UCSF

The U.S. Environmental Protection Agency (U.S. EPA) generally uses reference doses (RfDs) or reference concentrations (RfCs) to assess risks from exposure to toxic substances for noncancer health end points. RfDs and RfCs are supposed to represent lifetime inhalation or ingestion exposure with minimal appreciable risk, but they do not include information about the estimated risk from exposures equal to the RfD/RfC. We used results from benchmark dose modeling approaches recently adopted for use in developing RfDs/RfCs to estimate the risk levels associated with exposures at the RfD/RfC. We searched the U.S. EPA Integrated Risk Information System (IRIS) database and identified 11 chemicals with oral RfDs and 12 chemicals with inhalation RfCs that used benchmark dose modeling. For assessments with sufficient model information, we found that 16 of 21 (76%) of the dose-response models were linear or supralinear. We estimated the risk from exposures at the established RfDs and RfCs for these chemicals using a linear dose-response curve to characterize risk below the observed data. Risk estimates ranged from 1 in 10,000 to 5 in 1,000 for exposures at the RfDs, and from 1 in 10,000 to 3 in 1,000 for exposures at the RfCs. Risk estimates for exposures at the RfD/RfC values derived from sublinear dose-response curves ranged from 3 in 1,000,000,000 to 8 in 10,000. Twenty-four percent of reference values corresponded to estimated risk levels greater than 1 in 1,000; 10 of 14 assessments had points of departure greater than the no-observed-adverse-effect levels. For policy development regarding management of cancer risks, the U.S. EPA often uses 1 in 1,000,000 as a de minimis risk level. Although noncancer outcomes may in some instances be reversible and considered less severe than cancer, our findings call into question the assumption that established RfD and RfC values represent negligibly small risk levels.