The safety profile of HBOT is strong enough that a Phase II clinical trial for a given neurological disease could be readily initiated without a full Phase I step. One of the most frequent criticisms of HBOT as a therapy for conditions ranging from neurological disorders to age-related diseases is the potential for oxygen toxicity, usually framed in terms of oxygen-induced oxidative stress. We address the mechanistic question of oxidative stress in the Mechanisms section (Mechanisms underlying HBOT efficacy) below; here we summarize the clinical safety data.

Clinical safety data are substantial. A retrospective analysis of 2,334 patients by Hadanny and colleagues [5] found that serious adverse events are infrequent and typically non-serious. A larger retrospective dataset of 1.5 million treatments concluded that “occurrence of adverse events associated with HBO therapy is infrequent and typically not serious. The findings of this study suggest that when administered according to the appropriate therapeutic protocols HBO therapy is a safe and low-risk intervention” [6]. A 2023 systematic review and meta-analysis of 24 randomized controlled trials involving 1,497 participants concluded that “HBOT is more likely to cause adverse reactions when the chamber pressure is above 2.0 ATA” [1 atmosphere absolute (ATA) being the atmospheric pressure of 760 mmHg at sea level] [7]. A longitudinal observational study further reported that pulmonary function is preserved across repeated HBOT sessions, even in patients with pre-existing respiratory disease [8]. The most common events are aural fullness, transient visual changes, fatigue, and occasional patient-related adverse events, such as hypoglycaemia or hypotension before chamber entry.

In 1996, the recognized applications of HBOT included carbon monoxide poisoning, arterial gas embolism, decompression sickness, clostridial myonecrosis, necrotizing fasciitis, refractory osteomyelitis, acute traumatic ischemic injury, compromised skin grafts and flaps, anaemia due to exceptional blood loss, thermal burns, and difficult-to-heal wounds [9]. Three decades later, this therapeutic landscape has not broadened in any decisive way. Real-world use remains constrained by the limited strength of evidence for many indications and by the logistical dependence on dedicated hyperbaric facilities. HBOT has not been widely adopted into routine care across major public-health systems, where reimbursement or commissioning is generally confined to a limited set of approved indications. Apart from decompression illness, HBOT is approved and routinely funded for carbon monoxide poisoning and for the management of selected patients with complex wounds, particularly when conventional treatment alone is insufficient. A list of primary and adjunctive uses for HBOT, drawn directly from the 2023 UHMS recommendations [10] and FDA guidance, is provided in Supplementary Table S1.

Clinical trials and accumulated clinical experience suggest that HBOT may have therapeutic potential beyond its established indications, including in selected CNS conditions. However, the evidence remains insufficient for broad clinical acceptance. Trials conducted largely by a single Israeli research group have reported benefits in persistent post-concussion syndrome and post-stroke cognitive impairment, including medium-to-large effect sizes and improvements on validated cognitive scales [11–13]. These findings are encouraging but need to be amplified by further studies.

Harch (2022) applied the Centre for Evidence-Based Medicine (CEBM) levels of evidence and the American Society of Plastic Surgeons (ASPS) grading framework to this literature, concluding that the 1.5 ATA HBOT protocol meets CEBM Level 1 criteria and supports an ASPS Class A recommendation for persistent post-concussion syndrome [14]. This grading, however, derives from an independent systematic review and should not be interpreted as an official ASPS clinical practice guideline.

A 2022 meta-analysis provided solid evidence that HBOT is beneficial in the treatment of diabetic foot ulcers, particularly by improving healing and reducing the risk of amputation [15]. The UHMS/Wound clinical practice guideline [16] and the DAMO₂CLES randomized multicenter trial [17] together support the narrower and more defensible conclusion that HBOT benefits selected patients with ischemic or refractory diabetic foot ulcers, rather than all patients with diabetic foot ulcers, and this is now reflected in the way most health systems commission the therapy. As discussed below, even if routine use were recommended, ensuring equitable and timely access to HBOT would be difficult because treatment depends on the availability of specialized hyperbaric facilities.

