Return to ASDWA's Newsroom

Perchlorate Health Effects Model Review

This week a group of highly regarded peer reviewers met to evaluate the newly developed Biologically Based Dose-Response (BBDR) model that EPA plans to use to determine the appropriate MCLG for perchlorate.  The new model was developed at the recommendation of EPA’s Science Advisory Board as an improvement in how EPA was determining the MCLG for perchlorate, so that sensitive life stages were adequately accounted for.  The panel of peer reviewers for the model was assembled after a lengthy search that included nominations from the public.  In addition to this review of the model, EPA will convene another peer review panel later to consider the approach EPA will take to apply the model to the MCLG determination.
EPA worked with Food and Drug Administration (FDA) experts to develop the model which is based on data from several studies of perchlorate health effects and previously developed models.  The resulting model evaluates the effects of perchlorate hormones on lactating mothers, breast-fed infants and bottle fed infants under different iodide intake levels.  Reviewers were asked to specifically look at factors like the model structure, model calibration and sensitivity (especially at lower dose levels), hypothyroxinemic reference levels and model predictions, and characterization of model predictions, among others.   Reviewers recognized the complexity of the model and the challenges with developing a model that could give a definitive answer on the health impact.  However, the reviewers did have concerns about some aspects of the modeling process.  There were questions about how literature was selected to calibrate the model, whether enough quality data was found, especially when much new data has been generated but was not used in model calibration.   There was concern that the perchlorate doses where the model is the most accurate, don’t compare well to the levels found in drinking water.  The reviewers also questioned whether the model results should be extrapolated to three very different sensitive subpopulations.   Finally, the reviewers thought that a clearer explanation in the report on many of these issues was needed.  This could help alleviate some of the concerns.
The peer review team will continue to develop their report on the model and present the full findings to EPA.  EPA does not need to have a consensus opinion from the panel but just a report of their work, which could include majority and dissenting opinions.  EPA will take all this into account in refining the final model that will be used to support the MCLG determination.
For more details on the perchlorate modeling process and this peer review, see EPA’s perchlorate web page.