HBOT for diabetic peripheral neuropathy (≠ foot ulcers)

Hyperbaric oxygen therapy (HBOT) has a clear, insurance-covered role in one diabetic foot problem: severe, non-healing foot ulcers. Using it to treat diabetic nerve damage itself — the burning, numbness, and tingling of diabetic peripheral neuropathy (DPN) — is a different question, and a much weaker one. This atlas lays out what the actual trials show, why the strongest-sounding numbers come from small studies with real bias problems, and how HBOT stacks up against the drugs that already have approval for nerve pain.

The distinction that changes everything: ulcers vs. nerves

People mix these up constantly, and the mix-up matters for your money and your expectations.

A diabetic foot ulcer (DFU) is an open wound on the foot that won't heal. HBOT for DFUs is an approved indication — covered by Medicare and recognized by the Undersea and Hyperbaric Medical Society (UHMS) — but only for advanced wounds (Wagner grade 3 or higher) that have failed standard wound care for at least 30 days. The goal there is saving the limb, not fixing nerves.

Diabetic peripheral neuropathy (DPN) is the nerve disease underneath. High blood sugar over years damages the small blood vessels feeding your nerves and floods nerve tissue with oxidative stress. The result is the classic "stocking and glove" pattern: numbness, burning, pins-and-needles, and pain in the feet and hands. DPN can lead to ulcers (you stop feeling the wound), but treating the neuropathy and treating the ulcer are not the same goal.

HBOT for DPN-the-nerve-disease is investigational. No major guideline lists it as a recommended treatment for diabetic nerve pain. That is the single most important takeaway here, and the rest of this article explains why.

If a clinic markets HBOT to "heal your diabetic neuropathy," they are reaching past where the evidence stops. For the wound side of this story, see our HBOT for diabetic foot ulcer evidence atlas.

How HBOT might help nerves (the mechanism)

The mechanism is plausible. That's why researchers keep testing it. Plausible is not the same as proven, but it's worth understanding why the idea has legs.

In a hyperbaric chamber, you breathe near-100% oxygen at 2.0–2.5 times normal atmospheric pressure (ATA). Under that pressure, far more oxygen dissolves directly into your blood plasma — not just the oxygen riding on red blood cells. Plasma can carry oxygen into tissue beds where blood flow is poor and red cells struggle to reach.

DPN is driven partly by microvascular ischemia and hypoxia. The tiny vessels feeding peripheral nerves thicken and narrow. Nerve tissue starves for oxygen. That low-oxygen state cranks up oxidative stress and inflammation, which damages Schwann cells and the nerve fibers themselves. The proposed logic of HBOT:

• Flood the starved nerve environment with dissolved oxygen
• Stimulate growth of new small blood vessels (angiogenesis) over a course of sessions
• Dial down the oxidative and inflammatory cascade
• Give Schwann cells and axons a better environment to repair

That chain of events is well-documented in wound healing, which is exactly why HBOT works for some ulcers. The leap of faith is assuming the same oxygen boost meaningfully reverses or slows established nerve damage — and that's where the human data gets thin. The biology behind oxygen delivery is covered in depth in our guide on how HBOT increases tissue oxygenation.

Two cautions belong right next to the mechanism, because they're easy to skip past.

First, oxygen is a double-edged sword for nerves. DPN is partly an oxidative-stress disease — too many reactive oxygen species damaging nerve cells. Flooding tissue with oxygen could, in theory, add oxidative stress rather than relieve it. HBOT's defenders argue the body's antioxidant defenses ramp up to compensate over a course of sessions, and that the net effect on the microvasculature is protective. But the point stands: the mechanism is not a simple "more oxygen = better nerves" story. It's a balance, and we don't fully know which way it tips in established diabetic neuropathy.

Second, angiogenesis takes a full course to develop. A single session won't sprout new blood vessels. The new-vessel response that supposedly drives lasting benefit is thought to require dozens of sessions, which is part of why neuropathy protocols run 30–40 visits. That long, expensive commitment is being asked of patients on the strength of a mechanism that's biologically reasonable but clinically unproven for nerve symptoms.

What the evidence actually shows

Here's the honest picture. There is a published meta-analysis with a positive headline, and there are individual trials that found nothing. The gap between them tells you almost everything about how to read this field.

