Exposure Assessment

Tier 0 Assessment

Exposure Sources

Phenoxyethanol has several functions as an ingredient in consumer products, including as a preservative, antibacterial, solvent and fixative. Due to its multiple functions, it can be found in a number of consumer products, such as cosmetics and personal care products, pharmaceuticals and household care products.

In a project for the Consumer Product Safety Commission (CPSC), The American Chemistry Council Ethylene and Propylene Glycol Ethers Panel has tabulated uses of various ethylene glycol ethers in consumer products, suggesting common use in paints, coatings, dyes and cleaners (OECD SIDS, 2004), as shown in Table 13:

Table 13: Percentage of Phenoxyethanol Production used For Consumer Products. From OECD SIDS, 2004

Types of Consumer End Products	Consumer Products Vol         ( metric tons)	Consumer Products % Production	Consumer Products Approx. Weight	Percent Industrial/Consumer Use
Paint/coatings Cleaners Dyes	< 4.500 < 2.300 < 450	37.5% 19.0% 6.5%	37-15% 5-15% 5-15%	37/63

 

The highest human exposure to phenoxyethanol comes from nearfield exposure scenarios that have direct skin contact, and aggregate exposure will be greatest for products containing phenoxyethanol that are frequently used, particularly if used in combination with other phenoxyethanol-containing products. According to the Danish EPA, phenoxyethanol is the most commonly used preservative in cosmetic and personal care products, and a recent survey reported that phenoxyethanol was present in 40.7% of 629 products. In addition, the US Department of Health and Human Services Household Products Database, which is available online at http://householdproducts.nlm.nih.gov/cgi-bin/household/brands?tbl=chem&id=745, shows that phenoxyethanol is present in several household products were direct skin contact is expected, such as hand dishwashing liquids, general purpose cleaners and laundry liquids.

Exposure Pathways

Consumer use of phenoxyethanol may result in dermal and inhalation exposures. However, due to its low vapour pressure, inhalation exposure is much lower than dermal exposure. Inhalation exposure is also limited by its relatively rapid degradation in the atmosphere. Occasional systemic exposure may occur via vaccines containing phenoxyethanol as a preservative. Therefore, as inhalation is considered not to be the major route it is not selected for further study here. In reality, phenoxyethanol, will evaporate to some extent during the application of a product, reducing the amount remaining on the skin. Accounting for the evaporation could conceivably lead to a further reduction of the dermal exposure estimate, but with some additional inhalation exposure.

Tier 1 Assessment

The products selected are the daily use of cosmetics and personal care products and household care products, which have direct skin contact on a frequent basis.

Cosmetic and Personal Care Products

The tier 1 assessment was based on exposure data from the SCCS Notes of Guidance 8th Revision SCCS/1501/12 (SCCS, 2012), shown in Table 14, using the conservative assumption that all of the products contain phenoxyethanol at a maximum concentration of 1% in product (worse case), which is the maximum level permitted under EU regulation. Dermal penetration of 80% was assumed for the dermally exposed products and 90% oral bioavailability was assumed for products that might be consumed orally. The output of the tier 1 assessment is shown in Table 15. In order to estimate the aggregate exposure, the phenoxyethanol exposures from the individual products were summed.

Table 14: Estimated daily exposure levels for different cosmetic product types according to Colipa data [SCCNFP/0321/02; Hall et al 2007, 2011]

