Conceptual Models of Early-Life MNP Exposure
Transitioning from worst-case defaults to realistic probabilistic parameters.
BACKGROUND
The AURORA project (Actionable eUropean ROadmap for early-life health Risk Assessment of micro- and nanoplastics) is a five-year Horizon 2020 initiative focused on understanding the impact of micro- and nanoplastics (MNPs) on human health during pregnancy and early life.
The primary aim of Task 5 (Work Package 5) is to deliver an actionable roadmap for risk assessment by integrating the scientific findings from the project's other research objectives, and findings from CUSP (European Research Cluster to Understand the Health Effects of Micro- and Nanoplastics), as well as evidence gathered from external research initiatives.
The specific goals of Task 5 include:
- Integration of Findings: Synthesizing data from exposure assessments (measuring MNPs in tissues like the placenta and cord blood) and toxicological studies (identifying health effects like oxidative stress or endocrine disruption).
- Developing a Risk Framework: Creating a specialized framework specifically tailored to the unique properties of micro- and nanoplastics, which often require different assessment methods than traditional chemicals.
- Identifying Knowledge Gaps: Determining what information is still missing to conduct a comprehensive evaluation of MNP impacts on early-life health.
- Supporting Policy and Regulation: Providing the evidence base and methodological workflows necessary for European regulators to manage the use of MNPs and protect vulnerable populations.
The current state of knowledge (2026) is that, while we have robust evidence for Presence and Internalization of MNPs, evidence regarding the long-term health effects of MNPs are still largely based on in vitro (cell-based) and in vivo (animal-based) models. Studies have used idealised MNPs, predominantly Polystyrene particles, which do not account for the wide range of polymer-based particles which have been detected in the environment, and to which humans may be potentially exposed.
What is known is that the ubiquitous presence of NMPs in environmental compartments makes human exposure unavoidable via oral and respiratory routes. Even if a comprehensive human health risk assessment is currently hampered by significant knowledge gaps regarding the hazards of MNPs.
Therefore, currently, an exposure-led Risk Assessment approach is seen as most appropriate, while taking account of existimng and proposed risk assessment approaches for MNPs.
The POLYRISK approach is a modular and flexible human health risk assessment framework that prioritizes the inhalation exposure route by structuring its analysis around a source-to-outcome (StO) continuum. It utilizes a deterministic, tiered testing strategy to group particles based on specific physical and chemical hypotheses, such as the fibre pathogenicity paradigm (FPP) for rigid polymeric fibers and the poorly soluble low toxicity (PSLT) particle concept for other respirable plastics. This framework integrates Adverse Outcome Pathways (AOPs) and Integrated Approaches to Testing and Assessment (IATA) to evaluate mechanistic endpoints like inflammation and oxidative stress, moving from basic physico-chemical characterization to specialized biological assays. While it currently relies on semi-quantitative screening and real-world exposure scenarios, such as urban traffic or indoor sports environments, its modular design is intended to be future-proof, allowing for the addition of quantitative dose-response models as more robust empirical data becomes available.
The Aurora approach aims to expand on this by adopting an exposure-led probabalistic/deterministic modelling approach. This is necessary due to the practical and ethical constraints of expoisure screening of the foetus and early life stages.
To demonstrate applicable approaches, four exposure scenarios have been defined at the outset.
- Pregnancy: Models exposure through pregnancy and the potential for maternal MNP transfer to the foetus via the placenta.
- Neonate: 0-6 Months, focussing on MNP bottle shedding and formula intake.
- Infant: 6-12 Months, focussing on on crawling behavior and dust resuspension.
- Toddler: 1-3 Years, focussing on hand-to-mouth activity and toy chewing.
These scenarios attempt to model parameters specific to each life stage. They do not represent the full spectrum of MNP exposure, but are designed to indicate and allow testing of various exposure metrics that can later be combined into a comprehensive life-stage assessment.
HAZARD CHARACTERISATION
To be expanded with details of MNP characterisation database.
A Note on Normalization Metrics (Body Weight)
A critical question in particle toxicology is the choice of normalization metric. Traditional chemical risk assessment universally uses Body Weight (BW) to derive a comparable dose (e.g., mg/kg/day). This is based on the principle that a chemical's effect relates to its concentration within the body's total volume.
However, this may be less relevant for particles like MNPs, where toxicity can be driven by:
- Surface Area Interaction: Harm may be caused by particles interacting with an organ's surface (e.g., the gut lining), making organ surface area a more relevant denominator.
- Particle Flux: The risk might relate to the absolute number of particles crossing a biological barrier (like the placenta), where the total daily particle load is the most critical metric.
