This page exists because HBOT marketing routinely cites only the positive half of the literature. Here is the whole shape of it.
The full ledger
- Where evidence is established: the FDA-cleared indication list — settled, uncontroversial, and distinct from neurological-recovery claims.
Where trials are positive: a Phase I trial in persistent post-concussive symptoms and PTSD reported symptom, cognitive, and cerebral-blood-flow improvements after forty 1.5-ATA sessions (Harch et al.) [1]. A randomized crossover trial enrolled 56 adults one to five years after mild TBI with persistent post-concussion symptoms: forty 1.5-ATA sessions improved cognitive performance, quality of life, and SPECT-measured brain metabolism — notable because every participant was years past the injury, which is direct trial evidence that the window for change does not close at twelve months (Boussi-Gross et al., 2013) [2]. In severe TBI, a prospective randomized trial found hyperbaric oxygen produced a more robust effect on cerebral metabolism than either normobaric 100% oxygen or standard care (Rockswold et al., Journal of Neurosurgery, 2010) [3]. And in post-COVID condition — a different brain insult with overlapping symptoms — a randomized, sham-controlled, double-blind trial (n=73, 40 sessions) reported gains in global cognition, attention, executive function, energy, sleep, psychiatric symptoms, and pain interference, alongside measurable MRI perfusion and microstructural changes (Zilberman-Itskovich et al., Scientific Reports, 2022) [4].
The record keeps moving, in both directions. In 2025, a double-blind randomized trial gave 47 adults with persistent symptoms after brain injury (40 of them TBI) forty hyperbaric or forty sham sessions: the hyperbaric group improved roughly seven points more on the Neurobehavioral Symptom Inventory (95% CI 1.7–12.3, p=0.01), with specific gains in smell, anxiety, sleep, and vestibular complaints (Weaver et al., Scientific Reports, 2025) [7]. In children aged 8–15 with persistent post-concussion symptoms six months to ten years after injury, a randomized, sham-controlled, double-blind trial (n=25) reported improved cognitive and executive function and behavioral symptoms, with corresponding microstructural change on MRI (Hadanny et al., Scientific Reports, 2022) [8]. And in post-stroke cognitive impairment, a four-arm randomized trial (n=116) found hyperbaric oxygen combined with computerized cognitive training outperformed either alone — and standard care — on global cognition and daily-function scores (Liu et al., Journal of NeuroEngineering and Rehabilitation, 2026) [9]: early factorial evidence that pressure and active training pair, rather than compete.
Where trials are null: the largest sham-controlled trials in service members with mild TBI found no significant difference between hyperbaric and sham arms on cognitive measures or symptom scores — and a rapid evidence assessment across twelve randomized mild-TBI trials reached the same conclusion, while noting that both hyperbaric and sham groups outperformed standard care [5]. Why that reads as unresolved rather than settled: the “sham” in most of these trials was pressurized air at 1.3 atmospheres — an exposure that itself raises dissolved oxygen, and one some practitioners use therapeutically — so the same data supports two readings: a placebo response, or two active arms compared against each other. We don’t pick a side on this page; we report the fork [5]. The 2021 VA/DoD clinical practice guideline weighs those null trials and recommends against HBOT for post-acute mild TBI [10]. And the field’s answer to the sham problem is design: a 420-participant randomized trial in veterans and service members is now underway using a true room-air control at normal pressure, built specifically to remove the ambiguity the pressurized-air shams created (Neumann et al., Trials, 2026) [10]. When it reports, this page will say what it says.
Where trials are negative: for cerebral palsy, the evidence is unusually complete, and it says no. A double-blind randomized trial in 111 children found hyperbaric oxygen and pressurized air produced the same improvement (Collet et al., The Lancet, 2001), and a second trial designed to isolate oxygen from pressure was stopped early for futility (Lacey et al., Annals of Neurology, 2012) [11]. In autism spectrum disorder, the one rigorous randomized trial found no between-group benefit, and no positive finding has been replicated under standard conditions [11]. We cite the negative trials with the same care as the positive ones — a ledger only earns trust if it runs in both directions — and no one at MindForge will propose HBOT on the strength of those diagnoses.
