Graphene Oxide Side Effects in Humans: What We Know So Far

Most of what researchers know about graphene oxide side effects comes from cell cultures and animal experiments, not from watching people get sick. Only one controlled human inhalation study has been published so far, and in that trial, thin graphene oxide nanosheets were well tolerated at low doses with no overt harmful effects on heart rate, blood pressure, or lung function.¹ That single, carefully controlled experiment is reassuring but extremely narrow. The animal and lab data, by contrast, paint a more complicated picture in which dose, particle size, surface chemistry, and route of exposure all shift the safety profile dramatically.

The Only Human Data So Far

In 2023, researchers at the University of Edinburgh published the first controlled human exposure study. Healthy volunteers inhaled highly purified, thin graphene oxide nanosheets in two different size ranges inside a clinical facility. The team tracked heart rate, blood pressure, lung function, and blood inflammatory markers. None of those measures changed meaningfully compared to a clean-air control session. A more sensitive blood proteomics analysis turned up very few differences in plasma proteins, and an ex vivo model showed a mild increase in clot formation at the site of a simulated arterial injury.¹ The authors stressed that the study was designed as a proof of concept for doing this kind of testing safely, not as a comprehensive safety clearance. The doses were low, the exposure was brief, and the participants were young and healthy. For now, virtually everything else we know comes from laboratory cells and animal models.

Oxidative Stress as the Central Mechanism

If there is one thread connecting nearly every reported side effect of graphene oxide, it is oxidative stress. When GO particles interact with cells, they tend to ramp up the production of reactive oxygen species, which are chemically aggressive molecules that can damage fats, proteins, and DNA. In human lung cells, GO exposure raised reactive oxygen species levels and depleted the cells’ built-in antioxidant defenses, and blocking that oxidative burst with an antioxidant compound partially rescued cell survival.² A separate live-imaging study found that larger GO sheets, which tend to sit on the cell surface rather than being swallowed up, triggered stronger oxidative responses and more inflammation than smaller ones that were readily internalized.³

Oxidative stress does not just weaken cells generally. One particularly damaging consequence is lipid peroxidation, where reactive molecules chew through the fatty membranes that hold cells together. In a macrophage model, the most highly oxidized form of GO triggered lipid peroxidation in roughly a third of the cell population, eventually leading to membrane rupture and cell death.⁴ This same lipid peroxidation pathway has been implicated in heart tissue damage in animal studies, where markers of membrane fat breakdown climbed steadily with increasing GO dose.⁵

How the Immune System Responds

GO does not slip past immune cells unnoticed. Macrophages, the body’s frontline scavengers, recognize and attempt to engulf GO sheets. What happens next depends a lot on the size of those sheets. Larger GO flakes tend to drape across the macrophage surface rather than being fully swallowed, which provokes a stronger inflammatory signal. In both cell cultures and live animals, larger GO drove macrophages toward an aggressive inflammatory state with elevated production of signaling molecules that recruit more immune cells to the area.⁶ That response was consistent whether GO was introduced into the abdomen, the lungs, or the bloodstream.

Beyond general inflammation, GO can also flip a more specific immune switch called the inflammasome. In primary human macrophages that had been primed with a bacterial signal, GO triggered a cascade ending in the release of IL-1β, a potent inflammatory messenger. That process required the cells to actually take in the GO particles and involved both oxidative stress and potassium leaking out of the cell.⁷ The good news from materials engineering is that reducing GO (stripping away some of its oxygen groups through heat treatment) substantially dialed down both the oxidative stress and the inflammatory cytokine secretion in macrophage experiments, pointing toward a tunable lever for making GO safer.⁸

Lung Damage and Fibrosis

Inhalation is the exposure route that worries occupational-health researchers most. When rats received GO directly into their windpipes and were monitored for three months, the animals developed pulmonary fibrosis, a thickening and scarring of lung tissue that makes it harder to breathe. The severity tracked with dose, and the researchers linked it to activation of a signaling chain involved in tissue remodeling and scar formation.⁹ That three-month window matters because it hints at chronic consequences, not just a temporary irritation. And the damage did not stay confined to the lungs; other organs also showed signs of injury at higher doses.

This is worth keeping in perspective. Intratracheal instillation delivers a bolus of material straight into the airways, which is a much harsher scenario than breathing in trace airborne particles over a workday. Still, the fibrosis finding is the main reason occupational guidelines for graphene materials lean on precaution.