The positive meta-analysis (and its big asterisks)

The most-cited supportive paper is a 2024 systematic review and meta-analysis in the journal Medicine (Weng et al.). It pooled 14 randomized controlled trials: 675 patients on HBOT-plus-standard-care versus 648 on standard care alone. The findings sound strong on paper:

• HBOT had a significantly higher "effective treatment rate" (P < .001)
• Nerve conduction velocity improved across the median, ulnar, peroneal, and tibial nerves, for both motor and sensory fibers (most P-values < .01)
• Only 6 adverse events in the HBOT group, with no significant safety difference

Now the asterisks, which the authors themselves flag:

1. Publication bias was detected. The authors ran funnel plots plus Begg's and Egger's tests and found publication bias in several outcomes. That means negative or null studies were likely missing from the literature, inflating the apparent benefit.
2. Nearly all trials were small and conducted in China, drawing heavily on Chinese-language databases. Small single-center trials tend to overstate effects.
3. "Effective treatment rate" is a soft, composite endpoint. It is not a hard, patient-meaningful outcome like sustained pain reduction at one year or prevention of ulcers. Nerve conduction velocity is a physiologic measurement; it does not always track how a patient actually feels.
4. No long-term durability data. Short-term electrical improvements may fade.

So the meta-analysis is real and it's published, but it is low-to-moderate quality evidence with documented bias — not the kind of high-certainty proof that drives guideline recommendations.

It also helps to understand why "effective treatment rate" is so easy to inflate. In many of these trials, "effective" is a clinician's judgment that symptoms improved by some threshold, scored in an unblinded setting. When neither patient nor doctor is blinded — and in a chamber, true blinding is hard because you can feel the pressure — expectation alone can shift a soft outcome. Add the placebo effect (real and large in pain conditions) and a soft endpoint, and you can manufacture a "significant" benefit that wouldn't survive a rigorous sham-controlled design. That's not an accusation of fraud; it's the predictable behavior of weak study designs, and it's exactly why a positive pooled P-value here should not be read as proof.

The trials that found nothing

Pull away from pooled "effective rate" numbers and look at rigorous individual trials using patient-centered outcomes, and the optimism cools fast. A frequently cited controlled trial randomized 34 patients to either HBOT (100% oxygen at 2.4 ATA, 90-minute sessions, 30 sessions) or a control condition, measuring warm sensory threshold (a marker of small-fiber function) and validated pain questionnaires. The result: no significant differences in clinical outcomes out to one year. Two HBOT patients normalized their warm sensory threshold — interesting, but not a population-level win.

This is the recurring pattern in HBOT-for-neuropathy: surrogate electrical measures sometimes move, but the symptoms patients care about — pain, numbness, function — often don't budge in well-controlled studies.

The ulcer trials are a cautionary tale

Even on the approved ulcer side, the best modern trial humbled the field. A 2016 double-blind, randomized, placebo (sham-air) controlled trial in Diabetes Care (Fedorko et al.) tested HBOT in patients with chronic diabetic lower-limb ulcers and found HBOT did not reduce the need for amputation and did not significantly improve wound healing versus sham. A Cochrane review of HBOT for chronic wounds reached a similar caution: short-term ulcer-healing benefit at six weeks, no clear long-term benefit, and trials with design flaws that undercut confidence.

If the rigorous, blinded, sham-controlled evidence is shaky even for the approved wound indication, you should be especially skeptical of bold claims for the unapproved nerve indication.

Honest evidence grade

A staff reviewer's bottom line: the supportive signal exists but rests on weak, bias-prone studies, while the most rigorous trials are null. That's the definition of investigational.

What a "good" trial would need to settle this

It's worth naming what's missing, because it tells you how far the evidence has to travel before HBOT-for-DPN earns a real recommendation. A trial that could change minds would be:

• Large and multi-center, not a single clinic with 30 patients
• Sham-controlled — a control group that breathes pressurized air, not just "standard care," so the placebo and expectation effects are matched
• Patient-centered in its primary outcome — sustained pain reduction or improved function at 6–12 months, not a one-time nerve-conduction reading
• Pre-registered, so the outcomes can't be swapped after the data comes in
• Run outside a single country's literature, to break the small-trial, publication-bias pattern

No published trial in DPN checks all those boxes. Until one does, the honest label is "promising mechanism, unproven in people."

The HBOT protocol used in neuropathy studies

If you do explore HBOT — most sensibly for an ulcer, not the neuropathy itself — here's roughly what the clinical (not "mild") protocol looks like, based on the parameters used in the trials above. This is descriptive, not a recommendation.

The takeaway baked into this table: the studies that found any nerve-conduction signal used hard-shell chambers at 2.0–2.5 ATA. A home soft-shell unit at 1.3 ATA is a different intervention and can't be assumed to do the same thing. For the difference, see mild HBOT vs. medical HBOT.