Current guidance from the Wilderness Medical Society does not issue a formal recommendation for HBOT in frostbite [18], and Cochrane reviews have emphasized the absence of randomized controlled trials for this indication. Nevertheless, recent comparative and case-based studies, including those by Magnan and colleagues [19, 20], have suggested improved tissue preservation and digit salvage when HBOT is incorporated into multimodal management, particularly in combination with iloprost, a drug used to treat pulmonary arterial hypertension.

A recent review of HBOT in neurodegenerative disease summarizes preclinical data suggesting plausible benefit across cerebral ischemia, stroke, Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, MS, and amyotrophic lateral sclerosis (ALS) [21]. In addition, Hadanny and colleagues [22] reported cognitive improvements in a randomized controlled trial of healthy older adults.

At present, there is no neurological or neuropsychiatric condition for which HBOT has been conclusively established as an efficacious treatment. There is, however, observational evidence beyond the subjective reports of individuals who describe improved well-being and cognition after treatment. This evidence comes mainly from the experience accumulated in MS, where several thousand patients across long-established oxygen therapy centres in the UK and elsewhere continue to use HBOT and report consistent improvements in activities of daily living, although the studies were not placebo-controlled.

In several neurological conditions, HBOT is supported by an increasing mechanistic rationale and a growing body of preliminary clinical observations, yet robust evidence from adequately powered, indication-specific trials remains limited. As in frostbite, this creates a tension between therapeutic promise and evidentiary standards, such that HBOT may be regarded as a potentially valuable adjunctive strategy while still falling short of broad inclusion in formal clinical guidelines.

In neurological disorders such as ALS, the available evidence is restricted to two case series: a Phase I clinical trial involving only five patients that reported promising results, followed by a Phase II trial that failed to replicate those findings [23, 24]. Overall, the limited available evidence does not support a consistent benefit of HBOT in ALS. This conclusion may have contributed to a decline in further investigation in this area, despite the urgent need for efficacious therapies for diseases for which no treatments have proven effective in the long term.

HBOT has been the subject of investigation in MS for over four decades. Early randomized studies reported benefit in selected patients: Fischer and colleagues [25], in a double-blind placebo-controlled trial of 40 patients conducted between 1980 and 1982, observed improvement in 12 of 17 treated patients compared with 1 of 20 controls, with the clearest effect in less severely affected cases; Neubauer and colleagues [26] reported clinical benefit; and James [27] reviewed the rationale and early clinical experience. These studies have been influential, particularly in shaping patient advocacy and routine practice in the UK network of oxygen therapy centres, but they do not meet current standards for randomization, allocation concealment, blinding integrity, pre-registration, and patient-reported outcome selection.

The 2010 meta-analysis by Bennett and Heard [28], building on the earlier Cochrane review by Kleijnen and Knipschild, pooled nine randomized trials and concluded that the available evidence did not support routine clinical use of HBOT in MS. Their caution was appropriate given the heterogeneity of protocols, endpoints, and disease phenotypes across the pooled studies, and the absence of any single large, modern, multicenter trial. The National Center for Complementary and Integrative Health has reached a similar conclusion.

Continued HBOT use by several thousand people with MS, especially in long-established oxygen therapy centres in the UK, together with consistent uncontrolled reports of benefit in fatigue and daily functioning, represents a signal worth testing in a modern trial. Such a trial should prioritize outcomes that matter to patients, including fatigue, cognition, quality of life, and the Symbol Digit Modalities Test, rather than relying only on traditional disability scales. A complete history of HBOT in MS should cite the Neubauer et al. [26], James [27], and Fischer et al. [25] contributions alongside the later meta-analyses; Murray’s 2022 overview [29] is valuable as a history of MS diagnosis and treatment but does not, by design, cover the hyperbaric literature in depth.

The mechanisms of HBOT in decompression sickness are the best characterized, largely because the physics of gas behaviour under changing pressure is well understood. The way decompression chambers work is a useful starting point for thinking about the further benefits claimed for HBOT in other indications.

