Qualité des preuves 4.88/5
Score d’évaluation en huit dimensions selon la grille de qualité . Chaque dimension notée de 1 à 5.
- D1 Ancrage des sources
- 4/5
- D2 Autorité des sources
- 5/5
- D3 Arithmétique
- 5/5
- D4 Incertitude
- 5/5
- D5 Portée
- 5/5
- D6 Prose
- 5/5
- D7 Honnêteté sur la perception
- 5/5
- D8 Complétude des réserves
- 5/5
Comment le risque varie
Le chiffre principal est une moyenne sur des situations très différentes. Voici comment la probabilité varie selon le scénario ou le contexte :
—
No epidemiological study has measured attributable disease from dietary or airborne microplastic exposure in the general US population. Exposure is real; harm is unquantified.
—
Workers in synthetic textile and plastics manufacturing have higher inhalation exposure. Some occupational studies report elevated respiratory symptoms (flock worker's lung), but these involve specific fiber types at industrial concentrations, not general microplastic exposure. Signal is weak and confounded by co-exposures.
—
Shellfish and small pelagic fish bioaccumulate microplastics. Consumers of large quantities of mussels, oysters, and sardines have higher ingestion estimates, but no cohort study has linked this dietary pattern to measurable health outcomes in humans.
—
Bottled water contains roughly 10-100x more microplastic particles per liter than tap water in most studies, likely from packaging. WHO's 2019 review found no health risk from either source at observed levels.
La longueur et la teinte de la barre classent ces scénarios les uns par rapport aux autres, pas par rapport à d'autres risques. Les probabilités exactes sont indiquées à côté de chacun.
Le fossé entre la peur et les preuves concernant les microplastiques est large, mais il s’étend dans une direction inhabituelle : la peur n’est pas que la science affirme que les microplastiques sont sûrs — c’est que la science n’a pas encore grand-chose à dire de concret. L’examen de 2019 de l’OMS sur les microplastiques dans l’eau potable n’a trouvé « aucune preuve fiable » de risque pour la santé aux niveaux d’exposition actuels. L’EFSA, après avoir examiné plus de 100 études, n’a pas publié de chiffre de risque quantifié et ne prévoit pas de finaliser son avis formel avant la fin de 2027. Aucune cohorte épidémiologique n’a mesuré de maladie humaine attribuable à l’exposition aux microplastiques par l’alimentation ou l’air aux niveaux de la population générale. L’entrée porte la mention no_reliable_estimate non pas parce que le risque est connu pour être nul, mais parce que personne ne l’a mesuré suffisamment bien pour lui attribuer un chiffre.
Ce qui alimente la peur est une heuristique « détection égale danger ». Leslie et al. (2022) ont trouvé des particules de microplastiques dans 77 % des échantillons de sang humain — une découverte authentique et frappante qui prouve une exposition systémique. Ragusa et al. les ont trouvés dans les placentas. Marfella et al. (2024), publié dans le NEJM, ont rapporté que les patients présentant des micro- et nanoplastiques détectables dans la plaque de l’artère carotide avaient un taux d’événements cardiovasculaires 4,5 fois plus élevé que ceux qui n’en avaient pas. Mais la cohorte de Marfella était présélectionnée (tous avaient une maladie carotidienne asymptomatique nécessitant une intervention chirurgicale), l’étude manquait de contrôles anti-contamination dans l’environnement opératoire, et l’association dans une population malade n’est pas une causalité dans le grand public. Pendant ce temps, l’affirmation d’exposition d’« une carte de crédit par semaine » qui a ancré la perception du public s’est avérée surestimer l’ingestion réelle d’un à deux ordres de grandeur — les estimations réalistes se situent autour de 0,01-0,1 gramme par semaine, et non 5 grammes.
Le chiffre n’existe pas encore, et cela compte plus que n’importe quel substitut. L’exposition professionnelle (travailleurs du textile et du plastique inhalant des fibres synthétiques à des concentrations industrielles) dispose d’une base de preuves mince mais réelle pour les effets respiratoires, bien que confondue par des co-expositions. Les grands consommateurs de fruits de mer ingèrent plus de particules via la bioaccumulation dans les coquillages. L’eau en bouteille contient environ 10 à 100 fois plus de particules par litre que l’eau du robinet. Aucun de ces gradients d’exposition n’a été lié à des résultats de santé mesurables chez l’homme. Le domaine est pré-épidémiologique : nous savons que l’exposition est universelle, nous suspectons une plausibilité biologique à partir de modèles animaux et in vitro, et nous avons une étude d’association provocatrice mais méthodologiquement contestée. Ce n’est pas rien, mais ce n’est pas non plus une probabilité.
Pour aller plus loin
Les microplastiques se retrouvent dans le sang, les poumons et les placentas humains, et la majorité des adultes les classent parmi leurs principales craintes de sécurité alimentaire. Mais l'OMS n'a trouvé aucune preuve fiable que les microplastiques, aux niveaux d'exposition actuels, nuisent à la santé. Détecter n'équivaut pas à un risque mesuré.