- What that mix means: the effect of pressurized oxygen on neurological recovery remains genuinely contested in the literature — which is why we treat HBOT as an adjunct under measurement, never a headline therapy, and why our own patients' re-test data governs its use in their plan.
The proposed mechanism, plainly
The biology under study is specific. Fluctuating high-dose oxygen — raised during sessions, returned to normal between them — is read by tissue as a relative oxygen swing, activating the HIF-1α oxygen-sensing pathway (the discovery of that machinery won the 2019 Nobel Prize in Physiology or Medicine). Downstream, physiologic studies report VEGF-driven growth of new microvasculature and mobilization of stem cells — on a delayed, cumulative time-course, rising tens of hours into a course rather than after single sessions — and, in the post-COVID trial above, visible perfusion and microstructural change on MRI [4,6]. The plausible target is tissue that is metabolically impaired but not dead: necrotic tissue replaced by fluid has no infrastructure left to rebuild, while under-perfused, under-performing tissue around it does — which is why candidacy here is an evaluation question, never a sales question [6]. One more property matters: the response appears dose-shaped rather than more-is-better — the signal depends on the oxygen fluctuation, not the absolute level — which is why a course is engineered (pressure, session length, air breaks, session count) instead of maximized [6]. Mechanism is not outcome; the trial ledger above is what governs claims.
The risks, stated plainly
Pressurized oxygen is a drug with a dose, and it has side effects. In a retrospective analysis of about 1.5 million treatments, 0.68% were associated with any adverse event; the most common were middle-ear barotrauma and confinement anxiety, and medically severe events were rare — fewer than 0.05 oxygen-toxicity events per 1,000 treatments (Jokinen-Gordon et al., 2017) [12]. A temporary myopic shift (blurrier distance vision) is common across a long course and largely resolves in the weeks after it ends, with small persisting changes reported in some eyes (Riedl et al., Acta Ophthalmologica, 2019) [12]. And a small number of conditions rule pressurized oxygen out entirely — an untreated collapsed lung is the classic example — which is one more reason candidacy is a screening decision made at evaluation, before anything is scheduled.
Our operating posture
Evidence-informed adjunct within measured programs; candidacy clinical; claims limited to what the cited record supports — and this page updates as the literature does.
Sources: [1] Harch et al., Phase I trial, HBOT 1.5 ATA ×40 sessions in PCS/PTSD — per Concussions in Athletics (Springer), future-directions chapter. [2] Boussi-Gross R, et al. PLoS ONE 2013 — randomized crossover trial, n=56, mild TBI 1–5 years post-injury with persistent post-concussion symptoms; 40 sessions at 1.5 ATA; improved cognition (computerized battery), quality of life, and SPECT-measured cerebral metabolism. [3] Rockswold et al., Journal of Neurosurgery 2010 — prospective randomized trial in severe TBI (n=69): hyperbaric oxygen vs. normobaric hyperoxia vs. standard care; the hyperbaric arm showed the more robust effect on cerebral metabolism. [4] Zilberman-Itskovich S, et al. Hyperbaric oxygen therapy improves neurocognitive functions and symptoms of post-COVID condition: randomized controlled trial. Sci Rep. 2022;12:11252 — global cognition, attention, and executive function (d ≈ 0.5), energy, sleep, psychiatric and pain domains, with MRI perfusion and microstructural changes. [5] Sham-controlled military mTBI trials (2.4 ATA ×30 sessions: no difference vs. sham on cognition or symptoms) and a rapid evidence assessment of 12 randomized mild-TBI trials (no HBOT-vs-sham difference; both arms above standard care; shams typically pressurized air at 1.3 ATA) — per Concussions in Athletics (Springer) and