Product	External Product Exposure g/day	External Product Exposure mg/kg/day	Retention Factor	Calculated daily exposure (g/day)	Calculated relative daily exposure (mg/kg/day)
Shower Gel	18.67	279.20	0.01	0.19	2.79
Shampoo	10.46	150.49	0.01	0.11	1.51
Hair Conditioner*	3.92	-	0.01	0.04	0.67
Hair Styling	4,00	57.40		0.40	5.74
Liquid Foundation	0.51	7.90	1.0	0.51	7.90
Makeup Remover*	5.00	-	0.1	0.50	8.33
Hand Wash soap*	20.00	-	0.01	0.20	3.33
Body Lotion	7.82	123.20	1.0	7.82	123.20
Face Cream	1.54	24.14	1.0	1.54	24.14
Hand Cream	2.16	32.70	1.0	2.16	32.70
Deo non-spray	1.5	22.08	1.0	1.50	22.08
Eye makeup*	0.02	-	1.0	0.02	0.33
Mascara*	0.025	-	1.0	0.03	0.42
lipstick	0.057	0.90	1.0	0.06	0.90
Eyeliner*	0.005	-	1.0	0.01	0.08
Toothpaste	2.75	43.29	0.05	0.14	2.16
Mouthwash	21.62	325.40	0.1	2.16	32.54

 

*Product types not covered by the Colipa studies: existing daily application amounts are divided by the mean

human body weight of 60 kg.

The assessment assumes chronic (average, [mg/kg bw/day]) aggregate external exposure, and the results are shown in the table below.

 

 

Table 15: Tier 1 exposure data to cosmetic products. External exposures are calculated from SCCS 2012

Product	External Product Exposure g/day	External Product Exposure mg/kg/day	Fraction of Phenoxyethanol in Product	External Phenoxyethanol Exposure (mg/kg/day)	Estimated Absorbed Dose of Phenoxyethanol Assuming 80% Dermal Penetration, or 90% Oral Bioavailability* (mg/kg/day)
Shower gel	0.19	2.79	0.01	0.03	0.02
Shampoo	0.11	1.51	0.01	0.02	0.01
Hair conditioner	0.04	0.67	0.01	0.01	0.01
Hair styling	0.40	5.74	0.01	0.06	0.05
Liquid foundation	0.51	7.90	0.01	0.08	0.06
makeup remover	0.50	8.33	0.01	0.08	0.07
Hand wash soap	0.20	3.33	0.01	0.03	0.03
Body lotion	7.82	123.20	0.01	1.23	0.99
Face cream	1.54	24.14	0.01	0.24	0.19
Hand cream	2.16	32.70	0.01	0.33	0.26
Deo non-spray	1.50	22.08	0.01	0.22	0.18
Eye makeup	0.02	0.33	0.01	0.00	0.00
Mascara	0.03	0.42	0.01	0.00	0.00
lipstick	0.06	0.90	0.01	0.01	0.01*
Eyeliner	0.01	0.08	0.01	0.00	0.00
Toothpaste	0.14	2.16	0.01	0.02	0.02*
Mouthwash	2.16	32.54	0.01	0.33	0.29*
Total	17.38	268.82		2.69	2.19

 

The key contributors to aggregate chronic exposure to phenoxyethanol in this tier 1 estimate are the leave-on (non-rinse) products including body lotion, hand cream, face cream and non-spray deodorant (Deo) and also mouthwash, which has a high product exposure due to the fact that the habits and practices data suggest that a high amount of product, 21.62g/day, is applied with each use and it is assumed that 10% is swallowed (SCCS, 2012).

Household Products

The tier 1 assessment for household products was conducted using the REACH Exposure Assessment Consumer Tool (REACT) developed by The International Association for Soaps, Detergents and Maintenance Products (AISE) to specifically address washing and cleaning product exposure assessment. The tool was used to calculate Phenoxyethanol exposure via dermal and oral routes separately. Both direct dermal (i.e. during product use) and indirect dermal and oral (i.e. skin contact with residues in laundry and ingestion of residues from contact with crockery and cutlery) exposure was calculated.

 

 

 

The algorithms within AISE REACT for calculating the external exposure are as follows:

__________________________________________________________________________________

Direct dermal exposure in mg/kg/day

Expsys = F1 x C x Tder x F2 x F3 x F4 x Sder x n / BW

Where, F1: ingredient fraction by weight, C: concentration in wash solution (mg/cm³), Tder: thickness of product in contact with skin (cm), F2: fraction transferred from solution to skin, F3: fraction remaining on skin, F4: fraction absorbed through skin, Sder: dermal surface area (cm²), n: product use frequency (tasks/day) and BW: body weight (kg).