To address this, our primary models (e.g., the Pregnancy Model) now include a "Normalization Metric" selector. This allows researchers to choose between the standard practice (Body Weight) and assessing the absolute total particle load, enhancing the framework's scientific flexibility and transparency.
Scenario 1: Maternal (Pregnancy)
Multi-pathway ingestion and inhalation exposure during pregnancy.
This section details the exposure routes for a pregnant mother, combining direct ingestion (water, diet) with the secondary ingestion of inhaled particles that are trapped and swallowed via mucociliary clearance. It serves as the basis for estimating the maternal burden that could potentially be transferred to the fetus.
Methodological Justification: Empirical Anchoring & Reverse Dosimetry
Summary of Approach: The 5-Step Reverse Dosimetry Sequence
- Daily Maternal Intake: Estimates the total daily particle load the mother ingests and inhales from environmental sources (water, diet, and dust) based on the input parameters.
- Systemic Translocation: Applies a highly restrictive Gut Translocation Fraction (
f_gut) to calculate the Systemic Dose—the tiny fraction of particles that actually cross the gut wall and enter the maternal bloodstream daily. - Placental Accumulation over 280 Days: Calculates how many particles become trapped inside the placental tissue over a standard 280-day gestation period (Systemic Dose × Placental Trapping Fraction × 280 days).
- Empirical Filter (Zhu et al. 2023): Evaluates the total placental burden. If a simulated scenario results in a biologically impossible accumulation (exceeding the clinical maximum of ~4,000 particles per term placenta), that scenario is physically filtered out.
- Final Foetal Dose: For the surviving, clinically-validated scenarios, a final Foetal Transfer Fraction (a Beta distribution heavily skewed toward zero) is applied to the Systemic Dose to determine the actual number of particles reaching the foetus per day.
1. Introduction
The identification of MPs in the placenta and the fetal body has been confirmed by the findings of numerous recent studies (Sharma et al., 2024). Furthermore, studies have confirmed the placental translocation of microplastics (MPs) and nanoplastics (NPs), a process highly dependent on physicochemical properties such as size, charge, and chemical modification as well as protein corona formation (Medley et al., 2023). However, mathematically modeling maternal-foetal MNP exposure presents unique kinetic challenges. This document outlines the scientific rationale for utilizing a Reverse Dosimetry (empirical anchoring) approach over traditional Forward Dosimetry in our probabilistic (Monte Carlo) risk assessment model, drawing upon current experimental and observational literature.
2. Mechanistic Limitations of Forward Dosimetry
Traditional risk assessment models rely on a "Forward Dosimetry" approach: estimating environmental intake, applying absorption fractions, and subsequently applying a static Placental Transfer Index (PTI). For MNPs, applying generalized kinetic transfer rates yields highly uncertain outputs due to several critical biophysical confounders highlighted in the current literature:
- Size-Dependent Transport: Experimental studies demonstrate strict size-dependent transport of polystyrene particles across the placenta (Medley et al., 2023). For example, Wick et al. (2010) observed that beads sized 50, 80, and 240 nm were able to cross the ex vivo placenta to the fetal compartment, while 500-nm beads did not. Similarly, Cartwright et al. (2012) observed in an in vitro model that 50-nm fluoresbrite polystyrene particles were transported to the fetal compartment at a sixfold higher rate than 100-nm particles.
- Surface Functionalization and Charge: Particle translocation cannot be predicted by size alone. Kloet et al. (2015) found that polystyrene particles of the exact same size had significantly different translocation properties that were likely dependent on specific chemical functional groups.
- Dynamic Protein Coronas: In biological fluids, proteins can rapidly cover the surface of nanomaterials forming a protein corona (Medley et al., 2023). Gruber et al. (2020) found that dynamic protein coronas heavily influenced the translocation of plain 80-nm polystyrene nanoparticles across the placental barrier, identifying albumin (HSA) and immunoglobulin G (IgG) as major proteins facilitating this transfer.
- Lack of Environmental Representativeness: The majority of experimental studies utilized uniform, spherical, polystyrene particles (Medley et al., 2023). Given the diversity of findings using highly controlled particles, generalized forward transfer rates cannot accurately represent the heterogenous mixtures that define true environmental exposures.
Due to these confounding variables, a universally applied PTI fails to accurately model realistic human exposure, necessitating an alternative mass-balance approach.