How HBOT compares to treatments that actually have approval

This is where context matters most. Painful diabetic neuropathy has real, guideline-backed treatments. HBOT is competing against options with far stronger evidence and FDA approval.

The American Academy of Neurology's 2022 guideline on painful diabetic polyneuropathy recommends offering one of several drug classes as first-line: gabapentinoids (gabapentin, pregabalin), SNRIs (duloxetine, venlafaxine), tricyclic antidepressants (amitriptyline), and sodium-channel blockers. The AAN notes their effect sizes are broadly comparable, so the choice often comes down to side effects and other health conditions. Two of these — duloxetine and pregabalin — carry specific FDA approval for diabetic nerve pain.

The practical gap is stark. A first-line pill costs little, is covered by insurance, and has solid trials behind it. A full HBOT course is roughly 30–40 sessions of 60–90 minutes each, often running thousands of dollars out of pocket because insurers won't cover it for nerve symptoms. For a head-to-head on cost, see HBOT session packages and pricing. HBOT might one day prove to be a useful add-on for a subset of patients, but today it is not a substitute for the basics: tight glucose control, foot protection, and a guideline-backed pain medication.

There's one more comparison that often gets lost: the single most evidence-backed thing you can do for diabetic neuropathy isn't a treatment for the nerves at all — it's better blood sugar control, especially in type 1 diabetes, where tight glucose management clearly slows neuropathy progression. The benefit in type 2 is more modest, but glucose control plus careful foot care remains the foundation every guideline starts from. No one is selling a chamber package for that, which is part of why it gets less attention. It also costs almost nothing and has the strongest long-term evidence of anything in this article. Any honest discussion of HBOT for neuropathy has to put it next to that baseline, and against that baseline, an expensive, unproven chamber course is a hard sell.

It's also fair to note what the drugs don't do. Duloxetine, pregabalin, gabapentin, and amitriptyline manage pain — they don't repair nerves or reverse the underlying damage either. So in a sense, no current treatment fixes the nerve. The difference is that the pills have strong evidence for the thing they claim to do (reduce pain) at low cost and low risk, while HBOT claims more (improving the nerve itself) on much weaker evidence at far higher cost. That asymmetry is the whole decision.

Safety: what HBOT can and can't go wrong

HBOT is generally well-tolerated, and the DPN trials reported few serious problems. But it is a real medical procedure with real risks, and diabetes adds a wrinkle.

Common, usually minor:

• Middle-ear barotrauma — the most frequent side effect; pressure, pain, or fullness in the ears during compression. Equalizing technique helps.
• Temporary myopia (nearsightedness) — vision shifts during a long course (often after 20+ sessions) and typically reverses within weeks of finishing.
• Sinus discomfort, claustrophobia, fatigue.

Rare but serious:

• Oxygen-toxicity seizures — convulsions from breathing high-pressure oxygen, historically estimated near 1 in 10,000 treatments; they resolve when oxygen is removed.
• Pulmonary barotrauma / lung injury under pressure.

A diabetes-specific note: some patients see blood sugar drop during sessions, so glucose is often checked before and after. Discuss your medications and a snack plan with the staff.

Contraindications. The one absolute contraindication is an untreated pneumothorax (collapsed lung), where pressure changes can be life-threatening. Other cautions include certain chemotherapy drugs (such as doxorubicin and cisplatin), some lung disease, and a seizure history. A full pre-treatment screen is essential. Our HBOT contraindications guide covers who should not enter a chamber.

Who might reasonably consider it — and who shouldn't

Probably not worth it for most people with DPN. If your main problem is diabetic nerve pain or numbness, start with the proven, covered, lower-risk options: glucose control, foot care, and a first-line medication. There is no strong evidence HBOT beats these for nerve symptoms, and it costs far more.

A reasonable conversation if: you have an advanced, non-healing diabetic foot ulcer (Wagner grade 3+) that hasn't improved after a month of proper wound care. That's the ulcer indication, where HBOT has guideline support and possible insurance coverage — and where the goal is saving your foot, not curing neuropathy.

Approach with extra skepticism if: a clinic promises to "reverse" or "cure" diabetic neuropathy with HBOT, pushes a large cash package up front, uses a soft-shell "mild" chamber (1.3 ATA) that doesn't reach therapeutic pressures, or skips a medical screening. Those are marketing red flags. See HBOT clinic red flags before signing anything.

