Dose-Response Relationship of Therapeutic Oxygen: More Is Not Necessarily Better and May Be Inferior to No Supplemental Oxygen – Part 2: Implications and Consequences

Abstract

In the late 1980s to mid-1990s, the authors had consistently positive experience in the treatment of typical sports injuries and cosmetic surgical wounds with hyperbaric oxygen. The treatments in these cases generally consisted of oxygen at 2.0 atmospheres absolute (ATA) for 60 minutes. We were thus surprised to find that experts in the field of hyperbaric medicine did not believe this modality to be effective for such normal wounds. Consequently, we asked Eric Kindwall, M.D., a professional acquaintance and published proponent of this prevalent view, what the bases for his beliefs on this matter were. Starting with references provided by Dr. Kindwall, we began an extensive literature review to try to reconcile our practical successes with the prevailing hyperbaric medical dogma. The finding that the outcomes of oxygen therapy have a hormetic dose-response relationship explains the mistaken conclusions drawn concerning the treatment of uncompromised or normal wounds. Further, consideration of this and related facts provides insights which can assist in a more accurate understanding of published findings relating to hyperoxic therapy and in optimizing clinical treatments conducted with hyperoxic gases at both normobaric and hyperbaric pressures.

1. Introduction

This chapter is the second part of a publication discussing the aspects of therapeutic oxygen dose-response. Part 1 presents verification for the existence of oxygen dose-response and its nature which is hormetic. Part 2 below presents characteristics of the oxygen dose-response; the implications these have for the conduct of oxygen therapy at both normobaric and hyperbaric pressures; several beneficial applications of knowledge about oxygen dose-response.

2. Characteristics of oxygen dose-response

2.1 Oxygen dose-response is not static

In considering oxygen dose-response from the standpoint of wound healing, the perspectives of critical importance are the characteristics of the tissue of particular interest over time as suggested by Barr and Perrins [1]. Thus, though the dose of oxygen for a hyperoxic treatment may typically be thought of and prescribed as the partial pressure of inspired oxygen, the duration of the treatments, the frequency of the treatments, and the total number of treatments, these factors do not necessarily relate to the tissue of interest insofar as the outcome of the treatment is concerned (Figure 1). Rather, there must be a body of experience that the prescribing physician is familiar with and/or confident in that enables him to judge that the oxygen treatment dose parameters will provide the desired outcome for the specific wound he wishes to address. The oxygen pressure in the arterial blood will be factored at the wound site by any compromise of blood flow to that site because of capillary and/or other vascular damage, by any compromise of oxygen diffusion from the capillaries resulting from edema and material such as fibrin in the interstitial spaces local to that wound site, and by the degree of infection and inflammation in the wounded tissue which can greatly influence the amount of oxygen needed to effectively support healing. This view dictates that:

1. Uncompromised wounds to a particular tissue will be treated optimally with lower doses of oxygen than wounds to the same tissue where oxygen delivery has been compromised by such things as circulatory disruption, edema, and inflammation (Figure 2). This is because at a given partial pressure of oxygen, the amount of oxygen delivered to a wound site is inversely proportional to the degree of compromise to the circulation, diffusion, and inflammation at that site.

2. As events such as angiogenesis and the reduction of edema and/or inflammation occur during treatment at a wound site, relative local oxygen delivery will increase, and the optimal inspired oxygen level for therapy will decline correspondingly [2, 3]. This produces a leftward shift in the optimal dose-response relationship as shown in Figure 2.

3. Where tissues have inherently different perfusion rates and, therefore, oxygen deliveries, the tissue with the higher natural delivery of oxygen, independent of compromise (i.e., high perfusion dose-response in Figure 2), will most often be optimally treated with lower doses of oxygen than tissues with lower natural delivery of oxygen (i.e., Figure 2). As an example, Neubauer [4] determined that appropriate oxygen treatment pressures for a central nervous system (i.e., CNS) condition such as multiple sclerosis (MS) was 1.5–2.0 ATA. HBO₂ for other tissues and organs with inherently lower oxygen deliveries than the brain is invariably conducted between 2.0 and 3.0 ATA.

Figure 1.

Relationship between inspired oxygen dose and the actual local dose of oxygen at the wound site of concern.

Figure 2.

