Yoon Hang Kim, MD, MPH

Board-Certified in Preventive Medicine | Integrative & Functional Medicine Physician

Introduction: A Question Worth Asking Honestly

Extracorporeal Blood Oxygenation and Ozonation (EBOO) has gained significant traction in integrative and functional medicine. Blood is drawn from one vein, passed through a hollow-fiber dialysis-style cartridge where it meets an oxygen-ozone gas mixture, and returned through a second vein, processing roughly two liters over about an hour. Clinics often point to the dark material captured by the filter as proof of detoxification, and some market the therapy as capable of removing blood clots.

That framing raises a fair clinical question in the opposite direction: can EBOO itself cause blood clots? This article works through what the published evidence actually shows, where the mechanisms point, and what a responsible pre-treatment and post-event workup looks like.

How to read this article. Claims are labeled by evidence strength. Direct means ozone was applied to blood, the exposure EBOO actually delivers. Indirect means the finding comes from inhaled or ambient ozone air-pollution research, where ozone never contacts blood directly and effects are mediated through the lung. Mechanistic means an established pathway extrapolated to EBOO. Unverified means a claim whose only source is a conference presentation or personal communication. The Direct/Indirect distinction matters most: a great deal of the ozone-and-thrombosis literature studies air pollution at parts-per-billion inhaled, not micrograms-per-milliliter delivered into circulating blood, and the two should not be silently merged.

Section 1: What the Published Clinical Evidence Shows

There appear to be no published case reports of thrombotic events attributed specifically to EBOO. The clinical thrombotic and embolic cases in the ozone literature involve other modalities.

Acute myocardial infarction after ozonated autohemotherapy. A 46-year-old man without traditional atherosclerotic risk factors presented with acute inferior myocardial infarction hours after ozonated autohemotherapy. Angiography showed left main coronary vasospasm and thrombotic total occlusion of the right coronary artery (Üreyen, Baş, and Arslan, Cardiology, 2015).

Hemolytic anemia with deep venous thrombosis. A 36-year-old woman receiving blood ozone three times weekly for three months developed intravascular hemolytic anemia (hemoglobin 5 g/dL, dark urine, schistocytes) with deep venous thrombosis. She was not G6PD deficient. The ozone dose was unknown and this was a conference abstract, not a full paper (Cortázar-Benítez et al., Blood, 2011).

Gas embolism causing stroke and cord infarction. Three cases from ozone injection procedures: multifocal cerebral gas embolism with ischemic stroke, simultaneous spinal cord infarction and STEMI (both with patent foramen ovale as conduit), and posterior-circulation infarcts from vertebral-artery gas after cervical paravertebral injection (Khosravi and Mirzaasgari, 2024; He et al., 2019; Freund et al., 2019).

Section 2: Mechanisms — Direct Blood Ozonation Evidence

This section covers studies in which ozone actually contacted blood. These are the most relevant to EBOO and, notably, both show prothrombotic signals.

2.1. The heparin interaction: the most EBOO-relevant finding in the literature

This deserves prominence because it arose from precisely the phenomenon EBOO practitioners observe. Bocci's group investigated ozone's effect on platelets specifically because, during ozonated autohemotherapy using heparin, clots were occasionally retained in the filter during blood reinfusion. They demonstrated, by biochemical and morphological criteria, that heparin in the presence of ozone promotes dose-dependent platelet aggregation, and that after calcium chelation with citrate, platelet aggregation was markedly reduced. They attributed this to transient hydrogen peroxide formation in the presence of calcium (Bocci et al., Platelets, 1999).

The clinical implication is concrete and, to my knowledge, underdiscussed in the EBOO field: the standard anticoagulant used in EBOO may itself participate in ozone-induced platelet aggregation, while citrate anticoagulation appears to mitigate it. This is a single in vitro study from 1999 and has not, as far as I can determine, been replicated or tested in an EBOO circuit. But it is direct blood-ozonation evidence, it explains an observed clinical phenomenon, and it suggests a testable modification of practice.

