Does Ozone Therapy Work? What a 2026 Umbrella Review of the RCTs Found

Ozone therapy has been offered in wellness and complementary clinics for decades on the strength of a plausible-sounding mechanism: a brief, controlled burst of oxidative stress that is thought to switch on the body's own antioxidant defences, improve circulation and calm inflammation. That mechanism has been described in laboratory work for years, but a mechanism is not the same as a measured benefit in an actual patient, and the underlying trial evidence has stayed fragmented across dozens of small studies testing different conditions, doses and delivery routes.

In June 2026, a team from the University of Florida, the Universita Cattolica del Sacro Cuore in Rome and Saint Camillus International University of Health Sciences published an umbrella review in the journal Medical Sciences, an approach that does not run new trials but instead systematically gathers every existing systematic review and meta-analysis of ozone therapy randomised controlled trials, appraises how trustworthy each one is, and summarises what the totality of that evidence actually shows. It is, in effect, a review of reviews, designed to answer a broader question than any single meta-analysis can: across everything that has been tested in proper trials, does ozone therapy consistently work?

The researchers searched MEDLINE, Web of Science, Embase and the Cochrane Library from the earliest available records through February 2025, with an updated search run again in May 2026 to catch anything newly published. That search returned 1,243 records. After removing 217 duplicates and screening the remaining 1,026 for relevance, the team read the full text of 99 reports, and ultimately included seven meta-analyses that met their criteria: each one had to be a systematic review pooling randomised controlled trials that compared ozone therapy against a genuine non-active control, meaning placebo, sham treatment, saline or standard care, rather than against another active treatment.

Those seven meta-analyses, taken together, covered four distinct clinical populations: people with chronic periodontitis, patients hospitalised with COVID-19, people with diabetic foot ulcers, and patients recovering from surgical removal of an impacted wisdom tooth. Each included review was then independently graded for methodological quality using AMSTAR-2, a standard tool for rating how trustworthy a systematic review's own methods are, and every individual outcome within those reviews was graded for certainty using GRADE, the same framework used across evidence-based medicine generally, which accounts for risk of bias, inconsistency between studies and how imprecise the result is.

For chronic gum disease, the umbrella review found two eligible meta-analyses that partially overlapped in which underlying trials they included, so the researchers kept only the more recent and comprehensive one for their formal certainty grading, to avoid double-counting the same patients. That more recent meta-analysis, by Liu and colleagues, reported statistically significant improvements in probing depth and gingival index, two standard measures of gum health, when ozone was used alongside standard scaling and root planing. It found no significant improvement, however, in bleeding on probing, plaque index or clinical attachment level, three other measures usually tracked in periodontal trials.

An earlier meta-analysis on the same question, by Moraschini and colleagues, did not find a significant adjunctive benefit at all. The umbrella review's authors rated the certainty of the periodontitis findings as very low for two of the five outcomes examined and low for the remaining three, using GRADE. In plain terms, even the more favourable of the two periodontitis reviews found a partial, inconsistent pattern of improvement in a condition where ozone is already among the more widely used applications, and the review's authors describe this as a possible signal in isolated surrogate measures rather than something that supports treating ozone as an established part of periodontal care.

For patients hospitalised with COVID-19, two randomised trials found that people who received ozone therapy were less likely to test positive on a PCR test at follow-up than people who did not, a large relative difference. The review's authors were explicit, however, that this is a surrogate laboratory measure rather than something a patient actually feels or benefits from directly, and on the outcomes that matter most to a hospitalised patient, length of hospital stay, whether they needed intensive care, and whether they survived, ozone therapy showed no significant benefit over standard care. The certainty of the hospital-stay finding was rated very low, and the intensive-care and mortality findings were rated low.

For diabetic foot ulcers, a Cochrane review covering two randomised controlled trials and 111 patients found that ozone therapy was not significantly better than control treatment at achieving actual wound healing, and no significant benefit was found for reducing the size of the ulcer either, with the certainty of both findings rated low to very low. Diabetic foot ulcers are a serious, slow-healing complication where a genuinely effective adjunct therapy would be valuable, which makes this a meaningful negative finding rather than an incidental one.

