Possible Plastic Contamination Does Not Make the Link BetweenMicroplastic and Nanoplastic Exposure and Disease Risk a Hoax

Key Point: Glove and lab supply contamination can change exposure counts and exposure pattern. It does not erase studies that purposely dose characterized particles. Minimal-contact field screening is one way to measure with fewer of those pathways.

This paper is also available at: https://doi.org/10.5281/zenodo.23130162

Abstract

A 2026 University of Michigan study showed that dry contact with ordinary nitrile and latex gloves can leave stearate salts that look like polyethylene under the microscope and in vibrational spectra. Traditional library matching can count those nonplastic residues as microplastics, especially below 10 micrometres. Press coverage extended that finding to the suggestion that microplastic health risk may have been overestimated. Some readers have gone further and treated the headline as if all microplastic studies were fake, and therefore as if microplastics do not cause disease.

Those conclusions do not follow. The Michigan study is about counting. Toxicological studies that purposely administer characterized microplastic or nanoplastic particles are a different experiment. Contamination in an environmental-measurement workflow does not change the dose those studies gave, or the response they recorded. What a false positive can change is the exposure number attached to an association, and the exposure pattern inferred from it: chronic background versus an episodic spike, particle number versus mass, and the smallest size bins most of all.

The practical response is better exposure measurement, not dismissal of the toxicology. Spectral libraries that include glove stearates can support reanalysis of affected datasets. Minimal-contact field methods such as EcoExposure avoid the glove route, skip multi-step isolation, and make repeated screening fast enough to distinguish a chronic signal from a one-time spike. They do not identify polymer, and they do not confirm or refute a dosing study.

1. The Headline Is Being Read As A Hoax. That Is The Wrong Category.

Microplastic and nanoplastic evidence comes in two kinds. One kind estimates what is in water, air, food, dust, or a biological specimen. The other kind takes a known particle and gives it to cells or animals on purpose. Risk assessment needs both. They do not answer the same question.

Clough and colleagues found a real flaw in the first kind of study. Stearate salts used to release gloves from the mold can transfer on dry contact. They resemble polyethylene visually and spectrally. With ordinary library matching, dry glove contact produced on the order of 2,000 false polyethylene counts per square millimetre; a cleanroom nitrile glove produced far fewer. Reanalysis with stearate standards reduced false positives in the smallest sizes, below 10 micrometres [1]. The study authors did not say the pollution is imaginary. Anne McNeil’s public formulation was that counts may be overestimated, and that there should still be none [2].

Fast Company, and outlets that followed it, led with health risk [3]. A reader who stops at the headline can walk away thinking every microplastic paper was an artifact, and that disease findings fall with the artifact. That reading confuses a false count with a false dose. A glove residue on a filter is not the particle an investigator weighed, characterized, and put into a dish or an animal.

2. Purposeful Dosing Studies Are Not Undone By A Counting Error

2.1 What the Michigan Study Showed

The study showed that a quality-control habit, wearing gloves, can introduce an interferent that is misclassified as the analyte. Identification is only as specific as the whole system: gloves, surfaces, blanks, library contents, match threshold, size range, and analyst review. A polyethylene label from an incomplete library is a workflow result, not a chemical fact.

 

2.2 What It Did Not Show

•        It did not show that every polyethylene detection is a stearate.

•        It did not estimate what share of published drinking-water or tissue counts are false. Its unit is false positives per square millimetre of contacted surface, not particles per litre.

•        It did not repeat any dosing experiment, and it did not show that characterized particles are biologically inert.

•        Its largest demonstrated effect was below 10 micrometres. The European drinking-water method starts at 20 micrometres, so the study does not size the error inside that reporting window [4].

 

2.3 Why A Dosing Study Stands On Its Own

In a controlled experiment the investigator chooses the material, states the nominal dose, picks the route, and records the response. Polystyrene beads are common because size and composition can be held still. They are not a stand-in for every weathered fragment in a river. If that experiment reports inflammation, oxidative stress, barrier disruption, altered signaling, or tissue deposition, a stearate on a different laboratory’s filter does not explain the finding.

Dosing studies have their own limits. The dose may be higher than ordinary exposure. The particles may be too uniform. The route may be instillation or gavage rather than drinking or breathing. Nominal dose may not be the dose that reached the cell. Those limits are reasons to be careful about human translation. They are not the glove artifact.