2.2. Ozone increases circulating microparticles at therapeutic concentrations

Whole blood from 19 healthy men was treated with air, 15 µg/mL ozone, or 30 µg/mL ozone, concentrations in the range EBOO devices use. Ozone produced a statistically significant increase in microparticles derived from erythrocytes, platelets, leukocytes, and endothelial cells. Microparticles expose phosphatidylserine and provide a catalytic surface for thrombin generation. Critically, the same study found accompanying changes in prothrombin time, aPTT, D-dimer, and fibrinogen were mild and did not exceed normal values (Boczkowska-Radziwon et al., Biomolecules, 2022).

2.3. Ozone structurally modifies fibrinogen

Ozone directly oxidizes fibrinogen, with the D regions particularly affected, increasing hydroxyl, carbonyl, and carboxyl group content and modifying methionine, tryptophan, histidine, and phenylalanine residues. When oxidized fibrinogen is converted to fibrin, the resulting gels show a higher weight-to-length ratio of fibrils than native fibrin, meaning altered clot architecture (Rosenfeld et al., Biochemistry (Moscow), 2013; Free Radical Biology and Medicine, 2014; Yurina et al., Free Radical Research, 2019).

A caution on interpretation: these are purified-protein in vitro studies, and the source papers report altered fibrin architecture without establishing that the resulting clots resist fibrinolysis. Secondary summaries sometimes add that inference; the primary data do not support stating it as fact. What can fairly be said is that ozone changes the structure of the clot-forming protein, and that the functional consequence is not yet established.

Section 3: Mechanisms — Indirect Evidence from Inhaled Ozone

A large body of air-pollution research links ozone to prothrombotic changes. It is genuinely informative about ozone biology, but the exposure route differs fundamentally from EBOO: inhaled ozone never reaches the bloodstream as ozone, acting instead through lung epithelium and secondary oxidation products. Dose metrics are not interconvertible with EBOO's. These findings should inform suspicion, not substitute for EBOO data.

Platelet activation. In 89 healthy adults, a 10-ppb increase in 24-hour ozone was associated with a 36.3% increase in soluble P-selectin, a platelet activation marker, and a 61.1% increase over a two-week exposure window (Day et al., JAMA Internal Medicine, 2017).

Procoagulant and fibrinolytic shift. In a randomized crossover trial, 32 healthy young adults inhaled filtered air or 200-ppb ozone for two hours. Serum proteomics showed coagulation factor X increased 20.96% and factor VII-activating protease 28.35%, while the natural anticoagulants protein Z and protein Z-dependent protease inhibitor fell 13.62% and 33.54%, and plasminogen fell 10.47% (Niu et al., Journal of Hazardous Materials, 2022). A separate panel study of 152 healthy adults in Beijing linked low-level ambient ozone to neutrophil extracellular trap formation and pro-atherothrombotic markers (Xu et al., Atherosclerosis, 2024).

The counterweight: the strongest single trial was largely negative. MOSES was a double-blind, randomized crossover study of 87 healthy adults aged 55 to 70, exposed to 0, 70, and 120 ppb ozone. It found no effects on any of its three primary markers of systemic inflammation and prothrombotic state — C-reactive protein, monocyte-platelet conjugates, and microparticle-associated tissue factor activity. Among secondary endpoints, endothelin-1 rose slightly at 120 ppb (0.07 pg/mL, p = 0.008), though the 70-ppb comparison was negative (Balmes et al., PLoS ONE, 2019). Summaries of the ozone-thrombosis literature sometimes cite MOSES only for its endothelin-1 finding while omitting its null primary results. That is a selective reading, and readers should weigh the negative primary outcome of the field's largest controlled human exposure trial accordingly.

Section 4: Extracorporeal Circuit Effects

Independent of ozone, extracorporeal circuits are prothrombotic. Hemodialysis reliably generates thrombin, evidenced by rising thrombin-antithrombin complexes and prothrombin fragment 1+2, which is why anticoagulation is mandatory. The mechanism is more complex than textbook contact activation: a randomized crossover study across three modern dialyzer membranes found contact system markers unchanged throughout dialysis (François et al., Kidney International Reports, 2020), and a 2023 review titled “Beyond Contact-Activated Coagulation” attributes circuit clotting to interaction among coagulation factors, platelets, leukocytes, and complement (Engelen, Verhamme, and Vanassche, Seminars in Nephrology). Gas embolism is a further risk: any blood-gas barrier failure introduces bubbles, which in the roughly 25% of adults with a PFO can cross to the arterial circulation. Two peripheral cannulations for an hour also carry ordinary access-site thrombosis risk.

