If you have diabetes and you're starting hyperbaric oxygen therapy (HBOT), a hyperbaric nurse will likely check your blood sugar before you ever step into the chamber. That isn't a formality. Most diabetic patients see their blood glucose fall during a session, and a small number drop low enough to need a snack or a glucose tablet on the spot. This guide explains why glucose falls under pressure, how big the drop usually is, who is most at risk of going too low, and what a careful clinic does to keep you safe.
The short version, backed by the published data, is reassuring but not zero-risk: the average drop is modest, dangerous lows are rare, and the people who run into trouble are almost always insulin users who came in with a glucose level that was already on the low side. The longer version is worth your time, because the difference between a routine session and a scary one usually comes down to a few habits you and your clinic can control.
What Actually Happens to Blood Sugar Inside the Chamber
Walk into a hyperbaric chamber and the air pressure around you climbs above normal sea-level pressure, usually to somewhere between 1.3 and 2.5 atmospheres absolute (ATA), depending on the protocol. You breathe oxygen at a far higher concentration than the 21% in room air. The whole point is to dissolve much more oxygen into your blood plasma than normal, flooding tissues with the fuel they need to heal.
That oxygen flood is exactly why glucose tends to fall. Your cells use oxygen and glucose together to make energy. When tissues are suddenly bathed in extra oxygen, several glucose-lowering effects appear to stack up at once: muscle and other tissues pull more glucose out of the blood, insulin seems to work more effectively for a window of time, and the metabolic machinery that burns sugar runs harder. The result is a measurable dip in circulating blood glucose during and shortly after a treatment.
This is not a fringe observation. It's consistent enough that hyperbaric units treat it as a standard feature of caring for diabetic patients, not a surprise. A study validating glucose monitoring inside the chamber opens by stating plainly that "patients undergoing hyperbaric oxygen treatments have been shown to experience a reduction in blood glucose levels during a treatment," which "necessitates frequent assessment" (PMID 34547774). In other words, the drop is expected, and the monitoring exists because of it.
How Big Is the Drop? The Numbers From Real Patients
The most useful data on the size of the drop comes from a retrospective review of diabetic patients in a U.S. hyperbaric program: 77 patients across 1,825 treatments, with blood glucose recorded before and after each session (PMID 31509900). It's the single best snapshot of what to expect, so it's worth looking at the numbers closely.
| Measure | Finding in the 1,825-treatment review |
|---|---|
| Treatments where glucose fell | 75.4% |
| Median size of the drop | 25 mg/dL |
| Range of change | Down 374 mg/dL to up 240 mg/dL |
| Type 2 patients: treatments with a drop | 77.5% |
| Type 1 patients: treatments with a drop | 51.5% |
| Treatments ending below 90 mg/dL | 1.1% |
| Treatments ending below 70 mg/dL | 0.2% |
| Patients needing emergency care | 0 |
A few things stand out. First, most sessions did produce a fall, but the typical fall was small: 25 mg/dL is roughly the swing you might see from a brisk walk or a slightly late meal. Second, the spread was enormous. The range stretched from a 374 mg/dL crash to a 240 mg/dL rise, which tells you that individual responses vary wildly and that "average" hides a lot. Third, and most reassuring, true lows were uncommon. Only about 1 in 90 treatments ended below 90 mg/dL, and only about 1 in 500 ended below the 70 mg/dL line that defines hypoglycemia. No patient needed emergency care.
The authors' conclusion is deliberately measured: HBOT "does not cause a clinically significant decrease" in diabetic blood glucose for most patients. That's the honest headline. The drop is real and worth respecting, but for the average person on a standard wound-care protocol it is not a medical emergency waiting to happen.
Why the Average Hides the Real Risk
Here's the catch with a 25 mg/dL median. If you start a session at 180 mg/dL, a 25-point drop leaves you at a comfortable 155. If you start at 95, that same drop lands you at 70, the edge of hypoglycemia. And if you're one of the people in that long tail who drops 100, 200, or 374 points, your starting number is the only thing standing between you and the floor. This is why clinics obsess over your pre-treatment glucose rather than the size of the drop. You can't control how much you'll fall, but you can control how high you start.