Indirect dermal exposure from clothes in mg/kg/day

Expsys = F1 x (M x (F'/W) x FD x FL) x Sder x F2 x F3 x F4 / BW

Where, F1: Ingredient fraction by weight, M: amount of undiluted product used (grams), F: fraction remaining in final liquor before spinning, W: total fabric weight (grams), FD: fabric density (mg/cm²), Sder: dermal surface area (cm²), F2: fraction transferred from solution to skin, F3: fraction remaining on skin, F4: fraction absorbed through skin and BW: body weight (kg).

Oral exposure from residues in mg/kg/day

Exp_sys = F1 x C x Ta x SA / BW

Where, F1: Ingredient fraction by weight, C: concentration in product (mg/ml), Ta: amount of water left on dishes after rinsing (ml/cm²), SA: area of dishes in daily contact with food (cm²) and BW: body weight (kg).

__________________________________________________________________________________

The REACT tool was run using the defaults in the tool and assuming a maximum use concentration of 15% phenoxyethanol (OECD SIDS, 2004), a dermal penetration value of 80% and an oral bioavailability value of 90%. The default values in the REACT tool are largely based on habits and practices data developed by AISE within the Human and Environmental Risk Assessment (HERA) project. The approach assumes that consumers use all products simultaneously and that the substance is present in each product category. The results are shown in Table 16. As with the personal care product assessment, the aggregate exposure to Phenoxyethanol was estimated by summing the exposures of the individual products.

 

 

Table 16: Tier 1 assessment of consumer exposure to phenoxyethanol from household products

Product	Exposure Scenario	Fraction of Phenoxyethanol in Product	Absorbed Dose (mg/kg/day)
Laundry liquid	Direct skin contact hand wash laundry	0.15	0.583
	Direct skin contact pre-treatment laundry	0.15	8.580
	Indirect skin contact wearing clothes	0.15	0.002
Hand dish wash	Direct skin contact dishwashing	0.15	0.125
	Indirect oral exposure to residues	0.15	0.001
General Purpose Cleaner	Direct skin contact cleaning surfaces	0.15	0.377
Total			9.67

 

In this Tier 1 assessment the key contributor to the aggregate exposure to phenoxyethanol from household products is laundry liquid. The estimate could be refined through the application of specific consumer exposure determinants (SCEDs). The SCEDs were not specifically developed for the purposes of aggregate exposure but could reduce the conservatism related to consumer habits and practices (for example, amount used, frequency of use and skin surface area).

Tier 2 Assessment

For this example, no tool or data were readily identified that could be used to assess aggregate exposure at tier 2 for the household products. Therefore, the cosmetics, personal care products were assessed using Creme Care and Cosmetics and PACEM, but the household care assessment was not further refined, and remain a tier 1 assessment. Note however, work is underway for both models to be expanded to include household products at a higher tier, by integrating data at the individual subject level.

Cosmetic and Personal Care Products

In tier 2, exposure assessment is refined in two respects. Firstly, realistic data on use and co-use of cosmetics in the population is considered. The use of this information accounts for the fact that not all products are used by all consumers to the same extend. Secondly, more realistic information on the concentration of phenoxyethanol in the products was considered.

The assessment of exposure from a particular use of a single product itself was not further refined. As in the first tier assessment, product dermal retention factors are taken as suggested in the SCCS Notes of Guidance for the Testing of Cosmetic Ingredients and their Safety Evaluation (SCCS, 2012). In addition, the same dermal and oral absorption fractions as used in the tier 1 assessment are assumed (i.e. 80% dermal absorption and 90% oral absorption).

