- UHMS Approved Indications: The Regulatory Baseline That Matters
- Weaver 2002 (NEJM): The Landmark Carbon Monoxide RCT
- Efrati 2013: Post-Stroke Neuroplasticity (n=74, Tel Aviv Randomized Trial)
- Efrati/Hadanny 2020 Aging Cell: Telomere Lengthening and Senescent Cell Reduction
- Miller 2015 and Wolf 2012: The DoD TBI Null Results
- Rossignol 2009: The Autism RCT That Launched a Market (n=62)
UHMS Approved Indications: The Regulatory Baseline That Matters
The Undersea and Hyperbaric Medical Society (UHMS) is the professional and scientific body that evaluates evidence for hyperbaric oxygen applications and maintains the list of approved clinical indications. Fourteen conditions currently meet the UHMS evidence standard for recommendation. The list includes: arterial gas embolism, carbon monoxide poisoning (including cyanide poisoning), clostridial myositis and myonecrosis (gas gangrene), crush injury and acute traumatic ischemia, decompression sickness, arterial insufficiency (selected problem wounds), severe anemia, intracranial abscess, necrotizing soft tissue infections, refractory osteomyelitis, delayed radiation injury (soft tissue and bony necrosis), compromised skin grafts and flaps, thermal burns, and diabetic foot ulcers (as a component of a comprehensive wound care program). Each of these approvals is supported by controlled human studies demonstrating outcome benefit and a plausible mechanism for HBOT's role.
The UHMS position statements are notable for what they explicitly do not include. TBI and concussion are not on the approved list, with UHMS noting that available controlled evidence does not support benefit. Autism spectrum disorder is not on the list. Anti-aging and longevity are not on the list. Long COVID is not on the list. The UHMS is conservative by design — it represents clinicians whose professional reputation rests on evidence-based practice standards — and its explicit non-endorsement of the most commercially active off-label markets is a meaningful signal about where the evidence actually stands.
Insurance coverage in the United States follows the UHMS list closely: the 14 approved indications are generally covered by Medicare, Medicaid, and most commercial insurers when provided in a certified hyperbaric facility with a qualifying diagnosis. All off-label uses are cash-pay. This insurance boundary is the single clearest regulatory signal about where the evidence is strong and where it isn't. The clinic charging $9,600 for a 40-session TBI protocol is doing so in an insurance coverage gap that exists because the evidence gap is real.
Weaver 2002 (NEJM): The Landmark Carbon Monoxide RCT
Lindell Weaver and colleagues published the definitive randomized controlled trial of HBOT for carbon monoxide poisoning in the New England Journal of Medicine in 2002. The study enrolled 152 patients with CO poisoning presenting to emergency departments, randomizing them to either three sessions of HBOT (at 3.0 ATA for 100% oxygen) or three sessions of normobaric oxygen (100% oxygen at 1.0 ATA, the current standard of care at the time). The primary outcome was cognitive sequelae — neuropsychological impairment — at 6 weeks. The HBOT group showed significantly lower rates of cognitive sequelae at 6 weeks (25% vs. 46%, p=0.007), a clinically meaningful difference on a patient-important outcome in a serious condition. The effect was durable: at 12 months, the HBOT advantage remained significant.
The Weaver 2002 trial is the best evidence in the HBOT literature: large sample for an emergency intervention, randomized, blinded (patients received oxygen in a hyperbaric chamber whether in the HBOT or normobaric control group), objective primary outcome with validated measurement tools, multi-center design, and adequate follow-up duration. It established HBOT for CO poisoning as the clearest high-evidence application in the field, and its methodology defines what high-quality HBOT research looks like — a standard the off-label literature has rarely met.
The mechanism for CO poisoning is also the most straightforward in HBOT medicine: carbon monoxide binds hemoglobin with 200 times the affinity of oxygen, creating carboxyhemoglobin (COHb) that cannot carry oxygen. HBOT at 3.0 ATA dissolves enough oxygen directly into plasma — independent of hemoglobin — to sustain tissue oxygenation while reducing the COHb half-life from 5–6 hours (breathing room air) to approximately 20 minutes (breathing 100% oxygen at 3.0 ATA). The mechanism is quantified, the physics are established, and the Weaver trial confirmed the clinical benefit. This is what it looks like when HBOT works.