Figure 1. Environmental measurement workflows can introduce contamination or classification error, whereas controlled toxicology studies deliberately administer defined particles. A false environmental count does not alter the administered dose or recorded biological response.

3. What The False Positive Can Change: Level And Pattern Of Exposure

Risk interpretation depends on both hazard and exposure. A counting error changes the exposure estimate. It does not delete the hazard evidence produced by a dosing study.

If a survey counted stearates as polyethylene, the reported particle number is too high. Any association that used that number as the exposure is attached to the wrong concentration. Reanalysis can move the exposure axis down. It can also change the pattern. A glove touch is an episodic laboratory spike, not a chronic environmental background. A false-positive pile-up below 10 micrometres can make the smallest bin look dominant when part of that bin was residue. Number, mass, and surface area are not interchangeable: a small mass of nanoscale particles can dominate a count. A corrected count should be reported with mass, size distribution, and polymer confidence, not as one number per litre.

This is not a new claim that particles are more toxic per particle. It is a claim about the denominator. Downward recalibration of an association is plausible only if the outcome was linked to the biased measurement, the error was not differential in a way that invented the association, and the corrected metric is the one the association used. Inhalation, food, occupational exposure, missed nanoplastics, additives, and adsorbed chemicals can still dominate the measured water pathway. Confounding remains possible. A lower attached concentration is a hypothesis about exposure classification, not a demonstrated human threshold.

Figure 2. A glove-derived false positive can alter environmental abundance and exposure classification, but it cannot change the defined dose, route, duration, or observed response in a separate controlled toxicology experiment.

Table 1. What A Glove-Derived False Positive Can And Cannot Change.

Question	What a glove false positive can change	What it cannot change
Are the particles real in a dosing study?	Nothing. The particles were selected and administered.	Does not make a characterized dosing study a hoax.
Did the dose cause the reported effect?	Nothing about the administered dose or the observed response.	Does not erase inflammation, barrier change, oxidative stress, or deposition under the tested conditions.
What exposure level goes with an association?	The environmental count used as the exposure number, if that count included stearates called polyethylene.	Does not by itself set a human disease threshold.
What kind of exposure was it?	Whether the signal was chronic background, an episodic spike, or workflow residue. Small-size bins are the most exposed to this error.	Does not identify the true polymer mix unless the dataset is reanalyzed.
Is every published count wrong?	This specific concern applies most directly to glove-contact workflows that used spectral-library matching without adequate stearate controls; other false-positive pathways require separate evaluation.	Does not show that all polyethylene detections are false.

 

 

4. Standardized Spectroscopy Does Not Close The Gap By Itself

Commission Delegated Decision (EU) 2024/1441 uses optical microscopy or chemical mapping for size and shape, and vibrational microspectroscopy for composition, from 20 micrometres to 5 millimetres [4]. The library must include priority polymers, proteins, minerals, and natural polymers such as cellulose. It does not require glove-stearate standards. ISO 16094-2:2025 likewise couples microscopy with vibrational spectroscopy for low-solids waters [5]. Instrument class is not workflow specificity. A stearate that resembles polyethylene will still be called polyethylene if the library has no stearate and the match threshold lets it through.

World Health Organization reviews already separate the abundance problem from the health-evidence problem: exposure estimates are uncertain, particles are plausible hazards in experimental systems, and a human disease threshold is not established from either body of evidence alone [6, 7]. The glove paper sharpens the exposure side of that split. It does not close the health side.

 

 

5. A Practical Response: Count with Fewer Contamination Routes

Conventional laboratory characterization is still necessary when the question is polymer identity. It is a poor default when the question is whether today’s sample differs from last week’s, because every extra surface is another chance to count the method. Prior work mapped those routes across containers, filters, tubing, gloves, digestion, transfer, laboratory air, and analytical surfaces [8]. The Michigan study adds a concrete case: the control step meant to reduce contamination can cause it.

EcoExposure was built for home and field screening without that chain. Water stays in a glass vessel.

A plant-derived reagent is added from a paper sachet. The phone does not touch the water. There is no filtration, digestion, tubing, or transfer between vessels, and no glove on the analytical path. The user briefly handles the outside of the vessel. The design aim is pathway reduction and speed, so the same source can be checked repeatedly and a chronic signal can be told from a one-time spike.