The same personal care and cosmetic product types were included as used in the Tier 1 assessment, to allow consistency in approaches, and to enable the refinement of exposure by incorporating chemical occurrence values to be measured.

The higher tier models considered in this refinement are population based models, which means that they assess the distribution of exposure in a population. For comparison with the first tier approach, the 95th percentile was chosen as representative of population exposure in both the Creme and PACEM analyses.

Phenoxyethanol Concentration

Two exposure inputs were varied. Firstly, considering the concentration of phenoxyethanol in the products, in the more conservative estimates a maximum concentration of 1% was assumed in all personal care and cosmetics products (worse case), and in the more realistic scenario Danish EPA data, where the phenoxyethanol concentration was measured in a number of marketed formulations, was assumed (Danish EPA, 2015), as shown in Table 17.

Table 17: Danish EPA data on the analytical results of phenoxyethanol measurement in cosmetic and personal care products.

Product	Concentration of Phenoxyethanol in Cosmetic Formulations (%)
Shower gel	1.00*
Shampoo	1.00*
Hair conditioner	1.00*
Hair styling	1.00*
Liquid foundation	0.69
makeup remover	0.80
Hand wash soap	0.50
Body lotion	0.85
Face cream	0.84
Hand cream	0.40
Deodorant non-spray	0.51
Eye makeup	0.89
Mascara	1.00*
lipstick	1.00*
Eyeliner	1.00*
Toothpaste	1.00*
Mouthwash	0.30

 

*For product types that weren’t included a concentration of 1% phenoxyethanol was assumed.

Secondly, the chemical occurrence of phenoxyethanol was considered, and in the more conservative case all products were assumed to contain phenoxyethanol, where as in the more realistic scenario chemical occurrence data obtained from the Mintel GNPD database was assumed.

 

Chemical Occurrence

Chemical occurrence data was used to infer the proportion of formulations on the market that contained phenoxyethanol. Mintel GNPD (http://portal.mintel.com/) is an online database that tracks consumer product launches across the globe. The Global New Products Database monitors product innovation and retail success in consumer packaged goods markets worldwide. More than 20,000 new products are added every month from 50 countries worldwide, ensuring GNPD users have access to comprehensive coverage, reliable data and robust reporting on products on the market.

The database is divided into categories such as Face/Neck Care, Body Care, Eye Care, and Lip Care. Using the site’s search function returns the total number of products in that category. Adding the ingredient name (phenoxyethanol) gives the subset of those products containing this ingredient. By determining the number of SKUs within a product category that contain phenoxyethanol relative to the total number of SKUs, the chemical occurrence can be derived, which in this model is assumed to be the likelihood that a product phenoxyethanol.

This simple method was used to derive the presence probabilities for phenoxyethanol in the cosmetic categories. The presence probabilities for phenoxyethanol in Europe (EU) and North America (NA) were calculated using the method described, and the highest value was assumed for each product and then these values were rounded up to the nearest 10 to remain conservative as there is no understanding for relative market share of those products identified which could have an impact on likelihood for consumer use. The rounded up presence probabilities that were used in the tier 2 assessment are shown in Table 18.

Table 18: A summary of chemical occurrence data obtained from Mintel GNPD for Europe and the US. Calculated percentages are rounded up to the nearest 10%.

Product	Mintel Chemical Occurrence (%)
Shower gel	20
Shampoo	20
Hair conditioner	40
Hair styling	30
Liquid foundation	50
makeup remover	40
Hand wash soap	20
Body lotion	50
Face cream	50
Hand cream	50
Deodorant non-spray	10
Eye makeup	40
Mascara	40
Lipstick	10
Eyeliner	20
Toothpaste	10
Mouthwash	20

 

Creme Care & Cosmetics Model

The Creme Care and Cosmetics model is described on page 32, in the triclosan case study section.