Section 5: The G6PD Question

G6PD deficiency is the most common human enzymopathy, affecting an estimated 400 million people, with the 2021 Global Burden of Disease analysis putting prevalence near 443 million. G6PD generates the NADPH that maintains reduced glutathione; mature erythrocytes cannot synthesize new protein, so deficient cells are especially vulnerable to oxidative stress. Because ozone is an oxidative stressor, a G6PD-deficient client is where oxidative hemolysis, and hemolysis-mediated thrombosis, becomes most plausible.

This reasoning is mechanistic. Notably, the one published ozone case combining hemolysis with thrombosis occurred in a client who was not G6PD deficient, and its authors described ozone-associated hemolytic anemia in G6PD-deficient individuals as extremely rare. Standard references recommend G6PD screening before specific oxidant drugs, most explicitly rasburicase; no guideline addresses ozone. Extending that logic to EBOO is my clinical judgment, not a standard of care: given an inexpensive test, widespread unawareness of carrier status, and deliberate exposure of a large blood volume to an oxidant, quantitative G6PD screening before EBOO seems a reasonable precaution, with quantitative assay preferred over qualitative testing in heterozygous females.

Section 6: Practical Recommendations

6.1. Anticoagulation strategy

Given Bocci's finding, the choice of anticoagulant deserves explicit consideration rather than default. Heparin is standard in EBOO circuits, yet heparin plus ozone promoted platelet aggregation in vitro while citrate substantially reduced it. Clinicians using EBOO may wish to discuss citrate-based or citrate-supplemented anticoagulation with their device manufacturer, recognizing that this evidence is preliminary, unreplicated, and that citrate anticoagulation carries its own considerations including hypocalcemia and metabolic effects.

6.2. Pre-EBOO screening

Quantitative G6PD assay; CBC with differential and platelets; basic metabolic panel for baseline renal function; and a targeted history covering prior clotting events, known thrombophilia, estrogen use, active malignancy, recent immobilization, and known PFO or cardiac shunt.

6.3. Workup after a thrombotic event

Hemolysis panel. LDH, haptoglobin, reticulocytes, indirect bilirubin, peripheral smear.

Quantitative G6PD. If not already done. Activity can read falsely normal during acute hemolysis because surviving reticulocytes are enzyme-rich; repeat in two to three months.

Thrombophilia panel. Factor V Leiden, prothrombin G20210A, antiphospholipid antibodies. Interpret protein C, protein S, and antithrombin cautiously during acute events or anticoagulation.

HIT screen. If heparin was used and platelets fell: 4T score and PF4 antibodies.

Echocardiography with bubble study. For PFO or intracardiac shunt, especially if gas embolism is suspected.

6.4. Contraindication framework

Circuit-only clotting: relative contraindication. Rule out technical causes first (heparin dosing, flow rate, dwell time, access). Clotting despite adequate anticoagulation warrants workup before repeating.

Client thrombotic event: absolute contraindication pending full workup; permanent in my view if an inherited or acquired thrombophilia is identified.

Section 7: The Evidence Problem on Both Sides

Proponents cite an informal safety record that deserves stating alongside its provenance. The figure of more than 200,000 ozone dialysis treatments in Malaysia without significant toxicity traces to a conference demonstration, not a published series. The observation of transient rust-colored urine in a minority of clients comes from a practitioner survey presented at a professional meeting and describes high-dose 10-pass therapy rather than EBOO. Both are unverified in the sense defined at the outset. No centralized registry or mandatory reporting exists; the International Scientific Committee of Ozone Therapy publishes a voluntary form, and voluntary reporting cannot establish incidence.

The dosing context of the reassuring EBOO data is also routinely omitted. Di Paolo's controlled EBOO trial reported increased lipid-peroxidation markers without appreciable erythrocyte hemolysis, but the human EBOO studies of that era delivered ozone at 0.5 to 1.0 µg/mL, whereas contemporary practice runs 30 µg/mL and sometimes up to 50 µg/mL. A no-hemolysis finding at 1 µg/mL does not establish safety at 30 µg/mL.