Who Is Most at Risk of Going Too Low
Not everyone faces the same risk, and the data point to specific groups. The review above found that the patients who ended up below 90 mg/dL were overwhelmingly insulin users: about 70% of the low readings came from people managed on insulin alone, a statistically significant pattern. Type 1 patients, who are insulin-dependent by definition, were a notable subgroup, though interestingly their glucose dropped less often than Type 2 patients' did (51.5% versus 77.5% of treatments) — likely because their sugars are more variable and harder to control in general.
| Risk factor | Why it raises hypoglycemia risk |
|---|---|
| Insulin use (especially insulin alone) | Insulin already drives glucose down; HBOT's effect adds on top of it |
| Sulfonylureas (glipizide, glyburide, glimepiride) | Oral drugs that force insulin release and can cause lows on their own |
| Low pre-treatment glucose | A small drop from an already-low start crosses into hypoglycemia |
| Skipped or delayed meal before the session | No incoming glucose to buffer the fall |
| Tight glycemic control / aggressive A1c targets | Less cushion above the low threshold |
| Long or higher-pressure sessions | More time and more oxygen for the glucose-lowering effect to act |
The flip side is genuinely encouraging. In the same review, well-controlled diabetics — those whose glucose stayed within a 50 mg/dL band across all their pre-treatment checks — had zero readings below 70 and zero below 90. Stability protected them. The patients who got into trouble tended to be the ones whose sugars were already swinging unpredictably. If your diabetes is well managed and you're not on insulin or a sulfonylurea, your hypoglycemia risk in the chamber is low.
If you take other medications, it's worth understanding how HBOT interacts with the rest of your regimen, not just your diabetes drugs. Our guide to HBOT and medication interactions covers the drugs that matter most.
The Hidden Trap: Your Glucose Meter May Be Lying
There's a technical wrinkle that every diabetic HBOT patient should understand, because it can create a fake emergency. Many fingerstick glucose meters use an enzyme called glucose oxidase (GO). Those meters need oxygen to do their chemistry, and in the oxygen-saturated environment of a hyperbaric chamber, that extra oxygen throws off the reading — they report values that are erroneously low.
A study testing this directly found that during HBOT, a GO-based meter read about 10 mg/dL lower than a reference method, a statistically significant artifact, while a meter using a different enzyme (glucose dehydrogenase, or GD) was not fooled (PMID 31916854). The practical danger is obvious: a GO meter can flash a scary low number that isn't real, prompting unnecessary treatment, or it can mask how high your true glucose is. Continuous glucose monitors (CGMs) showed their own small shifts in the same study, reading slightly high under pressure.
What this means for you is simple. Ask whether the clinic's meter is GO-based or GD-based, and whether they account for the offset. Reputable units test glucose before you enter the chamber, in normal air, precisely to sidestep this problem. If a reading taken inside the chamber looks alarming, it should be confirmed at surface pressure before anyone acts on it. The good news from the validation research is that capillary fingersticks and CGMs, when used correctly, agree closely with lab serum values — the readings are trustworthy as long as you know the device's quirks (PMID 34547774).
What Good Clinics Do to Keep You Safe
Hyperbaric medicine has handled diabetic patients for decades, and the safety playbook is well established even if the exact numbers vary by unit. A survey of all 13 accredited hyperbaric units in Australia and New Zealand mapped what careful practice looks like, and it's a good template for what you should expect anywhere (PMID 37718297).
