambers exist if they don't replicate hospital HBOT?
Soft chambers are FDA-cleared for one indication: acute mountain sickness at 1.3 ATA. Marketing has expanded into off-label wellness use, but the FDA clearance scope did not expand. Soft chambers are real medical devices for a narrow indication. The off-label claims around them often overstate the evidence by citing hospital-pressure studies that do not transfer down to 1.3 ATA.
How safe is HBOT compared to other interventions?
For the FDA-approved indications, HBOT is well-tolerated when delivered at UHMS-accredited facilities. Common adverse events (ear barotrauma) are usually minor and resolve. Serious events (oxygen toxicity seizures, fire) are rare but possible. Risk meaningfully exceeds standard outpatient care, which is why provider selection matters.
Medical disclaimer
This article is for informational purposes only and is not a substitute for professional medical advice. Hyperbaric oxygen therapy is investigational for most off-label uses discussed here. Consult your doctor before starting any HBOT protocol, especially if you have pre-existing ear, lung, or cardiovascular conditions. The FDA has approved HBOT for 14 specific indications; uses beyond these are off-label.
-- The HBOT Finder Team