Registre des sources
Chaque chiffre ci-dessous correspond à ce que la source a rapporté, avec la citation textuelle sur laquelle nous nous sommes appuyés et la méthode de calcul. Cliquez sur un lien pour vérifier directement.
1/7 sources vérifiées de manière indépendante, mot pour mot, par rapport à la source citée
-
[1] World Health Organization — Microplastics in Drinking-Water
Microplastics in Drinking-WaterSee all 3 Likelier entries citing this source →
- Statistique
No reliable evidence that microplastics in drinking water at current exposure levels pose a risk to human health- Extrait
“"Based on the limited information we have, microplastics in drinking water don't appear to pose a health risk at current levels. But we need to find out more." ”
- Données source de
- 2019-08-22
- Consulté le
- 2026-04-18 · copie archivée
- Calcul
- WHO reviewed all available evidence on microplastic occurrence in drinking water (tap and bottled) and assessed potential health impacts from ingestion. The report concluded that particles larger than 150 micrometres are unlikely to be absorbed, uptake of smaller particles is expected to be limited, and no epidemiological data link drinking-water microplastic exposure to adverse health outcomes. WHO called for more research but did not identify a quantifiable risk, supporting the no_reliable_estimate classification.
- Indépendance
- WHO review synthesised global evidence independently of the Marfella, Leslie, and EFSA analyses cited below.
-
[2] New England Journal of Medicine / Marfella et al. — Microplastics and Nanoplastics in Atheromas and Cardiovascular Events Vérifié
Microplastics and Nanoplastics in Atheromas and Cardiovascular Events- Statistique
Patients with detectable MNPs in carotid plaque had HR 4.53 (95% CI 2.00-10.27) for composite cardiovascular endpoint vs those without- Extrait
“"Patients in whom microplastics and nanoplastics were detected within the atheroma had a higher risk of a composite of myocardial infarction, stroke, or death from any cause than those in whom microplastics and nanoplastics were not detected." ”
- Données source de
- 2024-03-07
- Consulté le
- 2026-04-18 · copie archivée
- Vérification
- Extrait récupéré et confirmé de manière indépendante, mot pour mot, par rapport à la source citée lors de notre audit de vérification.
- Calcul
- Marfella et al. prospectively followed 257 patients who underwent carotid endarterectomy and analyzed excised plaque for MNPs via pyrolysis-GC/MS. Polyethylene was detected in 58.4% of plaques (150 of 257); PVC in 12.1% (31 patients). After 34 months, the MNP-positive group had significantly higher cardiovascular events. However, this is an observational study in a high-risk surgical cohort (all had asymptomatic carotid disease), and critics note no pre-analytical anticontamination procedures were used — the surgical environment itself contains PE and PVC. The study demonstrates association in a diseased population, not causation in the general public, and cannot be converted to a population-level attributable risk.
- Indépendance
- Prospective clinical cohort study by Italian researchers using pyrolysis-GC/MS on excised plaque tissue; independent of WHO review and Leslie blood-detection study.
-
[3] Environment International / Leslie et al. — Discovery and quantification of plastic particle pollution in human blood
Discovery and quantification of plastic particle pollution in human blood- Statistique
Microplastics detected in 17 of 22 (77%) healthy adult blood samples at mean concentration of 1.6 µg/mL- Extrait
“"This is the first study to report quantitative data on plastic particle pollution in human blood, showing that plastic particles are bioavailable for uptake into the human bloodstream." ”
- Données source de
- 2022-03-24
- Consulté le
- 2026-04-18 · copie archivée
- Calcul
- Leslie et al. analyzed blood from 22 healthy Dutch volunteers using pyrolysis-GC/MS for five high-production polymers. PET was the most prevalent (50% of samples), followed by polystyrene (36%) and PE (23%). The study proves systemic exposure (detection) but does not measure health outcomes. No dose-response relationship, no disease endpoint, no follow-up. Detection of a substance in blood is a necessary but not sufficient condition for harm — the dose, persistence, and biological activity all remain unquantified for general-population microplastic exposure. This study is included because it is the most-cited driver of public fear, illustrating the detection-equals-danger heuristic.
- Indépendance
- Dutch laboratory study using independent pyrolysis-GC/MS methodology on volunteer blood samples; no overlap with the Marfella clinical cohort or WHO/EFSA reviews.
