In Europe, robust habits and practices data exist in the adult population for the common cosmetic products (including personal care products) covering the majority of daily exposure (Hall et al, 2007, Hall et al, 2011), which are published, together with other estimated values, in the SCCS’s Notes for Guidance for the Testing of Cosmetic Ingredients and their Safety Evaluation (SCCS, 2012). This document also details the exposure scenarios and equations for calculating the daily exposure to chemicals. Other key published data sources include Loretz et al, (2005, 2006, 2008) providing frequency and amount data in the US adult population, and Biesterbos et al, (2013) providing the same in a Dutch population.
For aggregate exposure estimates, while in general there is agreement that tiered approach should be used (Delmaar JE and van Engelen JGM, 2006; Meek et al, 2011), which begins with a rough deterministic estimation of exposure and evolves to a more complex person-orientated probabilistic approach, to date there is no agreed methodology on the way to approach this. For cosmetic ingredients, aggregate exposure is estimated using a simplistic approach of adding deterministic exposures from all the individual product types in which the chemical might be present (SCCS, 2012). This is referred to as a low tier (tier 1), and provides a rough and very conservative estimation, since it assumes that everybody in the population uses all the products containing the chemical (at the maximum allowed concentration) every day. This has been regularly used in the past for cosmetic ingredient risk assessments for preservatives (SCCS, 2012). When these aggregate exposure estimates are used to calculate acceptable “safe levels” and/or conduct quantitative human risk assessments they result in overly conservative risk assessments. Thus, there is a need to use methods that are capable of producing refined, realistic aggregate exposure estimations (high tier estimates).
The landscaping exercise revealed two high tier aggregate models for estimating external exposure via different routes. The Research Institute of Fragrance Materials (RIFM) have partnered with Creme Global to build a model called the “Creme RIFM Model” (also available as Creme Care & Cosmetics) that allows estimation of aggregate exposure to fragrance ingredients in cosmetic products in European and US consumers (Comiskey et al, 2015; Safford et al, 2015), and the National Institute for Public Health and the Environment in the Netherlands (RIVM) have developed a higher tier Probabilistic Aggregate Consumer Exposure Model (PACEM) for ingredients, which contains exposure information for a population in the Netherlands (Dudzina et al, 2015; Manová et al, 2015; Nijkamp et al, 2015).
Aggregate exposure assessment for cosmetic products requires information on the chemical concentration used across product categories. The availability of such data in the public domain is limited as information on specific inclusion values is typically proprietary. Where chemical concentrations are quoted for example, the EPA Chemical/Product Categories Database (CPCat), then ranges are often quoted or maximum inclusion values and data are often only available for a limited number of chemicals in a limited number of products. Alternatively, if the chemical is restricted under regulation, the maximum regulated amount could be assumed as a conservative worse case assumption.
A major challenge in estimating aggregate exposure is obtaining information on how a chemical is used. Understanding how commonly a chemical is used in a cosmetic product, which is known as the chemical occurrence (presence probability) can lead to a significant refinement in the estimated aggregate exposure. Some data can be obtained from market survey databases such as Mintel and Codecheck, which relies on the mandated ingredient labelling in Europe, although, some systems are not well maintained leading to the availability of out of date information.
A key enabler in the development of realistic aggregate exposure estimates is the incorporation of co-use data. Such data are not available for household products and therefore conservative assumptions are employed such as, all product types containing the chemical are used by the consumer simultaneously.