Efrati 2013: Post-Stroke Neuroplasticity (n=74, Tel Aviv Randomized Trial)
Shai Efrati and colleagues at the Shamir Medical Center and Tel Aviv University published a 2013 PLOS ONE randomized controlled trial of HBOT in patients with chronic post-stroke cognitive impairment. The study enrolled 74 patients who had experienced a stroke 6–36 months prior, randomizing them to 40 HBOT sessions (2.0 ATA, 100% oxygen, 90 minutes) or a 2-month observation period followed by crossover to HBOT. The primary outcomes were neuropsychological measures of cognitive function and SPECT imaging of cerebral blood flow. The HBOT group showed significant improvement in cognitive domains including memory, attention, and information processing, with corresponding SPECT imaging changes showing increased metabolic activity in previously hypoperfused regions — the imaging findings interpreted as neuroplasticity induction in chronically ischemic "dormant" tissue.
The 2013 Efrati trial became the foundation for a commercial HBOT neuroplasticity program at the Sagol Center in Tel Aviv, which has treated thousands of patients internationally. The scientific hypothesis — that HBOT activates neuroplasticity in chronically ischemic penumbral tissue that remains metabolically dormant but structurally intact long after stroke — is mechanistically coherent and was supported by the imaging data. The trial's limitations are real: 74 patients at a single center, with the crossover design creating potential confounds, the imaging interpretation requiring expert reviewers, and the follow-up period ending at 3 months post-treatment with no long-term outcome data.
The trial has not been independently replicated by researchers outside the Efrati lab at the scale and rigor required to establish clinical guidelines. The Cochrane review on HBOT for stroke [1] found insufficient evidence to determine whether HBOT improves outcomes in acute or chronic stroke compared to no treatment or sham. The gap between Efrati's trial-level findings and the Cochrane-level evidence summary is not unusual for single-center trials in this field: a 74-patient positive result at one center does not move the evidence needle until replicated, and the replication has not happened at adequate scale.
Efrati/Hadanny 2020 Aging Cell: Telomere Lengthening and Senescent Cell Reduction
In November 2020, Efrati and Hadanny published a study in Aging Cell that generated extraordinary attention in longevity and biohacker communities: a 35-patient prospective trial in healthy aging adults (ages 64+) receiving 60 HBOT sessions showed statistically significant increases in telomere length (average 20% increase) and reductions in senescent T-cell populations (37% reduction in CD28null CD8 T-cells, a marker of immunosenescence). The findings were framed as the first demonstration that a non-pharmacological intervention could reverse cellular hallmarks of aging — telomere elongation is not expected in aging adults under normal conditions, and senescent cell reduction was the primary target of an emerging class of senolytic pharmaceutical interventions.
The paper became the centerpiece of HBOT longevity marketing globally, cited in promotional materials for commercial HBOT centers and biohacker community discussions as evidence that HBOT is "anti-aging at the cellular level." The scientific interest in the findings is legitimate: if HBOT genuinely reverses telomere shortening in aging adults, it would represent a mechanistically significant discovery. The limitations are also significant: n=35 without a randomized control group (all patients received HBOT; there was no sham or active control arm), single-center design, short follow-up (measurements at 1–2 weeks post-treatment), unknown clinical relevance of the telomere and senescent cell markers as outcome measures, and no data on whether the observed changes persist or translate to any health outcome the patients would recognize as meaningful.
Critically, the same group (Efrati lab, Sagol Center) that published this finding also operates the commercial HBOT longevity program that markets to aging adults based substantially on this research. This dual role — research group and commercial operator of the service the research supports — creates a conflict of interest that should be part of any honest evaluation of the 2020 Aging Cell findings. The findings have not been independently replicated as of 2026.