Shift in oxygen dose-response relationship resulting from differences in local tissue oxygen delivery (circa 2006 by the authors). In the three cases illustrated, the left-most curve represents the tissue with the greatest perfusion rate or a particular tissue with the least compromise of oxygen delivery. The middle curve represents the tissue with the middle perfusion rate of the three or the particular tissue with the intermediate compromise of oxygen delivery. The right-hand curve represents the tissue with the lowest perfusion rate of the three or the particular tissue with the greatest degree of compromise to oxygen delivery of the three cases.

Point 1 above is implicit in our derivation of the oxygen dose-response relationship. In research, it has been demonstrated by Karapetian et al. [5] in their study of the healing of mandibular fractures in rats with adjunctive HBO₂. With uncompromised fractures, HBO₂ administered at 2.0 atm improved healing while HBO₂ delivered at 2.5 atm hindered it [5]. If the fractures were compromised by inflammation, however, Karapetian and associates suggest that HBO₂ delivered at 2.5 atm would be more effective, thus predicting a rightward shift in the dose-response relationship with greater initial compromise to oxygen delivery.

Point 2 follows logically and predicts a leftward shift in the oxygen dose-response relationship as healing occurs (Figure 7, Part 1). Such a shift has been demonstrated in research conducted by Quirinia and Viidik [6] and Sadeghani et al. [7] using the healing of ischemic flaps and incisional wounds in rats, respectively. Hyperbaric oxygen treatments conducted once a day at 2.4 ATA for 90 minutes over 3 days produced benefits in almost all force parameters of healing ischemic flaps in comparison to the untreated control animals [6]. After 9 days of HBO₂ with the hyperoxic treatment parameters as described previously, however, the force parameters of the HBO₂-treated wounds fell below those of the controls [6]. Thus, as healing progressed, it seems that the effective dose-response curve shifted to the left. Sadeghani and associates, on the other hand, created 2.5 cm incisions on the backs of rats with the surrounding skin dissected from the subcutaneous tissue. At 8 days post-wounding, the bursting strengths of the wounds treated with 1.0 atm of O₂ and 2.5 atm of O₂ were very close. At 14 days post-wounding, however, the wounds healed with the 1.0 atm O₂ treatments were significantly stronger than those healed with the 2.5 atm O₂ treatments [7] indicating a shift in the oxygen dose response to the left. With healing in progress in both of these studies, it is only logical that the dose-response shift be to the left so that continued hyperbaric oxygen treatments using the higher treatment parameters produced adverse outcomes rather than beneficial ones.

Point 3 above also suggests a set of dose-response curves like those shown in Figure 2. In this case, however, the curves represent different uncompromised tissues where the leftmost curve is for the tissue with the greatest perfusion rate, and the rightmost curve is for the tissue with the lowest perfusion rate. Local oxygen consumption will of course be an important factor and could impose shifts in the curves for specific tissues. This does not change the basic nature of the relationships, however.

This latter aspect of our dose-response model has been demonstrated practically. Based on the research of Holbach and associates [8, 9, 10], Pierce and Jacobson [11] concluded that lower pressures were advisable for the treatment of brain edema and recommended a range of 1.5–2.0 ATA. A similar range of hyperbaric oxygen therapy has commonly been utilized for several cerebral neurological applications including stroke [12] and multiple sclerosis [4]. In the latter study, 1.5 ATA was found to be a more effective treatment pressure than 2.0 ATA. Highly relevant to this situation, we believe, is the fact that the brain typically receives 15–20% of the total cardiac output in humans though its mass makes up only 2.5% of the total body mass, on average. Normal clinical pressures for the treatment of systemic conditions with hyperbaric oxygen therapy such as those endorsed by the Undersea and Hyperbaric Medical Society (UHMS) commonly range from 2.0 to 2.8 ATA (UHMS Committee Report HBO Indications 14th Edition [13]).

2.2 Increased pressure is not essential with hyperoxic PO2’s ≤ 1 ATA

Between 2005 and 2011, studies utilizing mild hyperbaric oxygen therapy (mHBO₂) in inflatable, zip-up “chambers” (Figure 3) were conducted for autism [14, 15, 16, 17, 18]. These chambers are typically flushed with enriched air having approximately 24% oxygen at 1.3 ATA, giving an inspired partial pressure of oxygen of 0.312 atm. The results of such studies, except for the two conducted by the Center for Autism and Related Disorders (CARD) whose business is to provide psychosocial treatments for autism based on the principals of applied behavior analysis (ABA) [17, 18], have not been appreciably different than those of HBO₂ at greater pressures. This was one of the bases upon which we decided to try hyperoxic therapy at normal barometric pressure for ASD [19, 20].