Figure 1 should be read not simply as a catalogue of possible HBOT effects, but as a physiological map of how pressure makes oxygen therapeutically available to tissues that are under-supplied. In Oxygen and the Brain, James [32] repeatedly returns to the central proposition that “there is no substitute for oxygen”, especially in the brain, and that the practical purpose of a pressure chamber is to correct oxygen deficiency by increasing the oxygen dissolved in plasma, not merely the oxygen carried by haemoglobin. This distinction is crucial for neurological diseases, because James’s model emphasizes impaired microcirculation, blood-brain-barrier disturbance, inflammation, oedema, and reduced tissue oxygenation as linked barriers to recovery. Figure 1, therefore, provides a visual bridge between the older gas-law logic of hyperbaric medicine and the broader biological mechanisms now proposed for HBOT: improved oxygen diffusion, reduced swelling, restoration of capillary function, modulation of inflammation, and adaptive signalling responses that may support repair in vulnerable CNS tissue.

Generating high-quality evidence for HBOT faces obstacles that are, in part, inherent to the nature of the intervention itself.

The gold standard for therapeutic evaluation—a double-blind, placebo-controlled randomized trial—is challenging, though not impossible, to implement in HBOT research. A key difficulty lies in the design of an appropriate sham condition. In some studies, “control” participants enter identical hyperbaric chambers pressurized to a lower, ostensibly sub-therapeutic level (typically 1.1–1.3 ATA) while breathing slightly enriched air. Yet even small increases in oxygen partial pressure may have physiological effects; changes of as little as 10% have been shown to be biologically active. Such sham conditions, therefore, may not function as true controls. For device-based interventions, blinding can be achieved through the use of an identical but inactivated apparatus, or an activated apparatus fitted with a barrier to block the therapeutic component. However, the sensory experience of pressurization—ear-popping and awareness of an altered environment—may compromise blinding integrity in a proportion of participants, particularly those with prior HBOT exposure. The success of blinding must therefore be formally tested in each trial, as recommended in the Phase-III MS trial methodology literature [44].

HBOT is currently administered across widely varying pressures (1.5–2.4 ATA), session durations, and frequencies, making cross-study comparisons unreliable and limiting the validity of meta-analyses. Before large efficacy trials can be meaningfully designed, dose-finding studies are required.

Outcome selection is critical. Trials must incorporate clinically meaningful primary endpoints. For MS, this means moving beyond traditional disability scales to include fatigue, cognition, and quality of life. The Symbol Digit Modalities Test has emerged as the cognitive measure of choice following comprehensive psychometric analyses across MS disease-modifying registration trials, given its sensitivity to impairment, correlation with brain pathology, and ability to detect clinically meaningful change. For mild cognitive impairment (MCI), parallel neuropsychological batteries with neuroimaging biomarkers as secondary endpoints are needed. Trial duration is equally critical: effects on neuroplasticity may require months to manifest, and trials of insufficient follow-up will systematically underestimate benefit. Comorbidity-adjusted outcome selection, as highlighted in the MS trial literature [45], is particularly important for a chronic multisystem condition of this kind.

A further obstacle is economic. HBOT is a non-patentable procedure, and no pharmaceutical or device company is positioned to recover the cost of large efficacy trials through market exclusivity. As with other non-patentable or repurposed interventions, rigorous evaluation is therefore likely to depend on public, philanthropic, or non-profit funding [46]. This is a textbook market failure: private-sector incentives for off-patent and repurposed therapies are structurally weak [47] and bridging the resulting evidence gap requires coordinated public funding [48].

The appropriate funders are national health research agencies—the National Institutes of Health (NIH), the National Institute for Health and Care Research (NIHR), the European Health and Digital Executive Agency (HaDEA) under Horizon Europe—together with disease-specific foundations such as the Michael J. Fox Foundation, the MS Society, the ALS Association, and the Alzheimer’s Association. From a health-economics perspective, this investment is defensible: HBOT, if proven efficacious, would be a generically deliverable intervention with no ongoing licensing costs, and the societal savings from even modest reductions in disability progression in MS, or delay of dementia conversion in MCI, would far exceed trial costs. Non-patentable therapies may prove cheaper in the long term than patented drugs of equivalent efficacy, making the investment in their rigorous evaluation doubly justified from a public-health standpoint [46, 48].