Always do this first: talk to the doctor managing your diabetes and an endocrinologist or neurologist. HBOT, if used at all, should be an add-on inside a real medical plan — never a replacement for blood-sugar control and standard nerve-pain care.

Questions to ask a clinic before you pay

If a clinic is pitching HBOT for your neuropathy, these questions cut through the marketing fast:

• "Is this for my foot ulcer or my nerve symptoms?" If it's for nerve symptoms, ask them to show you the guideline that recommends it. There isn't one.
• "What pressure does your chamber reach?" If the answer is 1.3 ATA (soft-shell), it's below the pressures used in any of the supportive trials.
• "Will my insurance cover this?" For nerve symptoms, expect no. For an advanced ulcer, maybe — and they should help you verify, not just sell a cash package.
• "What outcome should I expect, and by when?" Vague promises of "healing" or "reversal" are a red flag. Honest clinics talk in terms of trying an add-on, not curing.
• "Who screened me for contraindications?" A real facility does a medical screen before your first session.

The realistic expectation

If you and your doctor decide to try HBOT as an add-on, go in clear-eyed. The most likely outcome, based on the rigorous trials, is no dramatic change in your nerve pain or numbness. A minority of patients in studies saw a measurable nerve-conduction shift; few saw their day-to-day symptoms transformed. Set a defined trial period and a clear stop point with your doctor, keep up your glucose control and medications the whole time, and don't pour money into an open-ended package on the hope of a cure that the evidence doesn't support. Tracking your symptoms carefully — see our guide on how to track HBOT progress — helps you and your doctor decide honestly whether it's doing anything.

Frequently Asked Questions

Does HBOT cure diabetic neuropathy?

No. There is no good evidence that HBOT cures or reverses diabetic peripheral neuropathy. A pooled meta-analysis showed improved nerve conduction measurements, but that data is low quality with documented publication bias, and the most rigorous individual trials found no meaningful improvement in patients' pain or sensation at one year. Treat any "cure" claim as marketing.

Is HBOT for diabetic neuropathy covered by insurance?

Usually not. Medicare and most insurers cover HBOT for advanced diabetic foot ulcers (Wagner grade 3 or higher, failed 30 days of standard care) — not for nerve pain or numbness. Using HBOT to treat the neuropathy itself is considered off-label and investigational, so you'd typically pay out of pocket, often thousands of dollars for a full course.

How is HBOT for foot ulcers different from HBOT for neuropathy?

A foot ulcer is an open, non-healing wound; neuropathy is the underlying nerve damage. HBOT is an approved, sometimes-covered treatment for severe ulcers because added oxygen helps wounds heal and may help avoid amputation. Treating the nerve disease itself is a separate, unproven use. One has guideline backing; the other does not.

What's the better-proven alternative for diabetic nerve pain?

Guideline-backed first-line options include duloxetine, pregabalin, gabapentin, and amitriptyline — oral medications with strong trial evidence, and in some cases specific FDA approval for diabetic nerve pain. Combined with tight glucose control and foot care, these have far more evidence behind them than HBOT and cost a fraction as much.

Are "mild" home hyperbaric chambers useful for neuropathy?

There's no solid evidence they help diabetic neuropathy. Soft-shell "mild HBOT" chambers run around 1.3 ATA, well below the 2.0–2.5 ATA pressures used in the clinical neuropathy trials. The studies that showed any nerve-conduction signal used hard-shell medical chambers at therapeutic pressures, so home soft-shell units can't be assumed to deliver the same effect.

Medical disclaimer: This article is for general information only and is not medical advice. Diabetic neuropathy is a serious condition. Talk to your doctor before starting, stopping, or changing any treatment, including hyperbaric oxygen therapy.

Sources

• HBOT for diabetic peripheral neuropathy: systematic review and meta-analysis (Weng et al., Medicine 2024, PMID 39252242)
• HBOT does not reduce indications for amputation in diabetic non-healing ulcers: sham-controlled RCT (Fedorko et al., Diabetes Care 2016, PMID 26740639)
• AAN practice guideline: oral and topical treatment of painful diabetic polyneuropathy (2022, PMID 34965987)
• Cochrane review: hyperbaric oxygen therapy for treating chronic wounds
• UHMS approved hyperbaric oxygen indications
• Hyperbaric oxygen therapy contraindications (StatPearls, NCBI Bookshelf)
• Hyperbaric complications and side effects (StatPearls, NCBI Bookshelf)
• PubMed: HBOT and diabetic peripheral neuropathy nerve conduction studies
• PubMed: diabetic peripheral neuropathy pathogenesis, microvascular hypoxia, oxidative stress