Figure 3.

“Mild” hyperbaric oxygen chamber (courtesy of Atlanta Hyperbaric Center, Smyrna, GA, USA).

Research has also been conducted for other neurological conditions with hyperbaric oxygen therapy having sham controls at increased pressure. These control subjects have typically breathed air in standard clinical chambers with a partial pressure of oxygen greater than 0.21 atm (e.g., air at 1.3 ATA for a P_iO₂ of 0.272 atm). The research on autism using mHBO₂ and the controls for other neurological studies conducted with air at an increased pressure, though never exceeding an inspired oxygen pressure of air at 1.5 ATA (i.e., ≈0.314 atm) that we are aware of, have shown that such treatments may be beneficial for addressing at least some brain disorders. These include autism [21], mild traumatic brain injury [22] and persistent post-concussion symptoms in military personnel [23], chronic stroke [24], and cerebral palsy (CP)¹ (Table 1) [25, 26].

	Ref.	Condition	Control		Test HBO2		Best; result
			P; (ATA)	PiO2; (ATM)	P; (ATA)	PiO2; (ATM)
Sampanthavivat et al. [21]	[25]	ASD	1.15	0.241	1.50	1.50	NO DIFF
Wolf et al. [22]	[26]	Mild TBI	1.30	0.272	2.40	2.40	CONTROL
Miller et al. [23]	[27]	Post-Concussion	1.20	0.251	1.50	1.50	HBO2
Anderson et al. [24]	[28]	Chronic Stroke	1.50	0.314	1.50	1.50	CONTROL
Collet et al. [25]	[29]	Cerebral Palsy	1.30	0.272	1.75	1.75	NO DIFF

Table 1.

Studies conducted for neurological conditions with blinded sham controls treated with air at elevated pressures in a whole-body chamber. It is important to note that in all these studies, both the hyperbaric sham air and test oxygen treatments produced notable benefits in comparison to no treatments at all. Further, there was no statistically important difference between the outcomes of the test and sham treatments.

It is most interesting that in these studies, the control and hyperbaric oxygen test conditions provided benefits not appreciably different from each other, and both were clearly better than no treatments at all. Invariably, however, the principal investigators considering only their own study’s outcome each concluded that since the test hyperbaric oxygen treatments were not significantly better than the presumed hyperbaric hyperoxic placebo treatments, neither the control nor test treatments were of benefit. The explanations given for these outcomes were Hawthorne effects [22]; placebo effects [23]; participation effects [25]; an incomplete study, though the results obtained weighed heavily in favor of the control-treated stroke patients having the best result [24]; and the parents of autistic children overrating their child’s improvement because they were extremely anxious for them to make progress [21]. Thus, neither the hyperoxic hyperbaric control nor test treatments were given any credit, whatsoever, for improving the symptoms of the afflicted subjects. Taken as a group, however, we believe the impact made by the hyperoxic control treatments on the variety of cases studied is too much for coincidence. Further, in our own experience, the parents of one child with autism treated with our normobaric hyperoxic therapy were quite conscious of their desire for their child to improve so were very careful not to overrate his progress. Rather, they underrated it until the evidence of change overwhelmed their doubts.

A critical issue for the practical application of low-dose hyperoxia (i.e., a P_iO₂ less than or equal to 1.0 atm of oxygen) was whether or not increased pressures as provided by whole-body hyperbaric chambers in HBO₂ and mHBO₂ are necessary to achieve the desired outcomes. We believed increased ambient pressure would not be essential for conditions other than gas embolism and decompression sickness and determined to test this hypothesis [19].

Following a number of pilot studies of normobaric hyperoxia for autism [19, 20], cosmetic surgery (Figures 4 and 5)² [28], and hair transplant surgery (Figure 6)³ [29], some of which produced remarkable outcomes, it appears that for such cases, no matter how much below 1.0 atm the dose of oxygen is, it is the oxygen pressure, alone, and not hydrostatic pressure that is critical to producing these outcomes (Figure 6).

Figure 4.

A typical recovery following cosmetic sculpting of the face with autologous fat injection without adjunctive normobaric oxygen therapy being administered. Note bruising under eyes at one-week post-surgery. Source: Courtesy of James Goodnight, M.D., Everlasting Beauty, Ridgewood, N.J.

Figure 5.

