Phenol, 4-nonyl-, branched

Endpoint summary

Description of key information

A weight of evidence (WoE) approach has been taken to the determination of the potential for nonylphenol to bioaccumulate in aquatic organisms and is described in a separate technical report in Section 13 of the CSA.

For the purposes of the risk assessment, the BCF from the Ekelund et al (1990) study with the fish Gasterosteus aculeatus is used. Based on whole body tissue wet weight the BCF is 1200 for Test 1 and 1300 for Test 2. The results have been lipid normalised based on the lipid values provided in the report, with calculated BCFs of 833 to 896. The highest value of 896 and most conservative value is used in the risk assessment. Results of this study show the BCF of valid fish bioconcentration studies are lower than the REACH bioaccumulation criterion of < 2,000.

Additional information

A weight of evidence (WoE) approach has been taken to the determination of the potential for nonylphenol to bioaccumulate in aquatic organisms and is described in a separate technical report in Section 13 of the CSA.

In response to a request by ECHA an update has been provided for the bioaccumulation in water endpoint. To fulfil this update, various studies requested by ECHA (Diehl et al (2012) and Ekelund et al (1990)) have been added, a new literature search has been conducted on recent publications from 2015 to 2017, and additional consideration provided to data on trophic magnification. As also requested by ECHA a weight of evidence (WoE) approach was used. This was considered an appropriate method for assessing nonylphenol bioaccumulation for several reasons including: (i) a data set covering a range of organisms and trophic levels; (ii) study reports that, superficially at least, indicated different conclusions; (iii) many studies have been carried out using non-standard test methods and are considered of lower reliability (Klimisch 2 or 3) but still provide valuable information; (iv) existence of studies such as field studies, in vitro methods and non-testing data all of which may not be support derivation of exposure limit values for use in a risk assessment under REACH Regulation, but are suitable for a WoE approach.

A standard approach was taken to bring clarity, consistency and transparency to the WoE, based on a framework developed by Suter (2017). This included three steps; step one was collation of the evidence, step two was weighting of the evidence and step three was evaluation of the evidence. Collation of the evidence included studies as per the 2010 submission and 2015 update, in addition to evidence found through a literature search carried out in 2017. A standardised set of criteria were developed to weight the evidence, incorporating three parameters; reliability, relevance/adequacy and strength. Each of these three criteria were provided equal weight. The data was evaluated according to the following defined data groups, with greater relevance given to the first group and lesser relevance to each subsequent group:

1. Reliable laboratory in vivo fish data are considered the most relevant;
2. Other in vivo testing, including that on invertebrates and field data, should be considered the second most relevant;
3. In vitro data are third; and
4. Non-testing data, such as QSARs, other modelling data and indicators such as physico-chemical properties are the least relevant.

Evaluation comprised considering how the data fit with three defined hypotheses identifying low, moderate and high bioaccumulation potential respectively. A large number of confounding factors were identified during the evaluation which complicate the comparability of the data. The full list is provided in the main report, however particularly important ones included:

• variation between the methodologies used by different authors as few standard guideline studies were available for direct comparison;
• a lack of accepted criteria for evaluation of the bioaccumulation data, other than the criteria for Bioaccumulative substances (BCF >2000L/kg) and Very Bioaccumatlive criteria (BCF >5000L/kg) as per REACH Annex XIII (ECHA, 2006); 
• inconsistencies in the expression of the data, which make comparison and interpretation of the data challenging such as:
• variable reporting of studies in terms of wet weight or dry weight tissues for expression of BCFs, leading to lack of consistency, lack of qualification for the basis of BCF values (especially when reported as dry weights) and, often, mis-representation of bioaccumulation in standard terms when they relate to dry-weight results.
• variable reporting of studies in terms of wet weight or dry weight tissues for expression of BCFs, leading to lack of consistency, lack of qualification for the basis of BCF values (especially when reported as dry weights) and, often, mis-representation of bioaccumulation in standard terms when they relate to dry-weight results.
• variation in the calculation method for BSAF;
• the use of steady state versus kinetic BCF calculations;
• the use of particular organ only concentrations for derivation of bioaccumulation/magnification factors; and
• the expression of BCFs based on total radioactivity measurements in numerous studies, which do not account for the metabolism of nonylphenol.

For several pieces of evidence, these confounding factors lead to significantly over-inflated BCFs (which could incorrectly be categorised as in support of hypotheses 2 and 3, moderate and high strength bioaccumulation). In many cases dry tissue weights, rather than wet tissues weights were used in the calculations but this qualification was generally discarded from the final evaluation or presentation of the findings. Data without significant confounding factors was compared with the Annex XIII criteria and associated guidance to conclude of the potential for bioaccumulation of nonylphenol.

The clear majority of data, particularly after consideration of the confounding factors, pointed towards hypotheses one that nonylphenol is not bioaccumulative. Based on the weighing of all the evidence, it is determined that nonylphenol does not fulfil the bioaccumulation criterion (> 2,000) as defined in the PBT criteria in the REACH Regulation and ECHA guidance, in terms of bioaccumulation for fish. To be conservative, a worst case BCF value of 896 for fish (5% lipid normalised), based on Ekelund et al (1990) for stickleback (Gasterosteus aculeatus) is preferred for use in the risk assessment.

The potential for accumulation into other organisms, such as invertebrates, birds and mammals varies and may represent a broader range of concentrations factors than for fish species. The weight of this evidence does not generally indicate bioaccumulation or biomagnification for these species. It might be postulated that some very worst-case scenarios may occur for filter feeders, such as bivalves under certain unknown circumstances. 

For the purposes of the risk assessment, the BCF from the Ekelund et al (1990) study with the fish Gasterosteus aculeatus is used. Based on whole body tissue wet weight the BCF is 1200 for Test 1 and 1300 for Test 2. The results have been lipid normalised based on the lipid values provided in the report, with calculated BCFs of 833 to 896. The highest value of 896 and most conservative value is used in the risk assessment. Results of this study show the BCF of valid fish bioconcentration studies are lower than the REACH bioaccumulation criterion of < 2,000.

As no experimental data for the bioaccumulation in terrestrial species is available, the BCF for earthworms is calculated using measured Log K_OW(see log K_OWvalue of 5.4, Section 4.7; Partition Coefficient). The BCF, calculated by using the ECHA Equation R.16-76 (BCF_earthworm= (0.84+0.012Kow) / RHO_earthworm; with RHO_earthworm= 1 (kg wwt/L) by default), is 3015.

This indicates that nonylphenol has the potential to accumulate in terrestrial organisms.

According to Annex XIII of Regulation (EC) No 1907/2006 and to the Guidance on information requirements and chemical safety assessment Chapter R.11 (PBT Assessment, ECHA (2014)), a substance does not fulfil the criteria “bioaccumulative (B)” or “very bioaccumulative (vB)” if the bioconcentration factor (BCF) is below 2000 or 5000 respectively or the log K_OW is below 4.5. The BCF of nonylphenol was determined to be 896 for fish (5% lipid normalised), based on Ekelund et al (1990) for Gasterosteus aculeatus.  It is concluded that nonylphenol is not bioaccumulative (B) or very bioaccumulative (vB) within the PBT criteria.