Hyperbaric oxygen in the treatment of childhood autism: a randomised controlled trial

Hyperbaric Oxygen therapy (HBOT) has risen in recent years as an “alternative” therapy for many conditions, autism included. The logic behind HBOT is rather fuzzy. For example, there was some discussion of using HBOT to reduce oxidative stress a few years back. How increasing oxygen in the body would decrease oxidative stress was not clear. Some other discussions focused on oxygen perfusion. Basically, some studies have shown that some areas of the brain may be getting less oxygen in autistics than in non-autistics. The idea was that increasing the oxygen to those areas might result in some improvement in some measure or another.

This begs the question: are there areas of higher perfusion (hyperperfusion) in the brains of autistics? Seems an important question to pose when proposing increasing perfusion. But one can not find the term “hyperperfusion” in this review promoting HBOT and autism, for example. But the answer is, yes, people have measured hyperperfusion in autistic’s brains:

Brain perfusion SPECT and EEG findings in children with autism spectrum disorders and medically intractable epilepsy

In specific, they found

The areas of hypoperfusion and EEG focus were highly related in seven of 12 children, while the areas of hyperperfusion were highly related to EEG focus in six of 12 children. The areas of hypoperfusion were highly related to the focus observed on EEG, but were not always related.

Using the simplistic logic of HBOT/autism promotors, one then is left with the question of whether could HBOT make seizure activity worse? I wouldn’t put too much weight on this question other than to point out that it isn’t 100% clear that there can be no downside to HBOT. The logic “there is hypoperfusion therefore HBOT should benefit” isn’t strong; the idea that “there are areas of hyperperfusion, therefore HBOT could have a downside” is also not strong. There are three other studies mentioning hyperperfusion and autism. And, I was interested to see that there are 350 hits for a search of hyperperfusion and epilepsy in pubmed. Compare this to 30 hits for autism and hypoperfusion.

Back to HBOT. There isn’t much science for HBOT, to be frank. Most of the momentum, at least in publications, is from one source: Dan Rossignol. An early paper: Hyperbaric oxygen therapy may improve symptoms in autistic children. by Dr. Rossignol was published in Medical Hypotheses–a pseudo medical journal. I believe Dr. Rossignol’s clinic in Florida provides HBOT.

While there have been articles like the above and some small open label study reports, true randomized trials have been lacking. A recent review Hyperbaric oxygen therapy for treatment of children with autism: a systematic review of randomized trials reported

While some uncontrolled and controlled studies suggested that HBO therapy is effective for the treatment of autism, these promising effects are not replicated. Therefore, sham-controlled studies with rigorous methodology are required to be conducted in order to provide scientific evidence-based HBO therapy for autism treatment.

Also worth noting is that HBOT is currently rated as “non-accepted” by the European Committee for Hyperbaric Medicine. An indication that there was not good evidence either way at the time they prepared their statement.

Two recent studies (the one which is the focus of this article and another) have used a more randomized/blind methodology and one has looked at biomarkers considered important in HBOT and autism (cytokines). The results have not been encouraging.

First the more biomarker based study. This group studied autistic children given HBOT and looked at cytokine levels (Brief report: Hyperbaric oxygen therapy (HBOT) in children with autism spectrum disorder: a clinical trial.) Per that study,

Ten children completed 80 sessions of HBOT and all improved by 2 points on the clinician-rated CGI-I scale (much improved) as well as several parent-completed measures of behavior. The lack of a control group limits the ability to determine if improvements were related to HBOT.

and also:

“Although this study was limited by the small sample size and by the variable nature of cytokines, we found no evidence that HBOT affects cytokine levels or that cytokine levels were associated with behavioral changes”

So, if there is a benefit from HBOT, it isn’t due to changes in cytokines. Which HBOT doesn’t seem to affect.

Another somewhat recent study attempted a clinical trial as well
Controlled evaluation of the effects of hyperbaric oxygen therapy on the behavior of 16 children with autism spectrum disorders. This study, out of the old “Thoughtful House” including Andrew Wakefield as an author found ” No consistent effects were observed across any group or within any individual participant, demonstrating that HBOT was not an effective treatment for the participants in this study. This study represents the first relatively large-scale controlled study evaluating the effects of HBOT at the level of the individual participant, on a wide array of behaviors.”