The Probabilistic Aggregate Consumer Exposure Model (PACEM)

PACEM is a probabilistic exposure model based on product usage data collected in a survey on cosmetics use in the Dutch population. The survey included 512 adults (210 male and 302 female). PACEM contains information on use frequency and amounts used for 32 PCPs. Information on gender, age and body weight is available and linked with the product usage information for each individual in the model’s database.

To run the model input is required on:

• the subpopulation to be considered (e.g. gender, age cohort to be considered);
• the products that are to be included in the assessment;
• the concentration of the substance in the product. These can be single values for each product or distributions of values to account for variability or uncertainty in the concentration data;
• the exposure per product use. This is to be expressed as an exposure fraction: the fraction of substance that is applied per use that the person is actually exposed to (e.g. is absorbed, inhaled, or is retained on the skin). Exposure fractions are specified for each product separately and can either be single values or distributions.

To assess aggregate exposure, PACEM simulates daily product use for a population based on the realistic product usage data. Next, this daily product use is combined with the product concentration data to assess daily total exposures for each individual in the model population. Chronic exposures are assessed by simulating exposures over multiple days and determining the daily average over this period.

PACEM can be used to simulate acute (single day) and chronic (daily average) exposure. Metrics included external exposure, absorbed dose and dermal load (for the risk assessment of sensitising materials).

Aggregate exposure assessment in the second Tier:

In the second tier refinements in the assessment are made in two respects. First, realistic use and co-use data of cosmetic products is taken into account. Second, more realistic information on product concentrations of phenoxyethanol, including presence probabilities are considered. The Creme Global and PACEM models, that are used in this second Tier include realistic product use information. Product concentration refinements are included in four different scenarios:

1. Phenoxyethanol always present in products at a concentration of 1%;
2. Phenoxyethanol always present in all products, concentrations assumed from the Danish EPA report (see concentrations in Table 7);
3. Assumed Mintel chemical occurrence rounded up to the nearest 10% (see chemical occurrence in Table 17) concentration is 1%;
4. Assumed Mintel chemical occurrence rounded up to the nearest 10%, concentrations assumed from the Danish EPA report (see concentrations in Table 17.

 

Assessment using PACEM

In the PACEM assessments the following inputs were used:

• The subpopulation of female, adult consumers was selected. the same list of products as used in the first tier was selected;
• Exposure amounts were calculated based on the retention factors given in (SCCS, 2012) and the assumptions on the absorption fractions of 80% for the dermal route and 90% of the oral route;
• Concentrations of the substance in the product varied among the four scenarios considered. For each scenario, a set of products was simulated based on the scenario assumptions. This was done as follows: in scenarios in which a chemical occurrence of less than 100% was considered, the product set was constructed by adding products with zero concentration in the proportion specified in the Mintel database. For the other (non-zero concentration) products, the concentration was set to the level assumed in the scenario. This assures that when products are sampled at random from this set the correct proportion of products considered in the daily product use contain phenoxyethanol.

Using these inputs, chronic absorbed doses from cosmetics use were modelled by simulating daily exposures for a period of 28 days and averaging over this period.

Assessment using Creme Care & Cosmetics:

• the subpopulation of female, adult consumers from the EU and the US were selected. The same list of products as used in the first tier was selected;
• exposure amounts were calculated based on the retention factors given in (SCCS, 2012) and the assumptions on the absorption fractions of 80% for the dermal route and 90% of the oral route;
• concentrations of the substance in the product varied among the four scenarios considered. For each scenario, the model iterates through each subject in the habits and practices survey and simulates a daily exposure resulting from the products selected for inclusion. By iterating through each day of product use for each subject, total daily exposure is calculated for each individual, to give their seven-day average or chronic exposure and their one-day maximum or acute exposure. For concentrations that are in the form of statistical distributions and/or with presence probabilities, these are simulated at each exposure event using random sampling to reflect the concentration and presence parameters inputted.

Using these inputs, chronic absorbed doses from cosmetics use were modelled by simulating daily exposures for a period of 7 days and averaging over this period.

