Back to Top

Calculation of new human toxicity CFs by using ChemIDplus database

2 posts / 0 new
Last post
guvenhu
guvenhu's picture
Calculation of new human toxicity CFs by using ChemIDplus database

Dear USEtox team,

In my study, I want to calculate new human toxicity CFs. In some LCA-toxicity papers, ChemIDplus is suggested to develop especially non-carcinogenic ED50s and thus CFs. I have some questions regarding the use of this database.

1) USEtox manual says that ED50 values should be chronic-eq concentrations and if the found data is other than chronic data it should be converted to its chronic eq. On the other hand, for human concentrations, ChemIDplus generally has TDLo concentrations. According to the definition of TDLo given by ChemIDplus, it could be over any given period of time. Thus it could be either chronic or acute (or semichronic etc.) in my understanding. How should I convert these TDLo concentrations to ED50 values which should be put into USEtox?

2) In the same database, there are several oral concentrations for some chemicals. Should I take the minimum concentration or take an arithmetic mean of the whole values to calculate ED50ingestion values?

3) For some chemicals, there are no human concentrations but there are available LD50 animal concentrations. If I only apply "Interspecies conversion factors" given in USEtox manual to LD50s and convert them in the unit of kg/lifetime, is it ok? Or, should I also apply a "time of exposure" factor? I am not again certain that whether the LD50 concentrations in ChemIDplus are chronic or not.

Thanks in advance.

USEtox Team
USEtox Team's picture
Human toxicity effect factors

If TDLO (lowest dose causing a toxic effect) can be interpreted as LOEL (lowest observable effect level), then the usual extrapolation between LOEL and ED50 applies as described in the USEtox documentation, where also an extrapolation factor from sub-acute to chronic and from sub-chronic to chronic might apply.

When for a given exposure route (e.g. ingestion) there are >1 avaialble animal data points, they are all converted first to a human-equivalent ED50, and then the smalles values among those is chosen as per USEtox documentation, page 160. If there are human data available, these are prioritized. If more than one human data point is avaialble per exposure route, an arithmetic mean is suggested for obtaining an average or the lowest amongh those can be chosen to be more conservative.

Acute animal LD50 can be directly converted to human-equivalent chronic ED50 based on Figure 8 in Rosenbaum et al. (2011).