That is not a claim of zero contamination, and it is not a claim of polymer identification. Air, the vessel, the sachet, and the image classifier can still add error. A field image cannot do what a stearate-aware Raman or infrared library does: name polyethylene and reject stearate. EcoExposure is a screen. Samples that matter still go to a corrected laboratory method. The advantage claimed here is the one the glove paper makes relevant: reduce the opportunity to introduce confounding material during measurement, and screen frequently enough that exposure pattern is observable.

Table 2. Conventional counting versus EcoExposure, limited to what each workflow can actually say.

Step	Conventional laboratory count	EcoExposure field workflow
Gloves on the sample path	Ordinary nitrile or latex can deposit stearate that matches polyethylene.	Gloves are not required. The user handles the outside of a glass vessel.
Isolation	Filtration, digestion, transfer, and surface contact over 24 to 48 hours.	No filtration, digestion, tubing, or vessel-to-vessel transfer.
Identification	Vibrational spectrum and library match. Specific only if stearates and other interferents are in the library.	Smartphone image of the vessel. Not a polymer identification.
What a positive means	A particle assigned to a polymer class by that workflow.	A field signal for follow-up, not a confirmed polyethylene count.
Speed and repeat measures	Slow. Poor fit for daily or episodic sampling.	Minutes. Built for repeated screening of the same source.
What it is not	Not automatically wrong. A corrected library can separate stearate from polyethylene.	Not a replacement for spectroscopy. Not proof the toxicology is right or wrong.

No blank counts are reported in this preprint. The comparison is a design comparison. Procedural blanks with and without glove contact are the next measurement, not a result already in hand.

 

6. What Should Happen Next

•        Stop reading a methods paper as a retraction of toxicology. Quote the dosing study’s particle, dose, route, and duration, or do not cite it.

•        Reanalyze glove-handled infrared and Raman sets with stearate standards, especially below 10 micrometres [1].

•        Publish raw blank counts. A sentence that blanks were used is not a blank result.

•        Report number, mass, size distribution, and assignment confidence together. Do not treat one count per litre as the exposure.

•        Separate chronic field signal from episodic handling spikes by repeating the measurement. Minimal-contact screening is useful here because it can be repeated; confirmatory spectroscopy remains the identification step.

•        Keep toxicology on environmentally plausible doses, weathered particles, and ingestion or inhalation routes, so translation limits are tested directly rather than borrowed from a counting controversy.

7. Conclusion

Ordinary laboratory gloves can seed false polyethylene counts. That is a counting problem. It is not evidence that every microplastic study is a hoax, and it is not evidence that particles dosed on purpose did nothing. Purposeful administration studies still stand or fall on their own particle, dose, route, and duration. The glove result can change the exposure level and the exposure pattern attached to an association: how high the count was, whether it was chronic or episodic, and whether the smallest size bin was partly residue. Expanded libraries and stearate-aware reanalysis can help correct affected datasets. Minimal-contact field screening, including EcoExposure, is one way to avoid the glove route and to sample often enough to see the pattern. It is a proposed improvement for exposure assessment. It is not a substitute for toxicology, and it is not a substitute for polymer identification.

This paper is also available at: https://doi.org/10.5281/zenodo.23130162

References

1. Clough ME, Ochoa Rivera E, Ayala AM, et al. Avoiding and reducing microplastic false positives from dry glove contact. Analytical Methods. 2026;18:2914-2926. doi:10.1039/D5AY01801C.

2. University of Michigan. Nitrile and latex gloves may cause overestimation of microplastics. EurekAlert. 26 March 2026. Includes Anne McNeil: “We may be overestimating microplastics, but there should be none.”

3. Mattson J. New research suggests the microplastics health risk may not be as bad as we thought. Fast Company. 3 April 2026.

4. European Commission. Commission Delegated Decision (EU) 2024/1441 of 11 March 2024 laying down a methodology to measure microplastics in water intended for human consumption. Official Journal of the European Union. 2024.

5. International Organization for Standardization. ISO 16094-2:2025. Water quality — Analysis of microplastic in water — Part 2: Vibrational spectroscopy methods for waters with low content of suspended solids including drinking water. 2025.

6. World Health Organization. Microplastics in drinking-water. Geneva: World Health Organization; 2019.

7. World Health Organization. Dietary and inhalation exposure to nano- and microplastic particles and potential implications for human health. Geneva: World Health Organization; 2022.

8. Chu MB. Potential plastic contamination pathways in conventional lab-based analysis of microplastics and nanoplastics vs. EcoExposure intentionally designed simple low-contamination field-deployable smartphone workflow. Zenodo. 2026. doi:10.5281/zenodo.19338947.