Summary of Selected Results

Assessment in PACEM

The daily average absorbed dose of phenoxyethanol was assessed in each of the four refinement scenarios defined above. The 95th percentiles of the population for each scenario are shown in Table 19.

Table 19: Internal exposure estimates to phenoxyethanol generated in PACEM

Scenario	Statistic	Estimated Average Daily Absorbed Dose in Total Population of Adult Females (mg/kg/day)
Phenoxyethanol always present at 1%	P95	2.7
Phenoxyethanol always present at Danish EPA use concentrations	P95	2.3
Refined with phenoxyethanol chemical occurrence, when present 1% assumed	P95	1.5
Refined with phenoxyethanol chemical occurrence and Danish EPA use concentrations	P95	1.3

 

The results show that by progressively incorporating more realistic information on the product composition and chemical occurrence of phenoxyethanol, the aggregate exposure estimate is reduced. In scenario 1, which assumes that phenoxyethanol is present in every product used by the subjects, the mean internal aggregate exposure is 2.7 mg/kg/day, and by incorporating both the specific concentration data and the chemical occurrence data, exposure is refined. Indeed, in scenario 4 where both factors are considered in the model, the P95 internal exposure (absorbed dose) is reduced by a factor of two to 1.3 mg/kg/day.

Assessment in Creme Care & Cosmetics

The output of the Creme aggregate exposure model is the internal exposure to phenoxethanol shown in
Table 20.

Table 20: Internal exposure estimates to phenoxyethanol generated in Creme Care and Cosmetics

Scenario	Statistic	Estimated Average Daily Absorbed Dose in Total Population of Adult Females (mg/kg/day)
Phenoxyethanol always present at 1%	P95	1.14
Phenoxyethanol always present at Danish EPA use concentrations	P95	0.78
Refined with phenoxyethanol chemical occurrence, when present 1% assumed	P95	0.50
RCR Refined with phenoxyethanol chemical occurrence and Danish EPA use concentrations	P95	0.38

 

The results show a comparison of the P95 summary statistics for Creme scenarios 1 through 4.

From scenario 1, which assumes that phenoxyethanol is present in every product used by the subjects, the mean internal aggregate exposure is 1.14 mg/kg/day. When the exposure estimate is refined with the more specific product use concentrations for phenoxyethanol, the mean internal exposure is refined by about 30% to 0.78 mg/kg/day. Similarly, when chemical occurrence is incorporated, to take into account that not all cosmetic and personal care products on the market contain phenoxyethanol, the P95 internal exposure is refined by as 56% to 0.5 mg/kg/day compared to scenario 1. The greatest refinement is achieved by incorporating both the specific concentration data and the chemical occurrence data in scenario 4, where the P95 internal exposure is 0.38 mg/kg/day, which is a reduction of 67% as compared to scenario 1, and a reduction of 82% as compared to the deterministic tier 1 assessment.

Key summary statistics for the Creme scenarios are presented in Table 21 showing the distribution of exposure. In this case when the chemical occurrence data is included (in scenario 3 & 4) the P5 exposures become zero owing, to a portion of the population not using certain products in the aggregate scenario.

Table 21 Summary percentile statistics of aggregate exposure showing estimated average daily absorbed dose in total population (ug/kg/day)

	Product	P5	P25	P50	Mean	P75	P95
1	Phenoxyethanol always present at 1%	43.75 ± 2.04	120.54 ± 1.47	247.95 ± 2.11	377.93 ± 5.10	459.36 ± 5.52	1140.78 ± 26.06
2	Phenoxyethanol always present at Danish EPA use concentrations	20.10 ± 0.58	63.05 ± 0.86	152.04 ± 2.31	243.72 ± 2.40	300.99 ± 3.69	777.72 ± 10.62
3	Refined with phenoxyethanol chemical occurrence, when present 1% assumed	0.00 ± 0.00	2.18 ± 0.21	22.51 ± 1.04	116.06 ± 1.85	124.32 ± 3.16	499.25 ± 13.70
4	Refined with phenoxyethanol chemical occurrence and Danish EPA use concentrations	0.00 ± 0.00	1.36 ± 0.12	17.21 ± 0.75	91.23 ± 1.75	90.76 ± 4.75	378.05 ± 9.93