In that survey, 12 of 13 units routinely treated diabetic patients, and diabetes was common — about 26% of their patients, of whom 93% had Type 2. Most units (83%) had written protocols for managing blood glucose, and most (83%) tested glucose on every diabetic patient. Three-quarters defined hypoglycemia as a glucose below 4 mmol/L (about 72 mg/dL), close to the 70 mg/dL U.S. threshold. Preferred pre-treatment minimums ranged from about 4 to 8 mmol/L (roughly 72 to 144 mg/dL), meaning a patient below that floor would get a snack or glucose before being allowed to dive.
| Safety practice | What it looks like in a careful unit |
|---|---|
| Pre-treatment glucose check | Measured in normal air before every session for diabetic patients |
| Minimum glucose to enter | Often a floor around 100–120 mg/dL; below it, you eat first |
| Supplementation | Snack, juice, or glucose tablets to raise a low pre-treatment number |
| In-course monitoring | Continued testing through the treatment series, not just day one |
| Written protocol | Documented rules nursing and physicians follow |
| Post-treatment check | Confirming glucose hasn't dropped too far before you leave |
The survey also found real variability — different thresholds, different testing frequencies, different satisfaction with their own policies — which is a reminder that not every clinic is equally rigorous. Some units relax testing for patients whose sugars prove stable over many sessions, which is reasonable, but the baseline of checking before you dive should be non-negotiable. If a clinic offers you HBOT and never asks about your diabetes or checks your glucose, treat that as a red flag. Our HBOT side effects guide covers the broader safety picture beyond blood sugar.
Is the Glucose Drop Actually a Benefit?
This is where the topic gets genuinely interesting, and where it's easy to be misled by clinic marketing. If HBOT lowers blood sugar, could it treat diabetes? The honest answer: the evidence is suggestive but thin, and you should not buy HBOT as a diabetes treatment.
A systematic review pulled together every study on HBOT's effect on glycaemia in people with diabetes and found 10 eligible publications (PMID 34684171). Six reported a statistically significant reduction in blood glucose from HBOT. Two found a significant boost in peripheral insulin sensitivity. And two found a meaningful drop in HbA1c, the three-month average of blood sugar that matters most for long-term diabetes control. On paper, that's a consistent signal pointing in a helpful direction.
But the reviewers were blunt about the limits. The studies were small, potentially underpowered, and mostly done in patients already receiving HBOT for diabetic foot wounds — not designed to test diabetes treatment head-on. Their conclusion was that there is "emerging evidence" of a glycaemia benefit, but that large, properly powered trials are still needed before anyone can claim HBOT controls diabetes. Mechanistically, animal work supports the idea: in diabetic rats, mild hyperbaric oxygen prevented the rise in blood glucose and HbA1c and protected the muscle's oxygen-burning capacity, linked to genes like PGC-1α that govern metabolic fitness (PMID 29633563). Plausible mechanism, real animal data, and a few suggestive human studies — but not proof.
So the fair framing is this: the same effect that creates a small hypoglycemia risk during treatment may, over a course of sessions, nudge glucose control in a good direction. That's a pleasant side benefit if you're getting HBOT for an approved reason like a diabetic foot ulcer. It is not, on current evidence, a reason to pursue HBOT in place of proven diabetes care.
Practical Steps Before and During Treatment
Most hypoglycemia in the chamber is preventable with simple habits. Eat a normal meal or snack within an hour or two before your session — don't show up fasting. Bring fast-acting glucose (tablets or juice) and tell the staff you have it. Know your own pattern: after the first few sessions, ask the nurse how much your glucose tends to drop so you can plan around it. If you use insulin, talk to your prescriber about whether to adjust timing or dose on treatment days, since the chamber's effect stacks on top of your medication.
What to watch for inside the chamber: shakiness, sweating, sudden hunger, confusion, a racing heart, or feeling lightheaded are classic hypoglycemia warnings. Tell the chamber operator immediately if any of them appear — do not wait to "tough it out." Operators are trained to bring you up safely and treat a low. The 1,825-treatment review is reassuring here: across all those sessions, not a single patient needed emergency care, precisely because monitoring caught problems early.