3. The Placenta as a Sink and the Reverse Dosimetry Framework
Experimental evidence demonstrates that the placenta acts not merely as a conduit, but as a sink for particulate matter. Zurub et al. (2024) described the physical accumulation of plastic and non-plastic particles directly inside the human placenta. Additionally, ex vivo perfusion studies by both Grafmueller et al. (2015) and Gruber et al. (2020) explicitly observed the accumulation of polystyrene NPs trapped within the syncytiotrophoblast layer.
To account for this, the model employs a Reverse Dosimetry framework. Instead of projecting highly uncertain systemic absorption and placental transfer rates forward, the model anchors its kinetic distribution parameters to observed, end-of-gestation clinical tissue burdens. The Monte Carlo engine is mathematically constrained to generate transfer and accumulation fractions that align with these empirical mass-balance realities.
4. Empirical Calibration Targets
To anchor the reverse dosimetry equations, quantitative, weight-normalized concentration metrics are required. The model's parameters are driven by recent clinical biomonitoring data:
- Quantitative Anchoring: Zhu et al. (2023) assessed the presence and type of particles in 17 placentas. All placenta samples included MPs, providing a critical average abundance metric of 2.70 ± 2.65 particles/g. While other landmark studies successfully detected MPs—such as Ragusa et al. (2021), who found 12 pigmented MPs sized 5 and 10 µm in the placentas of 4 women—the data reported by Zhu et al. (2023) provides the explicit mass-normalized concentration required to calibrate the mathematical accumulation limits of the Monte Carlo engine.
- Size-Binning Justification: The Monte Carlo engine stratifies particle kinetics based on distinct size bins. This is directly justified by Amereh et al. (2022), who evaluated plastic particles in 43 pregnant women's fresh human placentas and found that up to 64% of MPs from both IUGR and normal pregnancies were smaller than 10 µm. This clinical size dependence, combined with the experimental evidence of size-restrictive transport (Wick et al., 2010), mathematically restricts higher translocation efficiencies to the smallest particle distributions.
5. Multi-Study Integration for Pathway Parameterization
The model's broader exposure algorithms are corroborated by additional clinical findings:
- Exposure Frequency: Weingrill et al. (2023) analyzed temporal trends and showed that 100% of analyzed placentas in 2021 had MP particles. This justifies the model's structural assumption of continuous, daily maternal exposure kinetics rather than isolated events.
- Ingestion Drivers: The model's reliance on specific dietary loads is validated by Xue et al. (2024), who observed that the frequency of seafood consumption (r=0.781) and the consumption of bottled water (r=0.386) were positively correlated with MP levels in maternal amniotic fluid.
6. Conclusion
By shifting to Reverse Dosimetry, this risk assessment model ensures that predicted foetal exposures are tethered to empirical human tissue data rather than theoretical kinetic assumptions. This approach effectively accounts for the complex physicochemical barriers—such as protein coronas, surface charge, and size exclusion (Medley et al., 2023)—that currently make generalized forward predictive modeling unreliable. As larger epidemiological cohorts provide further normalized placental mass-concentration metrics, these empirical target distributions will be iteratively refined.
Forward Exposure Parameters
Water Concentration
C_water [particles/L]
Lognormal
Water Intake Rate
IR_water [L/day]
Triangular
Indoor Dust Concentration
C_dust [particles/m³]
Lognormal
Oral Deposition Fraction
DF_oral
Triangular
Reverse Dosimetry Kinetics
Target Placental Burden (Filter Limit)
Max: ~4000 particles
Empirical Anchor
Size Binning & Translocation (f_gut)
f_gut
Lognormal
Foetal Transfer Fraction
PTI_foetal
Beta(1, 20)
Scenario 2: Neonate (0–6 Months)
Dietary and inhalation parameters during early infancy.
This section details the primary exposure routes for neonates, focusing heavily on dietary intake via heated polypropylene (PP) bottles. The data provided corrects the common error of using initial, highly elevated shedding rates from brand-new bottles by supplying the stabilized shedding rate over a bottle's lifespan.
Model Calculations & Justification
The Neonate Exposure Model evaluates acute exposure risk for newborns, prioritizing the high-concentration formula feeding pathway while accounting for resting inhalation.
- Formula Feeding Pathway (Bottle_Load): Models the massive absolute daily particle load released from heated polypropylene (PP) baby bottles during formula preparation. Size modifiers are applied exclusively to this high-concern pathway to adjust for the increased translocation potential of nanoscale fractions.
- Ambient Inhalation: Assumes a 24-hour resting exposure in the crib environment. Calculated using the ambient indoor dust concentration (
C_dust), resting breathing rate (IR_breath_rest), and the fraction of particles actually deposited in the respiratory tract (DF_inhale). - Normalization (BW): The absolute total particle load (particles/day) can be optionally normalized by the neonate's body weight distribution to yield the Estimated Daily Intake (EDI) in particles/kg-bw/day.
