Can Carbon Monoxide Cause Cancer? Potential Dangers

Carbon monoxide is not classified as a carcinogen by major agencies, but a growing body of research suggests it plays a more complicated role in cancer than that clean label implies. A large nationwide study in South Korea found that people who survived carbon monoxide poisoning had a statistically higher risk of developing solid organ cancers afterward, and the lung cancer risk was especially pronounced. At the cellular level, CO interacts with processes that can both protect healthy tissue and, under certain conditions, help tumors thrive. The relationship is not a simple cause-and-effect story, and the science is still catching up to the question.

What Population Studies Actually Show

The strongest epidemiological evidence linking CO exposure to cancer comes from two large cohort studies conducted with South Korean national health data. The first tracked patients who had been poisoned by carbon monoxide and compared their long-term cancer rates to the general population. Patients with CO poisoning had a small but statistically significant increase in overall internal malignancy risk, and when the researchers broke the data down further, the increase was concentrated in solid organ cancers rather than blood cancers like leukemia or lymphoma, which were actually lower in the CO-poisoned group.¹

The second study zeroed in on lung disease specifically. Among CO poisoning survivors, the risk of developing lung cancer roughly doubled compared to controls.² That is a substantial increase, though it comes with an important caveat: people who get CO poisoning often share other risk factors. Many poisoning cases involve fires, faulty gas appliances, or occupational exposures that come bundled with other toxic chemicals. Disentangling carbon monoxide’s independent contribution from the broader toxic stew is difficult, and both studies acknowledge this limitation. Still, the consistency of the signal across a large population is hard to dismiss.

How Carbon Monoxide Disrupts Cells

Carbon monoxide is best known for binding to hemoglobin and starving the body of oxygen. That alone causes tissue damage, but the molecule does more than just block oxygen delivery. At the cellular level, CO specifically targets an enzyme in the mitochondrial energy-production chain, reducing its activity in a dose-dependent way. Even at concentrations as low as 50 parts per million, researchers observed a measurable drop in this enzyme’s function, and at 500 ppm the reduction reached about 55%.³ When mitochondria cannot produce energy properly, cells generate more reactive oxygen species, which are unstable molecules that damage DNA. Accumulated DNA damage is one of the fundamental ways cancer gets started.

CO also influences how cells repair that damage. Research has shown that CO exposure activates a DNA repair pathway called homologous recombination through a protein involved in detecting broken DNA strands.⁴ At first glance, stimulating DNA repair sounds protective. And in healthy tissue, it may be. But in cells that are already on the path toward becoming cancerous, enhanced DNA repair can actually help them survive treatments like chemotherapy or radiation that are designed to kill them by breaking their DNA. The same protective mechanism that shields a normal cell can shield a tumor cell.

When the Body’s Own Carbon Monoxide Feeds Tumors

Your body produces carbon monoxide naturally. An enzyme called heme oxygenase-1 (HO-1) breaks down heme, the iron-containing molecule in hemoglobin, and releases CO as a byproduct. Under normal circumstances, this is part of a healthy stress response. The trouble is that many tumors hijack this system, cranking up HO-1 production to levels far beyond what healthy tissue needs.

In colorectal cancer, HO-1 has been identified as what researchers call a “conditional tumor promoter.” While its stress-protective role is normally beneficial, in the context of cancer it promotes cell survival, immune evasion, and the growth of new blood vessels to feed the tumor. Blocking HO-1 in experimental settings has reversed some of these tumor-promoting effects.⁵ The pattern is not unique to the colon. In glioblastoma, an aggressive brain cancer, experiments with CO-releasing molecules showed that CO had significant effects on cell migration and wound healing, with the specifics varying by cell type.⁶ In uveal melanoma, a rare eye cancer, CO-releasing molecules significantly sped up cell division, reducing the time it took for cancer cells to double.⁷

CO also promotes the growth of new blood vessels, a process tumors depend on to sustain themselves as they expand. In lab experiments, CO increased levels of a protein called VEGF, one of the most important signals for new blood vessel formation. It did this by stabilizing a master-regulator protein that normally gets broken down when oxygen is plentiful, essentially tricking cells into behaving as if they were in a low-oxygen environment.⁸ Tumors already tend to be oxygen-starved environments, and CO amplifies that signal, potentially accelerating the tumor’s ability to build its own blood supply.