One patient's transformation over the course of a single day: Times indicated are relative to the surgical procedure. This case illustrates the absence of bruising and rapid reduction in swelling following cosmetic sculpturing of the face with autologous fat injection and normobaric oxygen therapy being administered. The final picture is at 1 day post-surgery, not 1 week (courtesy of James Goodnight, M.D., Everlasting Beauty, Ridgewood, N.J.).

Figure 6.

Application of normobaric oxygen therapy following follicular unit extraction (FUE) hair transplant surgery: A. Immediately prior to the surgery. B. Immediately post-surgery. C. One day post-surgery. D. Eleven weeks post-surgery. At eleven weeks post-surgery, the hair transplant surgeon found it incredible that the patient was thrilled with the results so quickly and considered his appearance to be already “acceptable,” to the point that he no longer felt it necessary to wear a hat. The patient also said that for the first time in the last 16 years (i.e., since he was 22-years old), he “could actually spike his hair up” (courtesy of Jon Ballon, M.D., Esthetics Hair Restoration, Alpharetta, GA).

To date, properly controlled studies have not yet been carried out to establish the validity of such claims to current scientific standards. In the meantime, however, a 2019 publication by Ding and colleagues [27] has provided independent support for the potential offered by normobaric hyperoxic therapy administered at 1.0 ATA (i.e., NBO₂) for treating acute stroke and the associated ischemic penumbra as well as short-term symptoms and long-term outcomes for other conditions produced by chronic circulation insufficiency (CCCI). Other research by Van Allen et al. [28] indicates that normobaric oxygen therapy benefits bladder dysfunction in cases with multiple sclerosis.

3. Applications of knowledge of oxygen dose response

3.1 Treating wounds at normobaric pressure before they become intractable

In view of Kulonen and associates’ [29] research demonstrating the capacity of improved healing of normal wounds with hyperoxia at 1 ATA (i.e., 35 and 70% oxygen in comparison to 21% oxygen as in air), there could be capacity for P_iO₂s up to 1 atm to improve wound healing when there is some degree of compromise to oxygen delivery. Thus, it may be possible to head off problem wounds (i.e., non-healing wounds) with hyperoxic therapy administered at 1 ATA before such wounds progress to a point where routine care must give way to hyperbaric oxygen therapy (e.g., Wagner Grade 3 and Grade 4 diabetic ulcers). With the great difference in cost, convenience, and risk between normobaric oxygen and hyperbaric oxygen [19], the possibility of heading off intractable wounds before they require hyperbaric oxygen therapy would seem to present considerable opportunity to reduce overall healthcare costs and risk to patients as well as improve the patients’ ongoing quality of life. Confirmation of this possibility has been reported by Moon et al. [30], who have concluded that normobaric oxygen therapy significantly augments tissue oxygen pressure in the feet of human diabetic patients and may reinforce wound healing potential via impact on multiple healing processes.

Regarding pathological brain conditions, as shown with our autism spectrum disorder (ASD) pilot studies with normobaric oxygen therapy and almost 11 years of follow-up without regression at the time of this publication [19, 20], normobaric oxygen therapy may provide an opportunity to decrease costs and risk, and to improve convenience in treating certain neurological disorders. Conditions such as mild traumatic brain injury [22] and persistent post-concussion symptoms [23], chronic stroke [24], and cerebral palsy (CP)¹ [25, 26] have research findings with HBO₂ that parallel those of ASD. Further, in view of the sensitivity of the brain to changes in oxygen and the authors’ results with ASD [19, 20], it is likely that a course of hyperbaric oxygen therapy at any increased atmospheric pressure may ultimately provide too great a partial pressure of oxygen (i.e., greater than 1.0 atm) to carry therapies of such conditions to their optimal conclusions [27, 31].

3.2 More accurate interpretation of research

One incident stands out as illustrating the importance of understanding the general nature of oxygen dose response to more accurately interpret the results of research relating to clinical HBO₂. In late 2004, Djillali Annane and associates in France published the results of a randomized, prospective, double-blind study of hyperbaric oxygen therapy of osteoradionecrosis of the mandible (ORN) in the Journal of Clinical Oncology [32]. The study protocol included two 90-minute oxygen treatments at 2.4 ATA per day to minimize patient inconvenience and expenses related to the overall length of time to administer the requisite number of treatments for the established Marx therapy protocol [33] on a once-a-day basis.