Where this matters most is consent. When a clinic markets EBOO as capable of removing blood clots while the circuit requires heparin precisely because it is prothrombotic, and while the only direct study of ozone and heparin found they promote platelet aggregation together, both sides of that ledger should be disclosed. The dark material in the filter is largely fibrin, activated platelets, lipoproteins, and plasma proteins that fouled the membrane, much of it clotting that occurred within the circuit.

Conclusion

Can EBOO cause blood clots? It is biologically plausible and not demonstrated. The strongest specific concern is not a general claim that ozone is prothrombotic, but the narrower and better-supported observation that ozone and heparin together promote platelet aggregation, arising from clots seen in autohemotherapy filters. Ozone also increases circulating microparticles at clinically used concentrations and structurally modifies fibrinogen, and extracorporeal circuits are independently prothrombotic. Against that: the largest controlled human ozone exposure trial found no effect on its primary prothrombotic endpoints, ozonated autohemotherapy did not affect platelet function in dialysis clients, the microparticle study found coagulation markers within normal limits, and no EBOO-specific thrombotic case appears in the literature.

The reasonable posture follows from that balance: screen for G6PD deficiency and thrombophilic risk, revisit the anticoagulant choice rather than defaulting to heparin, use a verified anticoagulation protocol, investigate any thrombotic event fully, and be transparent about what is known, what is not, and what is being checked. This is not an argument against ozone therapy. It is an argument for holding EBOO to the evidentiary and safety standards expected of any procedure that circulates a person's blood outside their body.

References

1. Üreyen ÇM, Baş CY, Arslan Ş. Myocardial Infarction after Ozone Therapy: Is Ozone Therapy Dr. Jekyll or Mr. Hyde? Cardiology. 2015;132(2):101-104. doi:10.1159/000431078. PMID: 26139204.

2. Cortázar-Benítez LF, Vargas Viveros P, Aiza Alvarez A, Hurtado Monroy R. Hemolytic Intravascular Anemia (HIA) Microangiopathic-Like and Deep Venous Thrombosis (DVT) Due to Ozone Exposure: Case Report. Blood. 2011;118(21):5278. doi:10.1182/blood.V118.21.5278.5278. (ASH abstract.)

3. Khosravi S, Mirzaasgari Z. Cerebral gas embolism and multifocal ischemic stroke during oxygen-ozone therapy: a case report. BMJ Neurol Open. 2024;6(2):e000885. doi:10.1136/bmjno-2024-000885.

4. He R, Huang Q, Yan X, Liu Y, Yang J, Chen X. A Case of Paradoxical Embolism Causing Anterior Spinal Cord Syndrome and Acute Myocardial Infarction Following the Intradiscal Oxygen-Ozone Therapy. Front Neurol. 2019;10:137. doi:10.3389/fneur.2019.00137.

5. Freund PR, Alshafai L, Margolin EA. Multifocal Stroke From Ozone Gas Emboli. J Neuroophthalmol. 2019;39(4):518-519. doi:10.1097/WNO.0000000000000754.

6. Bocci V, Valacchi G, Rossi R, Giustarini D, Paccagnini E, Pucci AM, Di Simplicio P. Studies on the biological effects of ozone: 9. Effects of ozone on human platelets. Platelets. 1999;10(2-3):110-116. doi:10.1080/09537109909169172. PMID: 16801079.

7. Boczkowska-Radziwon B, Olbromski PJ, Rogowska A, et al. Ozonation of Whole Blood Results in an Increased Release of Microparticles from Blood Cells. Biomolecules. 2022;12(2):164. doi:10.3390/biom12020164. PMID: 35204669.

8. Rosenfeld MA, Shchegolikhin AN, Bychkova AV, et al. Ozone-induced oxidative modification of fibrinogen molecules. Biochemistry (Mosc). 2013;78(10):1171-1179. doi:10.1134/S000629791310012X. PMID: 24237152.

9. Rosenfeld MA, Shchegolikhin AN, Bychkova AV, Leonova VB, Biryukova MI, Kostanova EA. Ozone-induced oxidative modification of fibrinogen: role of the D regions. Free Radic Biol Med. 2014;77:106-120. doi:10.1016/j.freeradbiomed.2014.08.018. PMID: 25224034.