Effects on the Liver, Kidneys, and Spleen

Once GO enters the bloodstream, it does not distribute evenly. The liver, spleen, and kidneys consistently show up as the organs that accumulate the most material. In rats given GO nanoplatelets by injection, the liver developed visible clusters of particles at its surface, while kidney tissue showed inflammation, shrunken filtering units, and, at the highest dose tested, a granulomatous reaction with giant cells forming around trapped particles. The spleen accumulated particles along its outer capsule in a dose-dependent fashion.¹⁰

A separate study using a PEG-coated reduced GO found similar organ accumulation patterns and confirmed liver toxicity and immune activation during the early days after exposure. Critically, the material was not fully cleared from the body during the study period, raising questions about long-term effects from lingering particles.¹¹ These findings are why biocompatibility researchers pay so much attention to whether the body can break down and excrete GO, a topic covered further below.

Blood Compatibility and Cardiovascular Concerns

When GO meets red blood cells in a test tube, the outcome depends sharply on concentration. At very low levels, human red blood cells showed no visible changes. But at higher concentrations, the cells changed shape and began to rupture, and blood clotting parameters shifted abnormally.¹² In heart tissue studied in animals, GO and its reduced form both triggered oxidative damage to heart cells with worsening effects at higher doses, involving mitochondrial dysfunction alongside the lipid peroxidation described earlier.⁵ The mild uptick in clot formation noted in the human inhalation study aligns with these lab findings about blood interactions, though the clinical significance of that mild shift is still unclear.

The Brain and the Blood-Brain Barrier

One counterintuitive finding involves PEGylation, the process of attaching polyethylene glycol chains to a nanomaterial to make it more “stealthy” in the body. In many drug-delivery contexts, PEG coatings improve biocompatibility. But in a study of reduced GO’s interaction with brain cells, PEGylation actually made things worse. At the highest concentrations tested, PEG-coated reduced GO was more toxic to both astrocytes and endothelial cells than the uncoated version. In live rats, injecting PEG-coated reduced GO led to a progressive weakening of the blood-brain barrier over the following week, with key structural proteins declining and signs of oxidative damage accumulating.¹³ This was a single study, and the mechanisms are not fully understood, but it serves as a cautionary note that surface modifications assumed to be protective do not always behave as expected in every tissue.

Skin and Eye Safety

Skin contact appears to be relatively benign compared to inhalation or injection. A standardized sensitization test using both few-layer graphene and GO found no evidence of skin sensitization or irritation in mice. There was no immune cell infiltration and no allergic-type response.¹⁴

Eyes are more vulnerable. A single brief exposure to GO did not cause obvious irritation in rabbit eyes, and human corneal cells survived a two-hour exposure without dying. But over 24 hours in cell culture, GO caused significant toxicity to both corneal and conjunctival cells through oxidative stress. Repeated exposure over five days produced reversible corneal cloudiness and redness in rats, effects that an antioxidant could partially prevent.¹⁵ A mouse study found that repeated GO exposure caused noticeable inflammation inside the eye and corneal cell death, though the reduced form of GO did not cause similar damage.¹⁶ The takeaway for anyone handling GO in a lab is straightforward: gloves are probably fine, but eye protection is non-negotiable.

DNA Damage

Several studies have looked at whether GO can damage genetic material. The picture is mixed but leans toward concern at higher doses. In one experiment, human kidney cells exposed to elevated GO concentrations showed reduced survival and increased DNA strand breaks, and zebrafish embryos exposed to GO displayed similar damage along with activation of a DNA repair pathway.¹⁷ A broader review of genotoxicity studies concluded that some graphene-based materials do have the potential to cause genetic damage, though the type and severity depend heavily on the specific material tested.¹⁸ Interestingly, the oxygen-containing groups on GO’s surface may partly protect against DNA damage; when those groups were chemically stripped away, DNA damage increased, suggesting that GO is paradoxically less genotoxic than more pristine forms of graphene.¹⁹ No long-term carcinogenicity studies in humans exist, so whether these DNA effects translate to cancer risk remains an open question.

Reproductive and Developmental Effects

Animal studies on reproduction raise some red flags, though all come with heavy caveats about dose and timing. When pregnant mice were given reduced GO nanosheets by injection, those treated early in pregnancy could deliver live pups, but some litters contained malformed offspring. Mice injected late in pregnancy fared much worse: all experienced pregnancy loss at lower doses, and most died at the highest dose tested.²⁰ A more recent study using very low oral doses of GO during pregnancy and nursing found no changes in litter size or survival but did observe an altered sex ratio and behavioral changes in the mothers. The offspring generation showed amplified neurobehavioral effects, including increased stress sensitivity and reduced learning and memory.²¹ The intergenerational amplification in that second study is the kind of finding that grabs researchers’ attention because it suggests effects could compound across generations even at exposures that seem negligible for the mother.