When the results of the study were revealed and examined at an interim point as originally planned, however, it was discovered that the HBO₂-treated group was doing less well (i.e., only 19% success) than the untreated control group (i.e., 32% success) [32]. Though this relatively large difference was not statistically significant, the test cases were clearly doing no better than the controls. Thus, the study was terminated for ethical reasons [32].

After the study was published, Blue Cross/Blue Shield in several U.S. states announced that they would stop or actually did stop reimbursing for the treatment of osteoradionecrosis of the mandible with hyperbaric oxygen therapy. This was despite the facts that:

• A randomized, prospective clinical study established the effectiveness of the protocol developed by Marx for the treatment of ORN based on aggressive surgical principles and sequential HBO₂once a day [34].

• ORN was one of the relatively limited number of conditions for which HBO₂ was recognized by CMS and reimbursed for by Medicare, Medicaid, and other third-party health insurers in the U.S.

• ORN was one of the several most common conditions for which HBO₂ was prescribed, and its treatment with HBO₂ in the U.S. had been successfully conducted for many years.

In their attempts to explain Annane’s results in a positive light on behalf of the hyperbaric medical community in the U.S., several prominent members of the Undersea and Hyperbaric Medical Society (i.e., Richard E. Moon, M.D., a former president of the of the Society, together with two university-associated oral and maxillofacial surgeons; and John J. Feldmeier, D.O., who was to become president of the Society some years later) pointed out in separate communications to the editor of the UHMS’s journal, Undersea & Hyperbaric Medicine, that the general standard of care of Annane’s patients was below the existing level in the United States [35, 36]. In doing this, they made a strong case that a missing element in the study approach was aggressive debridement of the wounds, which had been shown by Marx to be essential for the effectiveness of HBO₂ in the treatment of osteonecrosis of the mandible [33]. On this basis, it would seem Marx’s and other positive research conducted in the U.S., as well as the long history of successful treatment of osteoradionecrosis with HBO₂, prevailed, and this indication continues to be recognized and reimbursed by health care insurers in the U.S. including Blue Cross and Blue Shield.

In another letter to the editor re Annane’s study [37], however, as well as in Feldmeier’s mentioned above [35], attention was drawn to the administration of two treatments a day in the study while Marx’s protocol calls for one treatment per day for twice the number of days. Feldmeier, though, seemed to excuse this deviation, pointing out that at least one well-known hyperbaric medical center in the U.S. routinely treats patients twice a day for the same reason as Annane did [32, 35]. Regardless, Annane responded to these queries about the use of two treatments per day in print saying that “there is no evidence from both experimental and clinical studies to suggest that one HBO₂ session per day is superior to two sessions per day” [38]. Obviously, this reply by Annane would seem to be in ignorance of and contrary to the outcome of the study of Barth et al. [39] and other publications concerning HBO₂ and bone healing, but it was never formally rebutted by any party we are aware of. In addition, angiogenesis, the promotion of which is the objective of the Marx treatment protocol for ORN, like any physiological process, requires time to occur. Thus, unless the rate of angiogenesis were doubled by two treatments per day, which seems highly unlikely, the ultimate outcome of HBO₂ administered twice a day would not seem comparable to administration once a day per Marx’s established protocol.

As a final point on this matter, there is the issue of Annane’s study producing substantially poorer results, though not statistically significantly so, for the HBO₂-treated cases of ORN [32]. Annane and his associates never addressed this matter, and as Marx’s explanation for the benefits provided by HBO₂ for the treatment of osteoradionecrosis was alteration of the existing hypoxia, hypocellularity, and hypovascularity [33], it is difficult to accept that even without adequate debridement, the HBO₂-treated cases would do less well than the placebo control cases rather than at least as well. Feldmeier [35] seems to have had the same thought, though he expressed it in opposite terms, noting that Annane and associates [32] offered no explanation for why the placebo arm of their study had a notably higher though not statistically significant rate of recovery from ORN. Based on the biphasic nature of the oxygen dose-response for the clinical use of oxygen established in this report and the two treatments per day administered in the Annane study, too much oxygen would certainly seem the logical explanation for why the HBO₂-treated patients in that study did not do as well as the untreated controls.