hyperbaric literature reviews in the MindForge clinical corpus. [6] Mechanism: HIF-1α oxygen-sensing pathway (2019 Nobel Prize in Physiology or Medicine); hyperoxic–hypoxic paradox, VEGF-mediated angiogenesis, delayed stem-cell mobilization, and the impaired-but-viable tissue target — Sagol Center for Hyperbaric Medicine and Research (Efrati et al.), publications and reviews. [7] Weaver LK, et al. A double-blind randomized trial of hyperbaric oxygen for persistent symptoms after brain injury. Sci Rep. 2025;15:6885. doi:10.1038/s41598-025-86631-6 — 47 adults analyzed (40 with TBI); 40 hyperbaric vs. 40 sham sessions over 12 weeks; Neurobehavioral Symptom Inventory change 10.6 vs. 3.6 (difference 7.0, 95% CI 1.7–12.3, p=0.01); gains in olfaction, anxiety, sleep, and vestibular complaints. [8] Hadanny A, et al. Hyperbaric oxygen therapy in children with post-concussion syndrome improves cognitive and behavioral function: a randomized controlled trial. Sci Rep. 2022;12:15233. doi:10.1038/s41598-022-19395-y — n=25, ages 8–15, persistent symptoms 6 months–10 years post-injury; 60 daily sessions vs. sham; improved general cognition, memory, executive function, and behavioral scores, with MRI microstructural correlates. [9] Liu J, et al. Hyperbaric oxygen therapy combined with computerized cognitive training improves global cognition and functional independence post-stroke: a randomized controlled trial. J NeuroEng Rehabil. 2026;23. doi:10.1186/s12984-026-01992-x — four-arm RCT, n=116, post-stroke cognitive impairment; the combined arm outperformed HBOT alone, training alone, and standard care on MMSE and Barthel Index, with prefrontal functional-connectivity change correlating with cognitive gain. [10] VA/DoD Clinical Practice Guideline: Management and Rehabilitation of Post-Acute Mild Traumatic Brain Injury (2021) — recommends against HBOT for post-acute mTBI based on the null sham-controlled trials. Neumann D, Kumar A, Van Loveren H. Efficacy of hyperbaric oxygen treatment in veterans and service members with traumatic brain injury — study protocol for a blinded three-stage group-sequential randomized controlled trial. Trials. 2026;27. doi:10.1186/s13063-026-09645-z — planned n=420; HBOT at 2.0 ATA vs. room-air sham at 1.0 ATA; primary outcome Neurobehavioral Symptom Inventory. [11] Collet JP, et al. Hyperbaric oxygen for children with cerebral palsy: a randomised multicentre trial. Lancet. 2001 — n=111; HBOT 1.75 ATA vs. pressurized air 1.3 ATA: equal improvement in both arms. Lacey DJ, Stolfi A, Pilati LE. Effects of hyperbaric oxygen on motor function in children with cerebral palsy. Ann Neurol. 2012 — equal-pressure design isolating the oxygen variable; stopped early for futility. Xiong T, et al. Hyperbaric oxygen therapy for people with autism spectrum disorder (ASD). Cochrane Database Syst Rev. 2016 — the one rigorous RCT shows no between-group benefit, with more adverse events. [12] Jokinen-Gordon H, et al. A retrospective analysis of adverse events in hyperbaric oxygen therapy (2012–2015): lessons learned from 1.5 million treatments. Adv Skin Wound Care. 2017;30:125–129. doi:10.1097/01.ASW.0000508712.86959.c9. Riedl P, et al. Myopic shift and lens turbidity following hyperbaric oxygen therapy — a prospective, longitudinal, observational cohort study. Acta Ophthalmol. 2019;97:596–602. doi:10.1111/aos.14010 — myopic shift ≥0.5 D in 77.6% of eyes across 40 treatments, largely resolved by 12-week follow-up.
Frequently asked questions
Does HBOT cure concussion or TBI?
No — no responsible reading of the literature supports "cure," and we don't claim it.
Why offer it at all?
Where the positive-trial signal, patient specifics, and measurement discipline align, an adjunct with plausible mechanism and manageable risk can earn a supervised place — earn being the operative word.
Wondering whether HBOT belongs in your plan? That's decided from your data, never as a standalone course — start with a free 15-minute call.