Gut Microbiome Disruption

If GO reaches the digestive tract, it does not just pass through inertly. An in vitro model of the human gut microbiome showed that GO exposure reshuffled bacterial community composition, with a notable drop in one major bacterial group and a shift in the ratio between dominant phyla. The metabolic output of the microbial community also changed: the bacteria processed fewer amino acids and carbohydrates and produced less of certain beneficial metabolites.²² In mice given GO orally, electron microscopy revealed damage to the colon’s cellular architecture, including swollen mitochondria, disrupted cell junctions, and shrunken nuclei. The gut microbial community also shifted, with several bacterial genera declining.²³ These studies are early-stage, but they add the gut to the list of systems where GO does not behave as a passive bystander.

Why Size, Surface Chemistry, and Dose Change Everything

One of the frustrating aspects of graphene oxide toxicology is that “graphene oxide” is not a single material. The stuff produced in one lab can differ enormously from another lab’s batch in terms of lateral size, thickness, degree of oxidation, residual impurities, and surface coatings. All of these properties influence how toxic the material is. A comprehensive review identified lateral size, surface structure, functionalization, charge, impurities, aggregation tendency, and the protein “corona” that forms around particles in biological fluids as key determinants of toxicity.²⁴

Size effects can run in surprising directions depending on the endpoint measured. While larger GO sheets provoke stronger inflammatory responses from macrophages, smaller GO particles have shown greater potential to reduce cell survival and cause DNA damage in other test systems.²⁵ This means there is no universally “safer” size. The best size depends on which organ is exposed and which kind of damage you are trying to avoid.

Surface modification is equally complicated. Coating GO with polymers like polyacrylic acid improved biocompatibility in both cell cultures and animals more effectively than the widely used PEG coating.²⁶ Yet PEGylated GO still triggered strong immune responses despite the assumption that PEG would “passivate” the surface.²⁷ The lesson is that strategies borrowed from other nanomaterial fields do not automatically transfer to graphene.

Can the Body Break Down and Clear Graphene Oxide?

This question matters enormously for long-term risk. If GO lingers in tissues indefinitely, even modest toxicity could compound over years. The evidence suggests the body does have some capacity to degrade it, but the process is selective. Human neutrophils produce an enzyme called myeloperoxidase that can completely break down well-dispersed, highly oxidized GO sheets in the presence of hydrogen peroxide. The catch is that the enzyme fails to degrade GO that is heavily aggregated or poorly dispersed.²⁸ Even pristine graphene (without the oxide groups) can be degraded by the same enzyme if the sheets are well-dispersed in water.²⁹

In mice injected with GO intravenously, the spleen emerged as a major collection point. Using Raman spectroscopy to track the material, researchers found GO concentrated inside a specific subset of splenic immune cells, the marginal zone macrophages, which are positioned to intercept foreign particles from the blood.³⁰ A pharmacokinetics study found that GO of all tested sizes was also excreted through the urine, and that sheet thickness was a bigger factor than width in determining whether particles got filtered out. Larger sheets cleared more slowly, though all three sizes tested showed significant urinary excretion.³¹ So the body uses at least two pathways: enzymatic breakdown by immune cells and physical filtration through the kidneys. Neither pathway is guaranteed to catch everything, which is why incomplete clearance and organ accumulation remain open concerns.

Occupational Exposure in the Real World

Most people are unlikely to encounter graphene oxide in their daily lives. The exposure scenarios that matter today are occupational: lab researchers, production workers, and people involved in scaling up graphene manufacturing. A 2025 review of workplace studies found that exposure during production and handling activities was generally low, largely because graphene materials are typically processed in small quantities, in liquid suspensions, or inside closed equipment. Workers using personal protective equipment faced low risk under current conditions.³² The review flagged two scenarios where risk increases: handling dry GO powder and scaling up production volumes. Both generate more airborne particles. Given the uncertainties around long-term exposure, the authors recommended a precautionary approach and adapting traditional industrial hygiene practices to account for the nanoscale size of the particles.

Environmental Accumulation and the Food Chain

As graphene manufacturing grows, some material will inevitably reach waterways and soil. Lab experiments have shown that graphene can accumulate in aquatic organisms from bacteria up through small crustaceans and fish, and that dietary transfer through the food chain can actually deliver more material to an organism than direct exposure to contaminated water.³³ That said, not every food-chain link amplifies the problem equally. In one experiment, shrimp fed algae that had aggregated with GO did not show stress or accumulation, suggesting that GO-algae clumps may not transfer efficiently to larger organisms.³⁴ Environmental risk assessment for graphene materials is still in its infancy, and real-world concentrations in waterways remain far below the levels used in these experiments. But the bioaccumulation potential means this is a question researchers cannot afford to ignore as production scales up.