-
[4] Nature Medicine / Nihart et al. — Bioaccumulation of microplastics in decedent human brains
Bioaccumulation of microplastics in decedent human brains- Statistique
MNP concentration 7-30x higher in brain (median ~4900 µg/g) than liver/kidney; rose ~50% from 2016 to 2024; up to 10x higher in brains with documented dementia- Extrait
“"Concentrations in decedent brain samples were 7 to 30 times higher than in the liver or kidney... brain tissue from people who had been diagnosed with dementia had up to 10 times as much plastic in their brains as everyone else... the study design cannot show whether higher levels of plastic in the brain caused the dementia symptoms — they may simply accumulate more due to the disease process itself." ”
- Données source de
- 2025-02-03
- Consulté le
- 2026-06-14 · copie archivée
- Calcul
- Nihart et al. applied pyrolysis-GC/MS (corroborated by FTIR and electron microscopy) to >50 autopsy samples from the Albuquerque medical examiner, from decedents in 2016 and 2024. Brain MNP load (predominantly polyethylene, largely nanoscale shard-like fragments) far exceeded liver and kidney, and rose ~50% across the eight-year window. Brains with a documented dementia diagnosis showed up to 10x more plastic, concentrated in cerebrovascular walls and immune cells. This is a cross-sectional autopsy series: the dementia association is explicitly correlation, not causation, and the authors flag reverse causation (impaired clearance / blood-brain-barrier breakdown in dementia could increase deposition). It quantifies tissue burden and its temporal rise, not attributable disease risk, so it does not change the no_reliable_estimate classification.
- Indépendance
- US autopsy-tissue study using pyrolysis-GC/MS plus FTIR and electron microscopy; methodologically independent of the WHO review, the Marfella plaque cohort, the Leslie blood study, and the EFSA assessment.
-
[5] European Food Safety Authority — Microplastics and nanoplastics in food
Microplastics and nanoplastics in food- Statistique
EFSA risk assessment on microplastics in food scheduled for completion by end of 2027; no quantified health risk established as of 2025- Extrait
“"The European Parliament has requested EFSA to deliver a scientific opinion on the potential health risks posed by microplastics in food, water and air." ”
- Données source de
- 2025-10-28
- Consulté le
- 2026-04-18 · copie archivée
- Calcul
- EFSA's 2025 literature review of micro- and nanoplastic release from food contact materials analyzed over 100 studies (2015-2025) and concluded that while microplastic release from FCMs is real, current studies likely overestimate quantities, and nanoplastic data remain insufficient for reliable exposure estimates. EFSA has not issued a quantified risk figure. Their formal scientific opinion is not expected until end of 2027. The absence of an EFSA opinion after years of review is itself evidence of the epistemic gap — not of safety, but of insufficient data to quantify risk in either direction.
- Indépendance
- EFSA risk assessment is an independent EU regulatory review, methodologically separate from the WHO 2019 review and from individual primary studies.
-
[6] Journal of Hazardous Materials Letters / Pletz (2022) — Ingested Microplastics: Do Humans Eat One Credit Card per Week?
Ingested Microplastics: Do Humans Eat One Credit Card per Week?- Statistique
The widely cited 5g/week estimate overestimates ingestion by several orders of magnitude; realistic estimates are 0.01-0.1g/week- Extrait
“"Our analysis suggests that the widely reported figure of 5 g per week overestimates microplastic ingestion by several orders of magnitude." ”
- Données source de
- 2022-12-01
- Consulté le
- 2026-04-18 · copie archivée
- Calcul
- Pletz (2022) critically reviewed the Senathirajah et al. (2021) analysis that generated the "credit card per week" claim. The 5g figure was the extreme upper bound of a 0.1-5g range, and the underlying methodology conflated particle counts with mass in ways that inflate estimates. More careful analyses suggest typical adult ingestion is roughly 0.01-0.1g per week — still nonzero, but 50-500x lower than the headline figure. This matters because the perceived threat level is anchored to an exposure estimate that is almost certainly wrong by orders of magnitude.
- Indépendance
- Critical reanalysis of the WWF/Newcastle exposure estimate by independent food-science researchers; not affiliated with WHO, EFSA, or the original 2019 study authors.
-
[7] Environment International / Ragusa et al. — Plasticenta: First evidence of microplastics in human placenta
Plasticenta: First evidence of microplastics in human placenta- Statistique
12 microplastic fragments (5-10 micrometers) detected via Raman microspectroscopy in 4 of 6 human placentas from uneventful pregnancies- Extrait
“"In total, 12 microplastic fragments (ranging from 5 to 10 [micrometers] in size)...were found in 4 placentas." ”
- Données source de
- 2021-01-01
- Consulté le
- 2026-07-03 · copie archivée
- Calcul
- Substantiates the body prose's "Ragusa et al. found them in placentas" claim. First reported detection of microplastics in human placental tissue (fetal side, maternal side, and chorioamniotic membranes) in a small sample of six placentas from uneventful pregnancies. Like the Leslie blood study, this proves detection/bioavailability, not attributable health harm -- no outcome data, no dose-response, no disease endpoint measured. Does not change the no_reliable_estimate classification.
- Indépendance
- Italian clinical/laboratory study using Raman microspectroscopy on placental tissue; methodologically independent of the WHO review, the Marfella plaque cohort, the Leslie blood study, the Nihart brain study, and the EFSA assessment.