What the HOT-LoCO trial found about HBOT and long COVID

The HOT-LoCO study is important because it tested HBOT against a sham chamber experience, rather than comparing treatment with doing nothing.

How the study worked

The phase II trial enrolled 80 previously healthy adults with long COVID. Participants were randomly assigned to ten HBOT sessions or sham treatment over six weeks. Participants and investigators were blinded to the allocation.

The HBOT protocol used 100% oxygen at 2.4 bar for 90 minutes. That is not the same protocol as our 1.5 ATA, 60-minute sessions.

The published trial report adds useful detail. The study ran at the hyperbaric outpatient clinic of Karolinska University Hospital between September 2021 and June 2023. Of the 80 participants, 65 were women and 15 were men, split equally between the two groups. The sham group breathed medical air at a much lower pressure, so the experience of entering and sitting in a chamber was similar for both groups. Sessions took place roughly two to three times a week.

The main outcomes came from a standard quality-of-life questionnaire, the RAND-36, focusing on physical functioning and on limitations in everyday roles caused by physical health.

What it found

At 13 weeks, the researchers found no statistically significant difference between HBOT and sham on the two primary RAND-36 outcomes. Both groups improved, but the study did not show that ten HBOT sessions caused more improvement than the placebo procedure.

In numbers, physical functioning scores rose by an average of 9 points with HBOT and 8.59 points with sham, a difference that could easily be chance. Role limitation scores rose by 6.25 and 3.85 points, again without a significant difference.

Side effects were similar in both groups. Adverse events, most often cough and chest discomfort, were reported by 49% of the HBOT group and 44% of the sham group. The authors described HBOT as having a favourable harm profile and noted a difference between men and women in how the groups improved.

You can read the full HOT-LoCO paper in BMJ Open.

What the result means

The study does not prove that every possible HBOT protocol is ineffective for every person with long COVID. It does provide controlled evidence against assuming that a ten-session course will improve long-COVID symptoms.

Other studies have reported more encouraging findings using different designs and longer schedules. Those differences do not cancel HOT-LoCO, and HOT-LoCO does not automatically settle every other protocol. This is why we describe the evidence for long COVID as contested. Our guide to what HBOT is explains the mechanism that researchers are testing.

What did the Dutch registry report?

The best known of those more encouraging studies is a prospective registry published in Scientific Reports in 2025. It followed 232 people with long COVID treated at six hyperbaric centres in the Netherlands. The usual protocol was 40 sessions at 2.4 to 2.5 ATA, five days a week for eight weeks.

Three months after treatment, average quality-of-life scores had risen on both the mental and physical parts of the SF-36 questionnaire. The limits are as important as the headline:

• There was no control or sham group, so the authors write that a causal effect of HBOT on the outcomes cannot be established.
• A minority of people reported a significant and relevant deterioration after treatment.
• Some people stopped early because treatment was too exhausting or made symptoms worse.
• Several authors were employed by the commercial hyperbaric providers that delivered the treatment.

Why do the two studies point in different directions?

It is tempting to read the registry as proof that 40 sessions work where ten do not. The evidence does not support that conclusion yet, for several reasons.

1. Design. HOT-LoCO compared HBOT with a convincing sham. The registry compared people with their own starting point. Without a comparison group, natural recovery, attention from an eight-week programme and expectation can all look like treatment effects.
2. Session count. The registry used four times as many sessions. That could matter, but it has not been tested against a sham at that dose.
3. Who took part. Registry participants chose and paid for treatment and completed a demanding schedule. They may differ from people in a randomised trial.

Both studies used much higher pressures than our chamber, so neither result describes a course at 1.5 ATA.

What would settle the question is a trial that combines the strengths of both: a longer course like the registry, tested against a convincing sham like HOT-LoCO, with enough participants to detect a real difference. Until that exists, the honest position is uncertainty rather than a verdict either way.

What does this mean if you have long COVID?

Start with the care that has evidence behind it. The NHS long COVID page says there is currently no cure, but that treatments such as fatigue advice, physiotherapy, psychological support and rehabilitation can help ease symptoms. NICE guideline NG188 sets out how GPs identify ongoing symptoms and refer people to assessment services.

If you are still considering HBOT after that, three things are worth doing first:

• Ask the provider which study they rely on, at what pressure and with how many sessions.
• Read who HBOT is not suitable for, and agree with your own clinicians how you would judge whether anything changes.
• Treat a course as an energy cost that needs pacing like any other activity.

The UK advertising regulators' guidance on oxygen therapy says clinics need documentary evidence from clinical trials before claiming a therapy treats a condition. We do not claim that HBOT treats long COVID. For the general background, see what hyperbaric oxygen therapy is.