MNP Load from PP Bottles
Bottle_Load [particles/day]
Resting Breathing Rate
IR_breath_rest [m³/day]
Cbottle Shedding Dynamics Over Time
Visualizing why the initial use values (often used in worst-case deterministic models) drastically overrepresent chronic daily exposure compared to the stabilized lifespan mean.
How the model accounts for this: Experimental data shows that brand-new polypropylene (PP) bottles release a massive spike of particles during their first few sterilization and heating cycles (Day 1-5). However, this shedding rapidly decreases and stabilizes to a lower, consistent baseline over the bottle's lifespan (Day 21+). Traditional deterministic models often erroneously apply the "Day 1" maximum shedding rate to the entire 6-month neonatal period, vastly overestimating the total particle burden. To provide a realistic risk assessment, this framework utilizes the stabilized lifespan mean (represented by the blue curve) to accurately model chronic daily exposure.
Scenario 3: Infant (6–12 Months)
Crawling, floor micro-environments, and early hand-to-mouth behaviors.
This section characterizes the unique exposure profile of the crawling infant. It provides parameters for the "Pig Pen effect" (localized particulate resuspension in the breathing zone) and accurate frequency metrics for indoor hand-to-mouth behaviors, avoiding the use of outdoor soil-pica extremes for standard indoor modeling.
Model Calculations & Justification
The Infant Exposure Model evaluates the specific risks associated with the crawling life stage, where close proximity to the floor and frequent hand-to-mouth contacts drive environmental exposure.
- Hand-to-Mouth Ingestion Pathway: Models the direct transfer of microplastics from the floor to the mouth. It multiplies the frequency of hand-to-mouth contacts (
FQ_htmin contacts/hr) by the active crawling duration (Hrs_active), the surface area of the hand that is mouthed (SA_hand), and the concentration of MNPs on the floor surface (C_floor). - Resuspension Inhalation (The "Pig Pen" Effect): Models the localized cloud of particles kicked up by the infant's crawling. It multiplies the resuspended dust concentration (
C_resus) by the active breathing rate (IR_breath), active crawling hours (Hrs_active), and the inhaled deposition fraction (DF_inhale). - Modifiers: A morphology and size weighting score is applied to the combined exposure load to account for the higher toxicity and translocation potential of smaller micro- and nanoplastics.
- Normalization (BW): The absolute total particle load (particles/day) can be optionally normalized by the infant's body weight distribution to yield the Estimated Daily Intake (EDI) in particles/kg-bw/day.
Indoor Dust Conc.
C_dust [particles/mg]
Resuspended Dust
C_dust_resusp [part/m³]
Hand-to-Mouth Freq.
FQ_htm [contacts/hr]
Surface Area Mouthed
SA_hand [cm²]
Hand Surface Area Proportions
Illustrating the subset of the hand actually subjected to mouthing (SA_hand) relative to total hand area, a frequent source of error in deterministic models.
How the model accounts for this: Traditional deterministic models often assume that 100% of a child's hand surface area is placed into the mouth during a contact event, which grossly overestimates soil and dust ingestion. Observational videography (Xue et al., 2007) demonstrates that only a small fraction of the hand—typically just the thumb or a few fingers (roughly 10-15% of the total surface area)—actually enters the mouth. This model strictly bounds the SA_hand parameter using a triangular distribution that reflects this empirical reality, preventing compounded conservatism.
Scenario 4: Toddler (1–3 Years)
Active mobility, increased dust ingestion, and direct toy chewing emissions.
This section provides critical parameters for active toddlers. It notably addresses the data gap in direct particle shedding from chewed toys (ER_toy) by providing the best available surrogate data derived from recent mechanical abrasion studies on silicone/PVC matrices.
Model Calculations & Justification
The Toddler Exposure Model captures the unique risks of increased mobility and oral exploratory behaviors, focusing on direct dust ingestion and the physical degradation of mouthed plastic toys.
- Dust Ingestion Pathway: Models the direct consumption of settled dust and soil resulting from active play and frequent hand-to-mouth contact. Calculated by multiplying the daily dust ingestion rate (
IR_dust) by the environmental MNP concentration in dust (C_dust). - Toy Mouthing Pathway: Evaluates the direct release of micro- and nanoplastics from plastic toys due to mechanical chewing and salivary interaction. It multiplies the particle emission rate (
ER_toy) by the daily duration of mouthing behaviors (Hrs_mouthing). Because direct human mastication shedding data is extremely limited, this model utilizes surrogate data extrapolated from targeted mechanical elastomeric abrasion studies. - Modifiers: A morphology and size weighting score is applied to the combined exposure load to account for the higher toxicity and translocation potential of smaller micro- and nanoplastics.
