Microplastics: What We Know and Don't Know
Microplastics are in human blood, lungs, placentas, and brains. Here is what the measurement science shows, what early health studies suggest, and where the evidence runs out.
By Velloreth
Plastic does not disappear. It fragments. A bottle left in the sun becomes smaller and smaller pieces until the pieces are invisible — microplastics, under 5 millimeters, and nanoplastics, under a micrometer. These particles are in ocean water, soil, Arctic ice, rain, drinking water, food, and air. They are also, by a growing body of direct measurement, in human blood, lungs, placentas, reproductive organs, and brains.
That presence is established. Whether it matters for health is the question researchers are racing to answer — and the honest current state of the evidence is more nuanced than either “plastic is killing us” or “it’s fine” suggests.
Definitions and sources
Microplastics are commonly defined as plastic particles smaller than 5 mm, with some standards setting the lower limit at 1 µm. Nanoplastics are smaller — roughly 1 nm to 1 µm — and behave differently: their surface-area-to-mass ratio is much higher, they cross biological barriers more easily, and they are harder to detect.
Two categories:
- Primary microplastics are manufactured small: microbeads in cosmetics (banned in many jurisdictions), industrial abrasives, pellets lost from plastic production and transport (nurdles), fibers shed from synthetic textiles in laundry.
- Secondary microplastics are fragments of larger items — bottles, bags, packaging, tires, fishing gear — broken down by UV light, mechanical wear, and weathering.
The polymers detected most often in human samples are polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC), polyethylene terephthalate (PET), and polystyrene (PS). These are the workhorses of packaging, textiles, and construction. Their ubiquity in the environment tracks directly with their ubiquity in the samples.
Humans are exposed through three routes: ingestion (food and water, especially from plastic packaging), inhalation (indoor air, synthetic fibers, tire wear particles), and, less characterized, dermal contact. Studies of plastic food containers and bottled water repeatedly find particle release, accelerated by heat, UV exposure, and mechanical wear. One study of bottled water brands reported average particle counts in the thousands per liter using newer Raman methods capable of resolving nanoplastics.
What detection science shows — and its limits
A recurring theme in the microplastics literature is that the measurement problem shapes the conclusions. Different labs use different methods, thresholds, and size ranges, and cross-study comparison is difficult.
The main analytical tools:
| Technique | What it does | Limitation |
|---|---|---|
| Pyrolysis GC/MS | Burns a sample, identifies polymer by chemical signature | Bulk measurement; no size or shape information |
| µ-FTIR | Infrared imaging identifies polymer types | Diffraction limit: poor below ~10–20 µm |
| Raman spectroscopy | Identifies polymers at smaller sizes than FTIR | Slow; fluorescence interference |
| SEM/TEM | Images particles at nanometer scale | Does not identify polymer chemistry alone |
The nanoplastic detection gap is the most consequential limitation: the smallest, most bioavailable fraction of plastic particles is systematically undercounted by most studies, because conventional spectroscopy cannot resolve individual particles below roughly 500–1000 nm, and biological samples contain a haze of natural nanoparticles (protein aggregates, lipoproteins, extracellular vesicles) that are hard to distinguish from synthetic ones. Recent advances — including stimulated Raman approaches and machine-learning-assisted spectral analysis — are beginning to close this gap, and early results suggest nanoplastic counts in some matrices are far higher than previous microplastic-only methods reported.
Contamination control is another issue. Plastic is everywhere in lab environments — gloves, tubing, air handling — so rigorous blank controls are essential. A study without contamination blanks is not trustworthy. This is one reason results vary so widely across the literature, and why healthy skepticism about individual outlier studies is warranted.
Where the human health evidence stands
A 2025 systematic review of in-vivo human studies — 25 studies, about 1,500 participants — captures the current landscape. Most evidence is early: small samples, cross-sectional designs, variable methods. But some signals are consistent.
Cardiovascular
The strongest human data. A 2024 New England Journal of Medicine study found polyethylene microplastics in atherosclerotic plaques removed during surgery; patients with microplastics in their plaques had a higher combined risk of myocardial infarction, stroke, or death over follow-up than patients without. Other cardiovascular studies have found microplastics in thrombi from stroke and heart attack patients, and associations between blood microplastic burden and inflammatory markers and coagulation changes.