For more on getting ready, see our first HBOT session preparation guide and the pre-session nutrition guidelines, which cover what and when to eat before a dive.
Who This Matters Most For
If you're a Type 2 diabetic managed with diet, metformin, or other drugs that don't directly force insulin out, your hypoglycemia risk in the chamber is low, even though your glucose is the most likely to drop. If you use insulin or a sulfonylurea, you're in the group that needs the closest watching — not because the therapy is unsafe for you, but because your floor is lower to begin with. If your diabetes runs unstable, with big daily swings, the chamber adds one more variable, so tighter monitoring and a higher pre-treatment target make sense. And if your diabetes is rock-steady, the data suggest you can often be managed much like a non-diabetic patient, with routine checks as a backstop.
Whatever group you fall into, the core message holds: HBOT lowers blood sugar in a way that is usually small, occasionally significant, almost always predictable, and entirely manageable with monitoring. The patients who get hurt are the ones whose clinics skip the basics. Choose a unit that takes your glucose seriously, and the risk shrinks to near nothing. If you're still deciding whether HBOT is right for you, our guide to who's a good candidate walks through the full picture.
Frequently Asked Questions
How much will my blood sugar drop during an HBOT session?
For most diabetic patients, the typical drop is modest — a median of about 25 mg/dL in the largest published review, with roughly three out of four sessions showing some decrease. But individual responses vary enormously, from large crashes to actual rises, which is why clinics check your glucose before every session rather than assuming a fixed number.
Is hypoglycemia in the chamber actually dangerous?
It can be, but serious lows are rare. In a review of 1,825 treatments, only about 0.2% ended below the 70 mg/dL hypoglycemia threshold, and no patient required emergency care. The risk concentrates in insulin users and people who start a session with already-low glucose. Standard monitoring catches problems early, which is why HBOT is considered safe for diabetics in supervised settings.
Can HBOT treat my diabetes or lower my A1c?
The evidence is promising but not proven. A systematic review found that most studies showed a glucose-lowering effect and a couple showed reduced HbA1c, but the studies were small and underpowered, and none were designed to test HBOT as a diabetes treatment. Do not pursue HBOT in place of proven diabetes care; treat any glucose benefit as a possible bonus, not a reason to start.
Why does the clinic check my glucose before I go in instead of during?
Two reasons. First, the most important number is where you start, since a small in-chamber drop only becomes dangerous if you began low. Second, some glucose meters that use the glucose-oxidase enzyme read falsely low in the oxygen-rich chamber, so a pre-treatment check in normal air gives a more reliable baseline. If a reading taken inside the chamber looks alarming, it should be confirmed at normal pressure.
Should I eat before my hyperbaric session?
Yes. Eating a normal meal or snack within an hour or two beforehand gives your body incoming glucose to buffer the expected drop. Showing up fasting is one of the most common avoidable causes of an in-chamber low. Bring fast-acting glucose with you and tell the staff, especially if you use insulin or a sulfonylurea.
This article is for educational purposes only and is not medical advice. If you have diabetes, talk to your physician and your hyperbaric medicine team before starting HBOT, and never adjust insulin or other medications without medical guidance.
Sources: Systematic review: impact of HBOT on glycaemia in diabetes (Medicina, 2021) · Blood glucose levels in diabetic patients undergoing HBOT (Undersea Hyperb Med, 2019) · Blood glucose management practices in hyperbaric units, Australia & New Zealand survey (Diving Hyperb Med, 2023) · Validation of venous, capillary, and interstitial glucose monitoring during HBOT (Diving Hyperb Med, 2021) · Reliability of glucose meters and CGM during hyperbaric oxygen exposure (Diabetes Technol Ther, 2020) · Mild hyperbaric oxygen and glucose metabolism in type 2 diabetic rats (J Diabetes, 2018) · PubMed search: hyperbaric oxygen, blood glucose, diabetes · Cleveland Clinic: Hyperbaric oxygen therapy overview*