Finally, a study out in the past couple of months again attempts a randomized controlled study: “Hyperbaric oxygen in the treatment of childhood autism: a randomised controlled trial.” The study is out of Thailand. One factor of note is the attempt to do a real control using a “sham” air group. Obviously HBOT studies are complicated in that study subjects can easily detect the changes in pressure. Of note, HBOT in this case is 100% oxygen:

This study was a prospective, randomised, double-blind, controlled trial of HBOT at 153 kPa (1.5 ATA) with 100% oxygen for one hour daily, weekdays to a total of 20 sessions, versus a sham air treatment consisting of pressurised room air at 116 kPa (1.15 ATA) on the same schedule.

While some reports have used 100% oxygen at 1.5ATA, many have used either air or enriched air and sometimes lower pressure (1.3ATA). I.e. this study involves higher oxygen exposure than in many studies. Air is about 21% oxygen. So, a 1.5 atmospheres of pure oxygen is about 7.5 times the oxygen partial pressure in air. Many reports in early studies were about 1.3 ATA air or slightly enriched air. 1.3 ATA (atmospheres absolute) at 25% O2 is about 32% oxygen. For divers, these levels of oxygen are comparable to Nitrox or enriched air. One can get a higher oxygen level from a mask at 1 atmosphere. Which has been a critique of HBOT from the start. Early anecdotal reports from HBOT practitioners claimed that oxygen delivered by mask was not effective. Only high pressures gave whatever results were claimed. Which is counter intuitive to the simple explanations of how HBOT should work.

But, with both high pressure and pure O2, the Thailand study should provide clarity in these questions. Which begs the question, what are those results? The full paper is online and the abstract is below.

BACKGROUND:
Promising results with hyperbaric therapy for children with autism have been reported, but most involved the use of only mild pressure with oxygen supplementation. To date, there has been no randomised, blinded trial of 100% oxygen administered at hyperbaric pressure. This study evaluated the efficacy of hyperbaric oxygen therapy (HBOT).

METHODS:
Sixty Thai children with autism, aged three to nine years, were randomly assigned to receive 20 one-hour sessions of either HBOT at 153 kPa (1.5 ATA) or sham air at 116 kPa (1.15 ATA). Effects on behaviour were measured using the Autism Treatment Evaluation Checklist score (ATEC) and clinical improvement was measured with the Clinical Global Impression (CGI) system; in particular the clinical change (CGIC) and severity (CGIS) sub-scores. These were evaluated by parents and clinicians, both of whom were blinded to the actual exposure.

RESULTS:
The mean total ATEC scores by both parents and clinicians were significantly improved after intervention in both arms of the study compared to the score before intervention (P <; 0.001 in both groups by parents, P = 0.015 in HBOT group and P = 0.004 in sham group by clinician). There were no statistically significant differences in average percentage changes of total ATEC score and all subscales scores when comparing the HBOT and sham air groups, either by parents or clinicians. Changes in the CGI scores following intervention were inconsistent between parents and clinicians. For severity scores (CGIS), parents rated their children as more improved following HBOT (P = 0.005), while the clinicians found no significant differences (P = 0.10). On the other hand, for change scores (CGIC) the clinicians indicated greater improvement following HBOT (P = 0.03), but the parents found no such difference (P = 0.28).

CONCLUSIONS:
Children with autism who received 20 sessions of either HBOT or a sham air exposure had significant improvements in overall behaviour but there were no significant differences in improvement between groups. The inconsistent changes on CGI sub-scores between parents and clinicians are difficult to interpret, but no overall clinically significant benefit from HBOT could be shown. Both interventions were safe and well tolerated with minimal side effect from middle ear barotraumas.

Repeat for emphasis: “Children with autism who received 20 sessions of either HBOT or a sham air exposure had significant improvements in overall behaviour but there were no significant differences in improvement between groups”

Reports from parents and clinicians did not agree and the authors conclude “no overall clinically significant benefit from HBOT could be shown”.

HBOT is not cheap. A single “dive” can cost in the neighborhood of $100. Parents have purchased portable chambers which run in the $10-20k range (depending on model and whether new). And have modified these to provide O2 enriched air, outside the manufacturer’s specifications. There is a resale value in these chambers so far, so the net cost is not going to be as high. But, all told, there is substantial outlay of funds and time in HBOT. The science pro is shaky at best. And now there are two negative controlled trials.

These results will likely do little to dampen the enthusiasm for HBOT. All studies of HBOT and autism in clinicaltrials.gov are completed, so future data may not be forthcoming.

—
By Matt Carey