These are associations in observational designs. They do not prove causation — it is possible that sicker people have more vascular inflammation that traps more particles, or that both are driven by a third factor (smoking, occupational exposure, air pollution generally). But the mechanistic plausibility is there: microplastics activate macrophages, promote oxidative stress, and in cell and animal models disrupt fibrinogen structure and promote clot formation.
Reproductive and developmental
Microplastics have been detected in placentas, meconium, breast milk, semen, ovarian follicular fluid, and testicular tissue. Animal studies of polystyrene exposure show effects on sperm parameters and pregnancy outcomes at high doses. Human data is limited; the placenta findings are particularly striking but not yet linked to clinical outcomes in adequately powered studies.
Respiratory and gastrointestinal
Lung tissue studies have found microplastics in both cancerous and non-cancerous tissue, with higher counts in some cancer samples — again, association without established causation. Occupational studies (shoe manufacturing, textile workers) show high lung burdens. In the gut, microplastics are consistently detected in stool; cell studies show pro-inflammatory effects and microbiome modulation, but clinical significance in humans is unknown.
Neurological
A 2025 study reported higher microplastic abundance in the brains of deceased individuals with dementia than without, particularly in blood vessel walls and immune cells. This is a single study of postmortem tissue. The blood-brain barrier crossing of nanoplastics has been demonstrated in mice. The leap from “particles are present in higher amounts in dementia brains” to “particles contribute to dementia” has not been made; it is a hypothesis now being actively tested.
What the toxicology adds — and its limits
Cell and animal studies consistently show oxidative stress, inflammation, and endocrine disruption at sufficient doses. Nanoplastics cross the gut barrier, lung barrier, placental barrier, and blood-brain barrier in model systems. Mitochondrial dysfunction and reactive oxygen species production appear to be core initiating events.
The dose problem is the major caveat: many laboratory exposures use concentrations and particle sizes that may not reflect real-world human exposure, and the particles used — pristine polystyrene spheres of uniform size — differ from the weathered, irregular, biofilm-coated fragments found in the environment. Extrapolating directly from a high-dose mouse study with perfect spheres to a human breathing city air for decades requires assumptions that have not been validated. This is not a reason to dismiss the toxicology; it is a reason to demand human-relevant dose-response data before making strong causal claims.
What you can reasonably do
Regulation is behind the science — there is no coordinated international standard for microplastic limits in food, water, or air, and no certified reference materials for measuring them in biological samples. Waiting for regulatory certainty could take years. The precautionary actions with the clearest mechanism:
- Do not heat food in plastic. Microwave in glass or ceramic. Heat increases leaching and particle release measurably.
- Reduce single-use plastic contact with food and drink. Not for the ocean alone — for your own exposure.
- Choose glass or stainless steel for water bottles and food storage, especially for acidic or fatty foods, which pick up more from plastic.
- Wash synthetic textiles less aggressively if you care about fiber shedding — full loads, lower spin speeds, and cold water reduce fiber release, and some washing-machine filters capture a large share of shed fibers.
- Ventilate and reduce indoor dust. Synthetic fibers and dust are major indoor sources; a HEPA vacuum and open windows help.
- Filtered water over bottled, where available — reverse osmosis or advanced carbon filters reduce particles; most basic filters do not.
None of these eliminate exposure — it is ambient — but they reduce the highest-leverage contact points, particularly heat-accelerated leaching and concentrated indoor sources.
What would change the picture
Three developments would move this field from “concerning presence” to “quantified risk”:
- Standardized measurement. Reference materials, agreed protocols for biological matrices, and nanoplastic-capable methods that every lab can run. Without this, studies cannot be pooled and dose-response curves cannot be built.
- Prospective cohort studies. Measuring microplastic burden in living people and following health outcomes for years, with proper confounder control. The cardiovascular plaque studies are a start; the design needs to scale.
- Mechanistic confirmation at realistic doses. Not “what happens when we expose cells to a gram of polystyrene spheres” but “what happens at the concentrations found in human tissue, with environmentally realistic particle shapes and chemistries.”
Until then, the accurate statement is this: microplastics are universally present in human bodies, the early human evidence points most strongly toward cardiovascular effects, the toxicology provides plausible mechanisms, and the population-level risk remains unquantified. That is an uncomfortable place to stand, but it is where the data actually is — and honesty about the gap is more useful than either panic or complacency.