As this situation now stands, and in view of the evidence assembled in this report relating to oxygen dose-response, it seems there are still unanswered questions concerning the efficacy of HBO₂ for osteoradionecrosis of the mandible related to two treatments per day. Research comparing the outcomes of one and two HBO₂ treatments per day for ORN with the appropriate standards of care, and other conditions for which two treatments per day are used but not validated through controlled clinical research, would seem appropriate to conduct to clarify the uncertainties about such therapeutic practices currently in use or that are being considered to be put into practice.

3.3 Treatment of sports injuries

We trust the lengthy discussion at the beginning of Part 1 of this publication has made the point that normal wounds, such as sports injuries, can be effectively treated with hyperoxia at an appropriate dose. There is one “injury” that has been frequently used as representative of sports injuries. This is delayed onset muscle soreness (DOMS). For the purpose of rounding out the discussion of normal sports injuries in this report, further comment on the treatment of this condition is warranted.

3.3.1 Delayed onset muscle soreness

DOMS is a mild muscle over-use or unaccustomed-use injury that is convenient to design research around for either human or animal subjects. Thus, when the authors’ efforts in the mid-1990s began to recruit professional sports teams in the United Kingdom and North America to utilize hyperbaric oxygen therapy to treat their injured athletes, a number of DOMS studies were initiated⁴ [40, 41, 42, 43, 44, 45, 46, 47, 48, 49]. For the most part, the conclusions of these studies were that HBO₂ was not an effective treatment for DOMS, and a review from the standpoint of evidence-based medicine by Bennett and associates in 2005 [50] found that the studies they included in their analyses provided an insufficient basis upon which to draw conclusions.

When examining these DOMS-based studies [43, 44, 45, 46, 47, 48, 49, 51, 52, 53] from the standpoint of oxygen dose-response, however, it is clear that these were primarily cases of overdosing very minor injuries with hyperoxia, in some cases utilizing full clinical doses for human subjects (i.e., 2.4 ATA oxygen for 120 minutes [44]; 2.5 ATA oxygen for 100 minute [46, 49].

Toward the lower end of the oxygen doses utilized to treat DOMS in human research relating to sports injuries [42, 45, 48], 2.0 atm oxygen for 60 minutes once a day has been an effective dose for a variety of serious but uncompromised sports injuries including severe muscle contusions and strains, as well as ligament and tendon injuries⁵,⁶ [52, 54]. Thus, even these relatively low doses of hyperbaric oxygen used in research would seem to be too high for DOMS. In view of this, perhaps normobaric oxygen or even a lower concentration of oxygen would prove to be an effective regimen for the treatment of DOMS. The Uchimaru study⁴ would seem to establish a dose that under-treats DOMS, however. In that study, 28% oxygen at 1.3 ATA (effectively mHBO₂ with a P_iO₂ of 0.364) did not significantly improve recovery compared to normobaric air. This matter really has little relevance to sports injuries, however. In our substantial experience in this regard, DOMS is not, as it is described by Mekjavic et al. [41] or Delaney and Montgomery [53], an injury that professional athletes have used hyperbaric oxygen therapy to treat or would likely do so, especially because such treatment has not proven to provide any enhancement of healing.

3.3.2 With mild hyperbaric oxygen therapy

Currently, some manufacturers of inflatable, mild HBO₂ “chambers,” with 510(k) clearance as substantially equivalent to the Gamow bag for treatment of acute mountain sickness, if they in fact have any FDA clearance at all [50], have promoted these systems for the treatment of neurological disorders and athletic injuries. Such promotion has attracted the interest of elite athletes who have adopted these systems for home use.⁷

While some studies have used mHBO₂ systems to administer therapy for neurological conditions such as autism spectrum disorders [14, 15, 16], no results of studies on the use of such equipment for typical sports injuries have been published to our knowledge. Rather, the promotion of mHBO₂ systems for the treatment of sports injuries would seem to depend on the results published for research with conventional clinical HBO₂ chambers using appropriate doses of oxygen as discussed above [50].

With respect to physical trauma to soft tissue, connective tissue, and bone, the two forms of treatment are not at all comparable. Further, several studies of sports injuries with control data reasonably like mHBO₂ have shown that the results of P_iO₂s ≤ 1.0 ATA are significantly inferior to HBO₂ treatments conducted at 2.0 and 2.5 ATA breathing 100% oxygen.

Abbadi and Elrefai [55] showed that treating recently occurring acute ankle sprains seven times for 90 minutes, each, over a five-day period produced significant pain relief, with HBO₂ administered at 2.0 ATA (80% pain relief) and 2.5 ATA (97% pain relief), and no pain relief attributable to the oxygen treatments given at 1.0 ATA. In comparison to this last treatment condition, however, typical mHBO₂ has less than one-third the P_iO₂. Thus, based on this study, mHBO₂ could not be expected to provide any benefits over and above no treatment administration.

