Does hyperbaric oxygen therapy reverse biological aging in healthy adults?

Strip away the press releases and the entire aging-reversal story reduces to a single paper: Hachmo et al., Aging (Albany NY), 2020. It was a single-arm, uncontrolled, unblinded, non-randomized prospective trial. There was no control group and no sham chamber, every participant received the same 60 daily HBOT sessions over three months, and they were compared only to themselves. The authors themselves list 'lack of control group' and 'limited sample size' among the study's limitations.

The sample was small and shrank. Thirty-five healthy adults aged 64 and over enrolled; only 25 were analyzed for telomere length and 20 for senescent cells, so the senescence result rests on roughly 57% of the people who started. With no control arm, there is nothing to separate a real drug-like effect from regression to the mean, ordinary biological variation, or the way a small sample can drift when low-quality specimens are dropped.

Crucially, every outcome was a surrogate marker measured in isolated peripheral blood cells. Telomere length rose by up to roughly 37.6% in some immune-cell subsets and the share of senescent T-helper cells fell about 37.3% after HBOT. Those are real measurements. But no clinical, functional, cognitive, disease, or lifespan outcome was assessed. Nobody in this study was shown to be healthier, sharper, or longer-lived. 'Reverses biological aging' is an interpretation laid on top of two blood markers, not a result the study measured.

The paper's competing-interests statement is unusually direct. It states verbatim that co-authors 'AH, BY, ZY work for AVIV Scientific LTD' and that 'ES is a shareholder at AVIV Scientific LTD', 'ES' being Efrati's initials. AVIV Scientific is the company that sells the anti-aging HBOT program this very study is used to market. The funding came from a private Sagol-network grant. So the one favourable result underpinning a five-figure product was generated, in part, by employees and a shareholder of the firm selling it.

This is not a personal accusation; it is the disclosure the authors printed themselves, and Caveat's standard is simply to weight it. A positive surrogate-marker finding from parties with a direct commercial stake in that finding is low-trust by default, and it is exactly the configuration that calls for independent replication before anyone treats it as established. That replication does not exist: more than five years on, no group outside Efrati's circle has reproduced the telomere or senescence finding, and the same group's later work (a 2022 transcriptome paper) is again a single-arm, uncontrolled study of yet another surrogate.

Two independent critiques, neither with any commercial stake in HBOT, go to the heart of the measurement. The longevity-science writer Reason (Fight Aging!) argues that telomere length in immune cells is 'a truly terrible measure of aging', it varies day to day and tracks infection status, so an apparent rise can reflect a shift in which immune cells are circulating rather than genuine elongation. He also notes the effect is confined to immune-cell behaviour, not the body's tissues, and that if HBOT really cleared a third of senescent cells body-wide it would already be a celebrated treatment for conditions like arthritis. It is not.

The science-integrity journalist Leonid Schneider (For Better Science), drawing on an analysis by statistician Nick Brown, raised concrete data problems. Brown's verbatim charge is that 'Table 1 is absurd because they are comparing different subsets of the same sample', the comparisons are between non-independent subsets, not separate groups. He also flagged inconsistent handling of poor specimens: 4 low-quality samples were excluded from the telomere analysis but 10 from the senescence analysis, despite identical blood preparations, which raises a selective-reporting concern. Per the report, Efrati did not respond to emailed criticism.

Note the symmetry Caveat owes the reader: the proposed mechanism is biologically plausible, HBOT genuinely drives HIF and angiogenic signalling, which is why it works for wound healing. But the same hyperoxia invoked as rejuvenating also generates reactive oxygen species, and independent reviews tie that oxidative stress to telomere attrition, mitochondrial-DNA damage and cataract risk. The mechanism, honestly stated, cuts both ways.

The most favourable synthesis available is a 2022 systematic review in Frontiers in Aging (the citation circulating with the claim mislabels the authors; the actual paper is by Tessema, Sack, Serebrovska, König and Egorov). It pulled together 17 studies, 9 of them human. Its own conclusion is the quietest possible verdict on the longevity claim: 'There is no direct research to prove that short-term hyperoxia actually increases life expectancy in humans.'

The review rated the underlying studies as having high or unclear risk of bias, and it catalogues real harms that the marketing tends to omit, deterioration in visual acuity, middle-ear barotrauma, mitochondrial-DNA damage and increased nuclear cataract risk. And it is itself conflicted: the article processing charge was funded by InterHypox e.V., the work was supported by the hyperoxia-equipment firm CellAir Construction GmbH, and one author ('EE', i.e. Egorov) is a co-owner of CellAir. In other words, the friendliest paper in the literature is seller-adjacent and still cannot endorse the headline.

As a matter of regulatory reality, anti-aging is not among the indications for which HBOT is cleared in the United States; regulators position consumer anti-aging use as unproven and caution against clinics marketing it. We flag the specific FDA consumer page for human confirmation below rather than cite it as audited, because the page would not load to our verification tool, but the substance is not in dispute among the independent sources here.