In March 2024, a group of Italian researchers published the most consequential paper on microplastics and human health in years. They had taken carotid artery plaque , the fatty deposits that build up in the main arteries supplying the brain , from 304 patients who had undergone surgery to clean out blocked vessels, and used pyrolysis gas chromatography-mass spectrometry to look for plastic particles inside. They found them. Then they followed the patients for a median of 34 months to see what happened next. The result, published in The New England Journal of Medicine, was the kind of headline that does not go away: patients with detectable microplastics in their plaque had 4.5 times the risk of heart attack, stroke, or death from any cause over the follow-up period, compared with patients whose plaque showed no detectable plastic [1].
For a field that had spent a decade producing a slow drip of "we found microplastics in another human tissue" papers, this was a different kind of finding. The 2022 Leslie study had detected microplastics in 77.4% of human blood samples from 22 healthy donors [2]. A 2021 Italian study found them in every placenta examined [3]. A 2023 Canadian study confirmed the placental finding and detected them in meconium , the first stool of newborns , direct evidence of foetal exposure [4]. A 2024 study of human testicular tissue found microplastics in every sample, at higher concentrations than in other tissues [5]. A 2022 study detected them in 75% of breast milk samples [14]. Microplastics are inside us. That part is no longer in serious doubt.
What the NEJM paper added was clinical: not just presence, but outcome. That changes the conversation.
What the Marfella study actually did
The study was prospective and multicentre, run out of Naples, and recruited 304 patients undergoing carotid endarterectomy , surgery to remove plaque from the carotid arteries, performed on people who have already had, or are at high risk of, a stroke. Each patient's excised plaque was analysed using pyrolysis gas chromatography-mass spectrometry, which breaks down a sample into its chemical constituents and is considered the gold standard for detecting and quantifying microplastics and nanoplastics in tissue. The technique identifies not just the presence of plastic but the specific polymer: polyethylene, polyvinyl chloride, polypropylene, and so on.
The researchers detected polyethylene in 58.4% of patients' plaque samples and polyvinyl chloride in 12.1%. Over a median 34 months of follow-up, the patients whose plaque contained detectable microplastics had a hazard ratio of 4.53 for the composite endpoint of myocardial infarction, stroke or death from any cause, after statistical adjustment for age, sex, BMI, smoking, diabetes, hypertension, hyperlipidaemia and prior cardiovascular events. The 95% confidence interval was wide (2.00 to 10.27), which reflects the relatively small cohort, but the association was strong enough to clear conventional significance thresholds by a comfortable margin [1].
The hazard ratio is the most quoted number from the paper, and it is the most likely to be misread. A hazard ratio of 4.53 does not mean a person with microplastics in their plaque is 4.5 times more likely, in any absolute sense, to have a heart attack. It means their relative risk of the composite outcome, after adjustment, was that much higher than the comparison group. In a carotid endarterectomy cohort , all of whom already had significant vascular disease , the baseline 3-year rate of major adverse cardiovascular events is on the order of 7 to 8%. If the hazard ratio is applied to that baseline, the MNP-positive group had a 3-year event rate closer to 25 to 30%. That is still a striking finding, but it is a relative shift in a high-risk population, not a population-wide claim [12].
What the study did not show
The same paper that put microplastics on the front page also set clear limits on what it could, and could not, conclude.
It was a single study, in 304 patients, all Italian, all already undergoing carotid surgery. It did not include a low-risk or general-population comparison group. It could not establish that removing microplastics would reduce risk, because nobody has yet figured out how to remove microplastics from vascular tissue. It did not measure microplastics in any other tissue, so it could not say whether plaque presence reflects broader body burden or some other process specific to the diseased vessel wall. The authors themselves were careful in their framing: this was an association in a high-risk cohort, not a causal mechanism [1].
The accompanying editorial in The Lancet Planetary Health, written by Kurunthachalam Kannan and Krishnamoorthi Vimalkumar, made the caveats more explicit. "The finding is a wake-up call," they wrote, "but replication in larger and more diverse cohorts is essential before generalising to the wider population" [9]. A 2024 systematic review in Environmental Science: Processes & Impacts, the journal of the Royal Society of Chemistry, went further, noting that the field still lacks a standardised method for detecting and quantifying microplastics in human tissue. Detection rates between studies vary by orders of magnitude, and some early results with very high detection rates have been challenged on contamination grounds [10].
The WHO's 2023 position on microplastics in drinking water is also worth weighing. Released before the NEJM paper, the WHO concluded that, on the available evidence, exposure to microplastics in drinking water at current levels is "low concern" for human health, and recommended that water utilities and regulators prioritise established contaminants over microplastics. The WHO explicitly called for more research, particularly on nanoplastics , particles smaller than one micrometre that are harder to detect and may behave differently in the body [6].
None of this is a refutation of the Marfella finding. It is the context required to read it correctly.