CO Exposure in Real-World Settings

Most people are not exposed to a single massive dose of carbon monoxide. The more common scenario is chronic low-level exposure from sources you might not think about. Faulty or poorly maintained gas appliances are a major culprit in the home, and the problem may be significantly underreported because the symptoms of mild CO exposure, including headaches, fatigue, and nausea, overlap with dozens of other conditions.⁹ A UK review noted that the long-term consequences of chronic low-level CO exposure from domestic appliances remain poorly understood, in part because the exposure often goes unrecognized for months or years.¹⁰

Occupational exposure adds another dimension. A decade-long cohort study of firefighters found significantly elevated cancer mortality after a five-year latency period. Lung cancer rates were roughly three times the expected level among firefighters aged 60 to 74, and non-lung cancers were elevated among younger firefighters as well.¹¹ Firefighters inhale a cocktail of toxins in combustion smoke, and isolating CO’s contribution is impossible in that context. But the study is a reminder that CO rarely acts alone in the real world. Tobacco smoke is another major source: research comparing cigarette and hookah emissions found that both produce CO alongside carcinogenic compounds, ultrafine particles, and aldehydes.¹² Whether the CO itself contributes to the cancer risk from smoking, or merely keeps bad company with the chemicals that do, is an open question.

Biomass smoke exposure, common in regions where wood, dung, or crop residue is burned for cooking and heating, carries its own CO burden. Research on women chronically exposed to biomass smoke has found altered expression of small RNA molecules and changes in gene activity linked to inflammation and breast cancer risk, though these effects involve the full suite of biomass combustion products, not CO alone.¹³

The Paradox of CO as a Cancer Treatment

Here is where the story gets genuinely strange. While the evidence above suggests CO can help tumors grow, a parallel line of research is exploring whether controlled doses of CO might actually fight cancer. The idea is not as contradictory as it sounds. Many biological molecules behave differently at different concentrations or when delivered in a targeted way. Water sustains life and also drowns people; the dose and context matter enormously.

Researchers have been developing carbon monoxide-releasing molecules, or CORMs, designed to deliver CO to tumor tissue specifically. A growing number of experimental studies suggest these molecules can target several cancer hallmarks at once, including uncontrolled cell division, resistance to cell death, blood vessel formation, and the ability of tumors to invade surrounding tissue.¹⁴ A review of CO-based cancer research noted increasing interest in next-generation CORMs designed for potential clinical use.¹⁵

One particularly inventive approach uses nanotechnology to release CO at a tumor site only when triggered by X-ray radiation. The idea is to make tumors more sensitive to radiation therapy. A high local dose of CO acts as a metabolic disruptor, interfering with the tumor’s ability to generate energy through glycolysis and making cancer cells more vulnerable to radiation damage.¹⁶ Another experimental system co-delivers CO and hydrogen gas: the CO sensitizes the tumor to radiation while the hydrogen protects the surrounding healthy intestinal tissue from radiation damage.¹⁷

These are still preclinical results, mostly from lab dishes and animal models, not human trials. But they illustrate a genuinely important point about CO biology: the molecule is not simply toxic or simply therapeutic. Its effects depend heavily on concentration, timing, and which cells it reaches. Low-level chronic exposure in the body may create conditions favorable for tumor growth, while a precisely aimed burst of CO at a tumor site may disrupt the cancer’s energy metabolism enough to kill it or soften it up for other treatments.