Miller 2015 and Wolf 2012: The DoD TBI Null Results
The two most rigorous trials in the off-label HBOT literature are the Department of Defense-funded studies for mild traumatic brain injury (mTBI) — and both found null results. Miller et al. 2015 enrolled 72 military service members with post-concussion syndrome in a sham-controlled randomized trial, comparing 40 sessions of HBOT (1.5 ATA, which is lower than the medical standard for approved indications but used in the mTBI protocols) against sham HBOT (air at 1.2 ATA, creating the sensation of pressure without therapeutic oxygen levels) and normal care. The primary outcome — post-concussion symptom scores — showed no significant difference between the HBOT and sham groups. Both HBOT and sham improved compared to normal care, consistent with a non-specific treatment effect. Wolf et al. 2012 conducted an earlier similar DoD trial in 50 military TBI patients with comparable results: no significant difference between HBOT and pressurized air sham on symptom outcomes.
The sham-controlled design is particularly important for interpreting these results. TBI symptom scales (Post-Concussion Symptom Scale, neurobehavioral symptom inventory) are highly susceptible to placebo response — the act of receiving any intensive treatment protocol, sitting in a pressurized chamber for 90 minutes daily over 8 weeks with attentive clinical staff, produces measurable improvement in self-reported symptoms regardless of whether oxygen is being delivered at therapeutic levels. The Miller and Wolf trials demonstrated that the improvement seen in open-label HBOT studies and clinic testimonials is primarily the non-specific response to receiving intensive treatment, not a physiological effect of pressurized oxygen at the concentrations used in these protocols.
The response from HBOT proponents to the null DoD trials has been multifaceted: some argue the 1.5 ATA protocols were below the therapeutic threshold (though higher pressures are more difficult to tolerate and have more side effects); others argue that the 8-week protocol was too short; others argue that the sham at 1.2 ATA with compressed air was itself partially therapeutic (the pressure itself, independent of oxygen concentration, may have biological effects — a legitimate but untested claim); and others argue that the outcome measures used don't capture the domains where HBOT provides benefit. Each of these counter-arguments has some scientific legitimacy and also provides an infinite regress that can prevent any negative trial from being accepted as definitive by motivated believers. The null results stand as the best controlled evidence available for mTBI.
Rossignol 2009: The Autism RCT That Launched a Market (n=62)
Daniel Rossignol and colleagues published a 2009 randomized, double-blind, controlled trial of HBOT in children with autism spectrum disorder, enrolling 62 children aged 2–7 and randomizing them to 40 sessions of HBOT (1.3 ATA, 24% oxygen — notably mild pressure and near-normal oxygen concentration, more similar to a slightly pressurized air environment than to medical HBOT) or sham (1.03 ATA, 21% oxygen, atmospheric pressure with slight ambient concentration change). The primary outcomes were parent and clinician behavior ratings. The HBOT group showed statistically significant improvements on several behavioral measures including irritability, stereotypy, hyperactivity, and social responsiveness scales.
The Rossignol 2009 trial entered autism parent communities during the peak of the DAN! (Defeat Autism Now!) movement and became one of the most-cited peer-reviewed studies supporting biological intervention in autism. Its influence was amplified by its publication in a peer-reviewed journal (BMC Pediatrics), its randomized and double-blind design, and the credibility of having a controlled comparator — features that distinguished it from most autism treatment claims. The trial drove adoption of home HBOT chambers (mild pressure, 1.3 ATA) and clinical HBOT protocols in the autism parent community through the mid-2010s.
The Rossignol study has not been independently replicated in a trial of comparable or larger size. A subsequent Cochrane systematic review of HBOT for autism [1] found insufficient evidence to recommend HBOT for ASD and noted that the Rossignol trial's mild pressure protocol (1.3 ATA, 24% oxygen) raises questions about whether any benefit is attributable to hyperbaric oxygen specifically or to non-specific effects of the treatment setting. The field has largely moved past the autism HBOT chapter — the major autism research centers did not adopt it, and the DAN! framework has been substantially revised — but the Rossignol study remains actively cited in parent communities and by clinics offering off-label HBOT for autism.
- 2016