Why the cardiovascular link is biologically plausible
The reason the paper got so much attention is not only the size of the effect, but that the mechanism is plausible. Microplastics in the bloodstream can act as foreign-body irritants. In cell culture and animal studies, microplastics and nanoplastics have been shown to trigger oxidative stress, endothelial dysfunction, and chronic low-grade inflammation , all of which are known contributors to atherosclerotic plaque progression and rupture [10]. Particles may also act as a scaffold for the deposition of lipids and proteins, accelerating the formation or destabilisation of plaque. None of these mechanisms has been confirmed as the operative one in humans, but the underlying biology is consistent with the clinical signal.
A separate strand of evidence comes from occupational and environmental exposure studies. Workers with high exposure to particulate air pollution, including the fine particulate matter PM2.5, have well-documented increases in cardiovascular risk; the biological pathways (systemic inflammation, oxidative stress, endothelial dysfunction) are similar to those proposed for microplastics. Microplastics are a relatively new entrant to a category of risk that cardiovascular epidemiologists have been studying for decades.
Where the science stands in 2026
In the two years since the Marfella paper, the field has continued to produce confirmatory and explanatory work. Larger cohort studies are in progress; basic-science work on mechanisms has accelerated; and the policy response has begun. The European Union, in September 2023, adopted a comprehensive restriction on intentionally added microplastics under its REACH chemicals regulation, targeting microplastics in cosmetics, detergents, and personal care products. When fully implemented, with staged deadlines through 2026 and beyond, the restriction is expected to prevent around 500,000 tonnes of microplastics from being released into the environment over 20 years [7]. It is the most ambitious microplastics policy in the world to date, though industry groups have challenged parts of it in court.
In the United States, the FDA has not set specific limits on microplastics in food or drinking water, but the National Institutes of Health has funded several large cohort studies that will help clarify exposure and outcomes. CDC NHANES data is being analysed to establish baseline exposure in the general US population. Australia, where this article is being written, has not yet adopted a microplastics-specific restriction beyond general product-safety regulation, though the federal government has funded research through the Minderoo Foundation's Global Plastics Map and similar efforts.
The honest assessment in mid-2026 is this: microplastics are real, they are inside essentially everyone tested, and the strongest single piece of clinical evidence , the Marfella NEJM paper , points to a meaningful cardiovascular signal in a high-risk population. The evidence in lower-risk populations is still thin, replication is ongoing, and we do not yet have a verified mechanism in humans. The hazard ratio is striking, but the absolute risk for an average healthy adult is much smaller than the relative risk suggests. Anyone who tells you that microplastics are a confirmed cause of heart attacks in the general population is over-reading the data. Anyone who tells you they are nothing to worry about is also over-reading the data, just in a different direction.
What the evidence actually supports doing
This is where the practical guidance has to get specific, because the public-health advice has to be defensible against the evidence we have, not the evidence we wish we had.
The interventions most consistently recommended by clinicians writing in mainstream outlets , Harvard Health [12], The Guardian [14], and the WHO [6] , are pragmatic and largely overlap. They are not about eliminating microplastic exposure, which is not currently possible. They are about reducing the most concentrated sources.
The first is heat. Heating food in plastic containers , putting a plastic takeaway box in the microwave, putting a plastic container in a dishwasher at high heat, leaving a plastic water bottle in a hot car , accelerates the migration of microplastics into food and water. The Harvard Health explainer on the Marfella study and the Guardian's practical guide both make this the single most evidence-backed personal-level action [12][15]. Glass or ceramic for microwave use, and stainless steel or glass for hot liquids, are the simple swaps.
The second is ultra-processed food. A 2023 study in Environmental Pollution found that a serving of ultra-processed food contains, on average, significantly more microplastics than a serving of minimally processed food. The mechanism appears to be contact with plastic packaging and processing equipment during manufacture. Reducing ultra-processed food intake is, in any case, good general health advice; reducing microplastic ingestion is a side benefit.
The third is drinking water. Where tap water is unfiltered, a high-quality activated carbon or reverse-osmosis filter reduces microplastic load, though it does not eliminate it. The WHO 2023 position is that current drinking-water exposure is below levels of established health concern, so this is a precautionary measure rather than a clinical imperative [6].
None of these interventions has been tested in a randomised trial for cardiovascular outcomes. The microplastics evidence base is observational, mechanistic, and still developing. But the actions being recommended are individually low-cost, low-risk, and have plausible mechanisms, and they overlap with general cardiovascular and dietary advice that is on firmer evidence. If you are already eating well, exercising, not smoking, and managing your blood pressure and cholesterol, the marginal benefit of these specific actions is probably small. If you have other cardiovascular risk factors and want to address every plausible contributor, they are reasonable additions to a clinician-supervised plan.
What the evidence does not yet support is panic. It does not support expensive "microplastic detox" products, none of which has been shown to remove microplastics from human tissue. It does not support foregoing evidence-based medical care in favour of unproven alternative approaches. It does support treating the Marfella finding as a serious signal that deserves replication, prudent personal action where the action is sensible for other reasons, and continued research investment at the policy level.
The 4.5x hazard ratio is the kind of number that spreads because it deserves attention. It is also the kind of number that needs context before it can be turned into sensible behaviour. The science is moving faster than the certainty. For now, the most useful response is to take the signal seriously, the absolute risk in perspective, and the action steps pragmatically.