- Normalization (BW): The absolute total particle load (particles/day) can be optionally normalized by the toddler's body weight distribution to yield the Estimated Daily Intake (EDI) in particles/kg-bw/day.
Dust Ingestion Rate
IR_dust [mg/day]
Toy Mouthing Duration
Duration_mouthing [min/day]
Emission Rate: Chewed Plastic Toys Surrogate Data
ER_toy [particles/min]
Bibliography
A curated list of key references used to parameterize the models.
- Amereh, F., Amjadi, N., Mohseni-Bandpei, A., et al. (2022). Placental plastics in young women from general population correlate with reduced foetal growth in IUGR pregnancies. Environmental Pollution, 314, 120174.
- Cartwright, L., et al. (2012). In vitro placental model optimization for nanoparticle transport studies. International Journal of Nanomedicine, 7, 497-510.
- Cox, K. D. et al. (2019). Human Consumption of Microplastics. *Environmental Science & Technology*. [DOI]
- Danopoulos, E. et al. (2020). Microplastic Contamination of Drinking Water: A Systematic Review. *PLOS ONE*. [DOI]
- Grafmueller, S., et al. (2015). Bidirectional transfer study of polystyrene nanoparticles across the placental barrier in an ex vivo human placental perfusion model. Environmental Health Perspectives, 123(12), 1280-1286.
- Gruber, M. M., et al. (2020). Plasma proteins facilitates placental transfer of polystyrene particles. Journal of Nanobiotechnology, 18(1), 1-14.
- Kloet, S. K., et al. (2015). Translocation of positively and negatively charged polystyrene nanoparticles in an in vitro placental model. Toxicology in Vitro, 29(7), 1701-1710.
- Li, D. et al. (2020). Microplastic release from the degradation of polypropylene feeding bottles during infant formula preparation. *Nature Food*. [DOI]
- Medley, E. A., Spratlen, M. J., Yan, B., Herbstman, J. B., & Deyssenroth, M. A. (2023). A systematic review of the placental translocation of micro- and nanoplastics. Current Environmental Health Reports, 10, 99-111.
- Ragusa, A., Svelato, A., Santacroce, C., et al. (2021). Plasticenta: first evidence of microplastics in human placenta. Environment International, 146, 106274.
- Sharma, R. K., Kumari, U., & Kumar, S. (2024). Impact of microplastics on pregnancy and fetal development: A systematic review. Cureus, 16(5), e60712.
- US EPA (2011). Exposure Factors Handbook: 2011 Edition. *U.S. Environmental Protection Agency*. [Link]
- Vethaak, A. D. & Legler, J. (2021). Microplastics and human health. *Science*. [PDF]
- Weingrill, R. B., Lee, M. J., Benny, P., et al. (2023). Temporal trends in microplastic accumulation in placentas from pregnancies in Hawaii. Environment International, 180, 108220.
- Wick, P., et al. (2010). Barrier capacity of human placenta for nanosized materials. Environmental Health Perspectives, 118(3), 432-436.
- Xue, J. et al. (2007). A Videotape-Based Analysis of Children's Hand and Mouthing Activity. *Risk Analysis*. [DOI]
- Xue, J., Xu, Z., Hu, X., Lu, Y., Zhao, Y., & Zhang, H. (2024). Microplastics in maternal amniotic fluid and their associations with gestational age. Science of The Total Environment, 920, 171044.
- Zhang, J. et al. (2023). Microplastics in Indoor Dust: A Review on Characteristics, Abundance, and Human Exposure. *In: Latest Updates in Microplastic Research and Their Environmental Applications*. [Link]
- Zhu, L., Zhu, J., Zuo, R., Xu, Q., Qian, Y., & An, L. (2023). Identification of microplastics in human placenta using laser direct infrared spectroscopy. Science of The Total Environment, 856, 159060.
- Zurub, R., Bainbridge, S., Rahman, L., Halappanavar, S., & Wade, M. G. (2024). Particulate contamination of human placenta: plastic and non-plastic. BioRxiv.
Interactive Exposure Scenario
Simulate the effect of percentile shifts on estimated daily intake.
This interactive tool demonstrates the core value of your updated probabilistic parameters. Adjust the slider to see how shifting from the Central Tendency (Mean) to the 95th Percentile (simulating older, deterministic worst-case methods) exponentially inflates the calculated Total Daily Intake (TDI) across age groups.