 

Discussion

In this exposure estimate for phenoxyethanol in cosmetic and personal care products, a tiered aggregate exposure assessment was conducted. At tier 1: simple deterministic addition of individual product exposures using the SCCS 2012 exposure values for the individual products in the EU population, and tier 2: a probabilistic person-orientated method incorporating product specific phenoxyethanol concentration and chemical occurrence data in the female population, using 2 models: Creme Care and Cosmetics, and PACEM. The female population was chosen at this more refined tier 2 in this instance to ensure that the exposure estimate was relevant for the population of cosmetic and personal care users, as it is known that females are higher users of these products. This study demonstrated the advantages of refining the exposure using subject-level probabilistic analysis, and specifically by incorporating chemical occurrence data. Using the deterministic simple addition with basic aggregation (tier 1) the P95 internal exposure is 2.19 mg/kg/day. By utilising the tier 2 approach the P95 internal exposure is refined to 1.38 in PACEM and 0.38 mg/kg/day in Creme Care and Cosmetics.

Comparing the outcomes of the higher tier aggregate exposure models with the first tier method for aggregate exposure should demonstrate that the incorporation of realistic information on use and co-use and the chemical occurrence of the substance lead to a reduction of the exposure estimate. Indeed, the higher tier Creme aggregate results at the P95 level range from about 40 – 114% of the highest screening level prediction for a single use (0.99 mg/kg/day, Table 6).

However, in the case of the PACEM assessment, the reduction is less than was observed in previous applications (Gosens et al, 2013, Dudzina et al, 2015) where the estimated exposure in proceeding from a low tier to a more realistic assessment was typically in the order of a factor 100. This could be due to the fact that this phenoxyethanol study looked at only a female population from the Netherlands, where exposure is mainly driven by the exposure to body lotion, (which accounts for about half of the total exposure estimated in the tier 1 assessment) and is not as much the result of a total aggregate assessment contributions from different sources of the aggregate exposure estimate.

In the case of PACEM, the first two refinement scenarios in the tier 2 assessment even lead to comparable estimate (even somewhat higher) as the SCCS approach used in Tier 1. There are several different reasons for this. The source of the exposure data in the two approaches is completely different; The SCCS data is based on European wide probabilistic exposure studies from the male and female population (Hall et al, 2007 & 2011), as compared to the PACEM input data which is from a survey in the Dutch population (Biesterbos et al, 2015).

In the case of the Creme Care and Cosmetics Tier 2 assessment, the exposure input data includes the same European data on amounts per use for the products, derived from Hall et al, (2007, 2011), but also included in the tier 2 assessment is the product co-use data at the subject level (is true for PACEM product use database), explaining why the exposure output is refined – through the use of a probabilistic model with which to assess consumer exposure that is based on real consumer habits and practices, as opposed to deterministically summing the contribution from each product category. All other assumptions regarding substance presence, concentration, product retention and penetration remain consistent.

In both the PACEM and Creme probabilistic exposure assessments the inclusion of product-specific concentration estimates and chemical occurrence data have considerable impact on refining the exposure estimate.

Household Products

To advance to a tier 2 estimation of aggregate exposure to phenoxyethanol from household products further data are required. The realism in the assessment can be increased by incorporating data on co-use patterns of product usage and the chemical occurrence of the ingredient. Some publicly available databases do exist that contain substance-specific information such as prevalence in products however this information is usually limited and important data required for the aggregation of exposure are missing for example, co-use data at an individual user level. Recently work completed under the CEFIC-LRI-B7 project has highlighted the paucity of co-use data for household products.