Why Measuring Long-Term CO Exposure Is So Difficult

One reason the cancer question remains unsettled is that measuring past CO exposure is surprisingly hard. The standard biomarker, carboxyhemoglobin (COHb), reflects how much CO is currently bound to your red blood cells. It is useful for diagnosing acute poisoning but tells you almost nothing about what happened last month or last year. A systematic review of the literature found that while COHb is the most commonly used biomarker for CO exposure, most studies were classified as having a risk of bias, and confounding factors like smoking were often not properly accounted for.¹⁸ Without a reliable way to quantify cumulative lifetime exposure, epidemiological studies have to rely on dramatic events like diagnosed CO poisoning as a proxy, which misses the much larger population of people living with chronic low-grade exposure from leaky furnaces, attached garages, or poorly ventilated cooking.

This measurement gap has practical consequences. If you have a CO detector that goes off occasionally but never at alarm-level concentrations, you might assume the exposure is harmless. The acute poisoning threshold is well defined, but the line between “safe” chronic exposure and “potentially harmful” chronic exposure is far less clear. The cellular research described above suggests that even modest CO levels can shift cell behavior in directions relevant to cancer, but translating those lab findings into a real-world exposure guideline would require long-term studies that do not yet exist.

Non-Cancer Harms Worth Knowing About

Even setting the cancer question aside, chronic or severe CO exposure carries other serious health risks that are better established. One of the most unsettling is delayed brain injury after acute poisoning. A person may recover from the initial episode, seem fine for days or weeks, and then develop severe neurological or psychiatric symptoms.¹⁹ This delayed deterioration is thought to result from ongoing damage to the brain’s white matter, the insulating material around nerve fibers. It can manifest as memory loss, difficulty concentrating, personality changes, or movement problems, and recovery is not guaranteed.

Cardiovascular effects of CO exposure are also well documented. The heart is especially sensitive to oxygen deprivation, and even moderate CO levels can stress the cardiovascular system in people who already have heart disease. Carbon monoxide’s ability to shift hemoglobin’s grip on oxygen means that even after the CO itself is cleared, tissues may have sustained damage from the period of reduced oxygen delivery.

For people living with chronic low-level exposure from household sources, the more immediate concern is usually persistent symptoms that mimic other conditions: recurring headaches, unexplained fatigue, difficulty thinking clearly, and mild nausea. Because these complaints are so nonspecific, the CO connection is easily missed by both patients and their doctors. A CO detector with a digital readout showing parts per million, rather than just an alarm-only model, provides more useful information about whether low-level exposure is occurring.

What Remains Genuinely Uncertain

The honest state of the science is that CO sits in an uncomfortable middle ground. It is not classified as a carcinogen, and there is no direct experimental proof that inhaling carbon monoxide alone, at levels below acute poisoning, causes cancer in humans. The population studies that do exist are observational, not controlled experiments, and the exposures they capture are bundled with other chemicals. The cellular research is compelling but largely conducted in petri dishes and animal models, where conditions are far more controlled and concentrations far more precise than anything the human body encounters in daily life.

At the same time, the weight of evidence points to CO being more than an innocent bystander. It disrupts mitochondrial energy production in ways that generate DNA-damaging molecules. It activates repair pathways that can inadvertently protect cancer cells. It drives blood vessel growth that tumors depend on. And the HO-1 enzyme system, which produces CO endogenously, behaves as a tumor promoter in multiple cancer types when overactive. None of these mechanisms individually proves that breathing CO causes cancer, but collectively they describe a molecule with real biological leverage over processes that matter for cancer development.

The research community has not ignored this tension. The fact that CO-releasing molecules are being actively developed as anticancer therapeutics speaks to how context-dependent the molecule’s effects are. The same gas that may nudge a pre-cancerous cell toward malignancy under chronic low-level exposure might, when delivered at the right dose directly to a tumor, overwhelm the cancer’s energy metabolism and make it more vulnerable to treatment. Getting from lab results to clinical applications will require sorting out these dose-response relationships with far more precision than current data allow, and whether any CO-based therapy ever reaches routine clinical use remains to be seen.