High-throughput estimates for mouthing exposure and related risk from chemicals in children’s products

Abstract

There are public concerns about the possibility of children’s products (e.g., pacifiers, toys) releasing during mouthing toxic chemical substances, which can pose serious risks to children’s health. However, in various assessment frameworks and tools, mouthing exposure is often missing, mainly due to the lack of chemical migration rates data from products into saliva. Measured migration rates are available only for few chemical groups (e.g. phthalates, brominated flame-retardants) and materials. Since conducting migration experiments is costly and time-consuming, mathematical estimation methods are needed to estimate migration rates for the thousands of marketed chemical-product combinations. We thus aim to develop a high-throughput mouthing exposure model to estimate chemical migration rates into saliva from different materials and quantify subsequent mouthing exposure and related risk. We collected data of chemical migration into saliva for substances found in different children’s products, and built a harmonized dataset. We then analyzed and identified the main properties influencing migration rates: chemical concentration in the products, chemical diffusion coefficients within materials and material-saliva partitioning coefficients. Adapting a food migration model to the specific case of saliva based on the identified properties, we developed a model able to estimate chemical-material specific migration rates and tested it against the collected experimental results. For different mouthing scenarios, we finally determine exposure doses expressed in μg/kgBW/d and Hazard Quotients (HQ) for each chemical-material combination. We collected 398 data points of measured migration rates covering 62 different chemical-material combinations at different chemical concentrations. Migration rates range widely from a median of 0.0034 μg/10cm2-min to a maximum of 32.7 μg/10cm2-min, with phthalate and other plasticizers in PVC showing the highest migration rates. Resulting mouthing exposure doses with experimental migration rates vary from a median of 0.005 μg/kgBW/d to 50.6 μg/kgBW/d for an average mouthing duration of 3.4 min/h. Combining exposure estimates with toxicity information, we derive widely ranging HQs of up to 15 across chemical-material combinations and average and worst-case mouthing scenarios. The developed model for estimating migration rates shows good fitting of the experimental dataset (adjusted R2=0.88). Our results demonstrate that we are able to systematically estimate mouthing exposure from chemicals in different chemical-material combinations and rank possible alternatives based on common chemical function. In addition, the results show that mouthing exposure is limited for most chemical-material combinations, but still a relevant exposure pathway that need to be assessed for children risks associated with plasticizers in PVC and high exposure durations.