Chen et al. [56] treated first-degree muscle strains in baseball players with HBO₂ at 2.5 ATA for 70 minutes per treatment twice a week for 10 weeks. Each treatment had two 5-minute air-breaks taken from the 70 minutes and two appended 15-minute periods, one at the start for compression to the treatment pressure, and the other at the end for decompression from the treatment pressure to surface pressure. These subjects were compared to blinded control subjects breathing air at 1.3 ATA on the same treatment schedule, though the control group breathed oxygen during the 15-minute decompression time to surface pressure. The P_iO₂ during the control treatment phase was 0.2724 atm, just slightly below the typical PiO2 of mHBO₂, 0.312 atm. At the end of this course of therapy, the HBO₂-treated subjects were significantly better than the control subjects in most assessment categories and inferior in none. Those assessment categories with superior outcomes were:

• Muscular enzymes including reduced serum levels of CK (creatine kinase), GOT (glutamic-oxaloacetic transaminase), and MB (myoglobin), which were positive indications of recovery from muscular injury.

• Subscales of the Brief Pain Inventory including improved general activity, mood, sleep, and enjoyment of life.

As with the Abbadi and Elrefai [55] research, Chen and associates’ [56] control data failed to show results that would suggest benefits for mHBO₂ in the treatment of typical sports injuries.

Another study to include in this discussion was conducted by Horie et al. [57]. They investigated recovery with HBO₂ from injury to the tibialis anterior muscles of rats induced by injecting cardiotoxin (CTX) into them. The HBO₂ treatments consisted of 120 minutes of oxygen at 2.5 ATA once a day, 5 days per week for 2 weeks. As well as untreated controls, other post-injury management utilized air breathing at 2.5 ATA (having a P_iO₂ of about 0.524 atm) and oxygen breathing at 1.0 ATA, with the other treatment parameters like those for the HBO₂-treated subjects [57]. It should be noted that both the hyperbaric air and normobaric oxygen treatments had P_iO₂s substantially greater than those typical of mHBO₂. Assessments comparing the HBO₂-treated rats to untreated controls, normobaric oxygen-treated animals, and hyperbaric-air-treated animals included histological analyses and measurement of the maximum force producing capacity of the regenerating muscle fibers. The results showed that HBO₂ treatments, but not hyperbaric air or normobaric oxygen treatments, accelerated satellite cell proliferation and myofiber maturation in rat muscle that was injured by a CTX injection, particularly at the end of the 2-week course of treatments. Thus, again, there was no indication of benefits from oxygen breathing conditions somewhat greater than those of mHBO₂.

To this point, a primary focus of this report has been the significant over-treating of normal wounds with hyperoxia and, thus, giving a false impression that such therapy is not beneficial for uncompromised trauma. In the fourth study to be discussed below, however, we have a first case in which all the hyperoxic treatments administered to subjects would appear to have been under-treatments. In this particular study reported by Uchimaru and associates (2010), 28% oxygen at 1.3 ATA (effectively mHBO₂ with a PiO₂ of 0.364 atm) did not significantly improve recovery from DOMS compared to normobaric air. This strongly suggests, therefore, that mHBO₂, usually with a slightly lower PiO2 (i.e., 0.312 ATA), would be an under-treatment for typical, more serious (i.e., than DOMS) sports injuries. On this basis, it seems that mHBO₂ is not likely to benefit any athlete seeking to expedite recovery from an injury he has sustained through participation in sport that we would expect to be more severe than DOMS.

4. Conclusions

Through extensive review of the scientific and clinical literature examining the issue of whether or not hyperbaric oxygen therapy could be an effective treatment for normal wounds (i.e., wounds that heal normally [58], it became evident to the authors that for therapeutic applications, hyperoxia is dose dependent. It also became apparent that the dose-response relationship for the clinical use of oxygen is biphasic in character, with beneficial effects in the lower dose range and negative impact in the higher dose range (i.e., hormetic in nature) when starting with the breathing of room air at one atmosphere absolute. The descending limb at higher doses in the dose-response relationship ultimately produces outcomes inferior to no hyperoxic therapy being administered at all (Figure 4, Part 1). This result is caused by the toxic effects of oxygen, which, in relative overdose, can have negative impact on all cells [59].