10. Yurina L, Vasilyeva A, Indeykina M, et al. Ozone-induced damage of fibrinogen molecules: identification of oxidation sites by high-resolution mass spectrometry. Free Radic Res. 2019;53(4):430-455. doi:10.1080/10715762.2019.1600686. PMID: 30935261.

11. Day DB, Xiang J, Mo J, et al. Association of Ozone Exposure With Cardiorespiratory Pathophysiologic Mechanisms in Healthy Adults. JAMA Intern Med. 2017;177(9):1344-1353. doi:10.1001/jamainternmed.2017.2842.

12. Niu Y, Li H, Wang W, et al. Ozone exposure and prothrombosis: Mechanistic insights from a randomized controlled exposure trial. J Hazard Mater. 2022;429:128322. doi:10.1016/j.jhazmat.2022.128322. PMID: 35086041.

13. Xu H, He X, Zhang B, et al. Low-level ambient ozone exposure associated with neutrophil extracellular traps and pro-atherothrombotic biomarkers in healthy adults. Atherosclerosis. 2024;395:117509. doi:10.1016/j.atherosclerosis.2024.117509. PMID: 38523002.

14. Balmes JR, Arjomandi M, Bromberg PA, et al. Ozone effects on blood biomarkers of systemic inflammation, oxidative stress, endothelial function, and thrombosis: The Multicenter Ozone Study in oldEr Subjects (MOSES). PLoS One. 2019;14(9):e0222601. doi:10.1371/journal.pone.0222601. PMID: 31553765.

15. Tylicki L, Lizakowski S, Biedunkiewicz B, et al. Platelet function unaffected by ozonated autohaemotherapy in chronically haemodialysed patients. Blood Coagul Fibrinolysis. 2004;15(7):619-622. PMID: 15389131.

16. Martínez-Sánchez G, Delgado-Roche L, Díaz-Batista A, Pérez-Davison G, Re L. Effects of ozone therapy on haemostatic and oxidative stress index in coronary artery disease. Eur J Pharmacol. 2012;691(1-3):156-162. doi:10.1016/j.ejphar.2012.07.010. PMID: 22796450.

17. François K, Orlando C, Jochmans K, et al. Hemodialysis Does Not Induce Detectable Activation of the Contact System of Coagulation. Kidney Int Rep. 2020;5(6):831-838. doi:10.1016/j.ekir.2020.03.010. PMID: 32518865.

18. Engelen MM, Verhamme P, Vanassche T. Clotting of the Extracorporeal Circuit in Hemodialysis: Beyond Contact-Activated Coagulation. Semin Nephrol. 2023;43(6):151473. doi:10.1016/j.semnephrol.2023.151473. PMID: 38233291.

19. Di Paolo N, Bocci V, Salvo DP, et al. Extracorporeal blood oxygenation and ozonation (EBOO): a controlled trial in patients with peripheral artery disease. Int J Artif Organs. 2005;28(10):1039-1050. PMID: 16288443.

20. Rowen RJ, Grabovac S, Su TB. Ozone dialysis delivers three or more times the ozone than other forms of ozone blood treatment. Med Gas Res. 2022;13(2):67-71. doi:10.4103/2045-9912.356474. PMID: 36204785.

21. Nkhoma ET, Poole C, Vannappagari V, Hall SA, Beutler E. The global prevalence of glucose-6-phosphate dehydrogenase deficiency: a systematic review and meta-analysis. Blood Cells Mol Dis. 2009;42(3):267-278.

22. Merck Manual Professional Edition. Glucose-6-Phosphate Dehydrogenase (G6PD) Deficiency. Accessed 2026.

About Dr. Kim

Dr. Yoon Hang “John” Kim is a board-certified physician with over 20 years of clinical experience. He completed his fellowship training under Dr. Andrew Weil at the University of Arizona and holds board certifications in Preventive Medicine, Medical Acupuncture, and Integrative/Holistic Medicine. He specializes in low dose naltrexone (LDN), autoimmune conditions, chronic pain, integrative oncology, fibromyalgia, chronic fatigue syndrome, mast cell activation syndrome (MCAS), and mold toxicity. He is the author of three books and over 20 peer-reviewed articles.

Professional: www.yoonhangkim.com

Clinical: www.directintegrativecare.com