This section contains the paper’s principal practical recommendation. While the design and funding of definitive randomized controlled trials remain a medium-to-long-term objective, a meaningful and mechanistically informative research agenda can be initiated immediately, at low cost, and without modification of existing clinical protocols. Patients already undergoing routine HBOT sessions represent an underutilized observational opportunity: the controlled, time-limited, and reproducible nature of each session makes it an ideal setting for continuous physiological profiling. Figure 2 summarizes the proposed core parameters, each of which is either already recorded in most hyperbaric facilities or obtainable with inexpensive, widely available equipment.

The core proposal is the systematic collection of the following parameters during each session (Figure 2). Cardiorespiratory continuous monitoring: includes heart rate, respiratory rate, peripheral oxygen saturation (SpO₂), and non-invasive continuous blood pressure (achievable via finger photoplethysmography at negligible additional burden). Metabolic monitoring: glycaemia, body temperature, and end-tidal carbon dioxide (EtCO₂) as a practical proxy for CO₂ production, directly reflecting the ventilatory response to acute hyperoxia. Autonomic response: heart rate variability (HRV), derivable from the electrocardiographic signal without additional hardware, provides a continuous index of autonomic nervous system modulation—a dimension of particular relevance given well-documented dysautonomia in both MS and MCI populations. Cerebral monitoring: includes near-infrared spectroscopy (NIRS) for non-invasive, continuous measurement of cerebral oxygenation and regional haemodynamics, which is fully compatible with the hyperbaric environment; and, where hyperbaric-compatible equipment is available, electroencephalography (EEG) for session-by-session characterization of cortical activity.

Cerebral monitoring deserves particular emphasis. Given that the putative therapeutic rationale for HBOT in neurological conditions centres on correction of cerebral hypoxia and on the promotion of adaptive neuroplasticity—including activity-dependent synaptic remodeling, angiogenesis, modulation of neuroinflammation, and metabolic support for repair—monitoring should extend beyond systemic oxygen exposure to include brain-relevant physiological responses. EEG data would be particularly informative for cognitive outcomes in MCI.

Taken together, even a modest observational dataset, systematically collected across patients, sessions, and clinical profiles, would yield several scientifically valuable outputs. It would characterize the normal physiological trajectory of an HBOT session in neurological patients, identify inter-individual variability in physiological response, generate hypotheses regarding which response patterns correlate with clinical benefit, and provide the biological-plausibility data that funding bodies increasingly require before committing to large-trial investment. Crucially, this program requires no randomization, no control arm, no additional clinical intervention, and no substantial funding—only the systematic application of available technology and a commitment to record, share, and analyze the data thus generated.

This Perspective does not present HBOT as an established neuroprotective therapy. Rather, it makes a narrower and, in our view, more defensible argument: that HBOT’s strong safety record, combined with a plausible—though still not demonstrated—mechanistic rationale, justifies a focused, modern program of clinical and physiological investigation in chronic neurological disease.

MS may represent the most appropriate starting point for such a program, given the existence of previous clinical experience, patient-reported signs of benefit, and a relatively well-defined need for improved symptomatic and disease-monitoring strategies. However, this should not preclude parallel consideration of other chronic neurological disorders, including Alzheimer’s disease and related dementias, where effective disease-modifying therapies remain limited and even modest, reproducible improvements in cognition or function would be clinically meaningful. Because HBOT has a favorable safety profile when delivered under modern protocols, any potential cognitive benefits in these conditions should be evaluated as early and rigorously as possible, rather than dismissed because definitive efficacy data are not yet available.

The practical recommendation is the continuous-monitoring program outlined in the previous section and summarized in Figure 2. It is low-cost, requires no randomization, does not disrupt existing protocols, and would begin generating the mechanistic and biological-plausibility data that funders now reasonably expect before committing to Phase III trials. Alongside this, rigorous, independently led, multicentre efficacy trials—using patient-centred outcomes—remain the essential medium-term objective. Given HBOT’s non-patentable nature, coordinated public funding is likely the only realistic pathway to achieving them.