Another aspect of the dose-response relationship of hyperoxic therapy is that medical applications have a continuum of doses beginning at just greater than the partial pressure of oxygen in air at one atmosphere (i.e., 0.20954 atm) and extending to 3.0 atm of oxygen, which is the highest partial pressure deemed safe for clinical use over a limited time period. With this model, it is our experience that hydrostatic pressure is only essential to raise the oxygen partial pressure to values greater than 1.0 atm to reach a PiO₂ sufficient to get a positive response from the tissue or tissues being addressed, except for the treatment of gas phase/embolic (i.e., bubble) disorders (i.e., arterial and venous gas embolism, and decompression sickness). In these latter cases, the physical size of the gas phase can be reduced by increased pressure in a whole body chamber to accomplish such things as clearing gas-phase obstructions and thereby improving oxygen delivery to tissues through the circulation, enhancing gas phase elimination by concentrating gas molecules in smaller bubble volumes, and reducing pain produced by the physical size of the gas phase.

Obviously, if it will positively impact a given condition, normobaric oxygen therapy is more convenient, less risky, and less costly to administer than hyperbaric oxygen therapy. As we have shown in our autism studies over the last 11+ years years (as of early 2024), with an appropriate breathing gas assembly, normobaric hyperoxic therapy prescribed by a physician is suitable and safe for use with children in their homes [19] and, thus, we contend, for adults in assisted living situations. The administration, per se, can be easily and properly done by nonprofessionals, including suitably trained parents of children who are receiving the therapy.

In considering conditions that may be effectively treated with hyperoxia at normobaric pressure, some neurological disorders including autism spectrum disorder [19], Alzheimer’s and other dementia [60], traumatic brain injury [22], chronic stroke [24], post-traumatic stress disorder [23], and cerebral palsy [25, 26] have promise. Also, wounds that do not respond to routine care such as diabetic ulcers of the lower extremities might benefit from early intervention with normobaric hyperoxia before they progress to intractable Wagner Grade 3 and 4 wounds that require hyperbaric oxygen therapy for management, with this latter therapy increasing the cost of care and decreasing the quality of life of the patients involved.

In conclusion, understanding the nature of oxygen dose-response and that the healing of normal wounds can be enhanced with oxygen therapy in proper dose in either normobaric or hyperbaric conditions provides opportunities to expand the overall applications of hyperoxic therapy and improve outcomes for patients as well as reduce their overall cost of care. This could include improved treatment protocols for hyperbaric oxygen therapy resulting from the leftward shift of the relevant oxygen dose-response relationship over a course of therapy and the possibility of physicians being able to treat some conditions more effectively with normobaric oxygen therapy or even normobaric gases with less than 1.0 ATA of oxygen.

Acknowledgments

By Russell E. Peterson, Ph.D.: I would like to thank my mother, Pauline W. Peterson, whose interest in medicine and the medical sciences led to my career in diving and hyperbaric medicine and physiology. I would also like to acknowledge my graduate school mentor, Christian J. Lambertsen, M.D., whose guidance to believe in one’s own observations rather than the beliefs of others was a critical element in the evolution of this publication.

By Michael W. Allen: I would like to thank my parents, Phyllis P. and William G. Allen, who blessed me with an inquisitive mind and a need for constant learning. I would also like to thank Philip B. James M.D., Ph.D, my early mentor and lifelong friend who fostered my interest in pressure physiology and the potential offered by the use of oxygen in medicine.

Conflict of interest

The authors are the Founding Partners of Microbaric^® Oxygen Systems, LLC, now known as NEMO Therapeutics, LLC, which was established to provide an organization through which research in applications of normobaric oxygen therapy, particularly autism spectrum disorders, could be investigated and, as possible and appropriate, made available as a mainstream medical therapy for autism and other neurological conditions for which there is currently no effective treatment. This organization has been funded in full by the Founding Partners. No outside funding or income has been received to support the research and writing of this review of oxygen dose response.

Funding statement

The entire literature review, development of concepts, and preparation of this report were funded by the authors. No outside funding was sought or received from any source.

Data availability

All publications cited as journal or book references in this manuscript are available from the National Center for Biotechnology Information, U.S. National Library of Medicine: 8600 Rockville Pike, Bethesda MD, 20894. www.ncbi.nlm.nih.gov/pubmed and other scientific and university databases.