The fourth condition studied was recovery after surgical removal of an impacted wisdom tooth, covering nine separate outcomes. Ozone therapy did not reduce facial swelling or improve how far patients could open their mouth at 24 hours, 72 hours or seven days after surgery. It was associated with better patient-reported quality of life at each of those time points and with patients using fewer painkillers, though the review's authors noted there is no agreed threshold for what counts as a meaningful improvement on the quality-of-life scale used, so they treated this patient-reported signal as being of uncertain clinical relevance. All nine outcomes were rated very low certainty.

Across all seven meta-analyses, the review also formally rated how trustworthy each review's own methods were, using AMSTAR-2. Only two of the seven reached a high-confidence rating. Three were rated low confidence, mainly due to gaps such as not registering a review protocol in advance or not properly assessing publication bias, and two were rated critically low, meaning they had multiple serious methodological weaknesses. That matters because a positive-looking result pulled from a critically-low-confidence review carries far less weight than the same result from a well-conducted one, and more than half of the evidence base here falls into that weaker category.

It is worth being direct about a gap between what this review actually tested and what Irish ozone therapy clinics typically advertise. None of the four conditions in this review, chronic gum disease, hospitalised COVID-19, diabetic foot ulcers or wisdom tooth recovery, are the primary reasons most Irish wellness clinics offer ozone therapy, which tends to be marketed around general pain relief, immune support, detoxification and longevity. This review does not directly test any of those specific marketed uses, so it cannot be read as disproving them outright.

What it does establish is a broader pattern: across every condition where enough randomised trial evidence exists to actually pool it into a meta-analysis, ozone therapy has not shown a consistent, high-certainty benefit over placebo, sham or standard care. Ozone therapy is not officially endorsed or regulated as a specific medical treatment by the Medical Council of Ireland and is not available through the Health Service Executive, which places it in the same category as many complementary therapies offered in the private sector without having been formally adopted into mainstream medical practice. That regulatory gap does not by itself mean the treatment is unsafe or ineffective, but it does mean the burden of demonstrating benefit rests on trial evidence like this review, and on that measure the evidence for the uses actually studied has not yet cleared the bar.

An umbrella review can only summarise the trials that exist, and for many of the uses ozone therapy is actually sold for in Ireland, principally general pain, fatigue and immune support in otherwise healthy people, there simply are not enough well-designed randomised controlled trials to pool into a meta-analysis at all. Absence of evidence in that sense is not the same as evidence that ozone therapy does not help with those specific uses. It does mean, however, that confident marketing claims for those unstudied applications rest on an even thinner evidence base than the four indications this review actually examined, all four of which still came up short on the outcomes that matter most.

The review's authors also flagged that ozone administration varies considerably between studies and, by extension, between clinics: different routes such as rectal insufflation, topical application or major autohemotherapy, and different doses and schedules, were used across the pooled trials. That inconsistency in how ozone is actually delivered is itself part of why the trial evidence has struggled to produce a clear, reproducible signal, and it is a reasonable question to put to any clinic offering ozone therapy: which specific protocol are they using, and is there trial evidence behind that particular version of it.

A 2026 umbrella review that pooled everything usable in the randomised trial literature on ozone therapy, across four different conditions, found no consistent, high-certainty benefit over placebo, sham or standard care on the outcomes patients actually care about, and its own authors concluded that routine clinical use is not yet justified pending better-designed trials. A handful of secondary and surrogate measures moved in ozone's favour, but every one of them was rated low or very low certainty, and more than half of the underlying reviews had serious methodological weaknesses of their own.

None of this proves ozone therapy does not work for the reasons Irish clinics most commonly market it for, since those specific uses largely have not been tested in the kind of trials this review could pool. What it does mean is that anyone considering ozone therapy in Ireland today is choosing a treatment that sits outside mainstream medical endorsement and, on the conditions where proper trials do exist, has not yet demonstrated the kind of clear, replicated benefit that would justify treating it as more than an unproven complementary option.