Review Tox Data for Other Health Effects with Doses or Air Concentrations Similar to Those for Your Site

During this step, you will review studies for other target organs and systems with doses or air concentrations similar to those doses or air concentrations for your site. Compare site-specific doses or concentrations to the Levels of Significant Exposure (LSE) tables and figures presented in ATSDR’s Toxicological Profiles. The LSE tables and figures summarize the doses or air concentrations associated with health effects reported in scientific studies. Remember to also review the MRL worksheet for health effects from other sensitive target organs and systems. Quite often, the MRL worksheet will use the dose responses from other studies to generate BMDLs/BMCLs and HEDs/HECs. These toxicity values can be very useful in deciding other possible health effects.

The User’s Guide, which is in the appendix of each Toxicological Profile, provides sample LSE tables and figures and a legend to understand each component of the graphics. You can use the LSE tables and figures to quickly identify health effects associated with specific doses and concentrations for acute, intermediate, and chronic exposure periods. See a sample User’s Guide for reference in Appendix C of the Dinitrophenols Toxicological Profile.

The doses and concentrations in the LSE tables and figures cover these components:

• Health effects observed at different doses (mg/kg/day) or concentrations (ppm, mg/m³) and durations (i.e., acute, intermediate, and chronic)
• Differences in response by species
• The study used to derive duration-specific MRLs
• Cancer effect levels (CELs) from animal and human studies
• EPA’s estimated range for doses and concentrations associated with an upper-bound individual lifetime cancer risk of 1 in 10,000 to 1 in 1,000,000

All entries in these tables and figures represent studies that provide reliable, quantitative estimates of NOAELs, LOAELs, or CELs. When making decisions about health effects, always read the information provided in the MRL worksheet (Appendix A). The MRL worksheet provides the basis and describes the study used to derive the MRL and includes important information about other studies that support the MRL or document other harmful effects.

When a Toxicological Profile is available for a contaminant being evaluated, use the LSE tables and figures for a quick review of the health effects and to locate data for a specific exposure scenario. Comparing your site doses and air concentrations to those from the studies provided in the LSE tables and figures can provide perspective of where your site-specific doses and air concentrations fall in relation to those seen in published scientific studies. Toxicological reviews in EPA IRIS can also be a resource. If neither ATSDR nor EPA has this information available, work with a toxicologist to determine if other reliable sources have similar information published for your contaminant of interest. You may wish to conduct a literature search to see if toxicological information is available.

As previously discussed, health guidelines are frequently based on either the NOAEL or LOAEL from data reported in the literature, very often from a single study. In addition, there is benchmark dose modeling that uses the NOAELs and LOAELs to plot a dose or concentration response curve, which is used to derive a benchmark dose (BMD) or benchmark concentration (BMC) and the 95% lower confidence limit of the BMD (i.e., a BMDL) or BMC (i.e., a BMCL). When sufficient information is available for an animal species and from the study, dosimetric methods are used to derive a human equivalent dose (HED) or a human equivalent concentration (HEC). Thus, MRLs can be based on a NOAEL, LOAEL, BMDL/BMCL, or HED/HEC.

In addition to the critical study, other studies can provide health effects information associated with different doses and concentrations. Health assessors are not expected to exhaustively review the studies for contaminants of potential concern at a given site, but reviewing the larger toxicologic and epidemiologic database (i.e., the LSE tables and figures summarized in the Toxicological Profiles) may yield important supporting information for your documents.

In the in-depth analysis, health assessors are looking beyond single points on the dose-response curve to more fully understand the range of effects and effect levels observed in experimental studies. Both the shape and slope of the dose-response curve can help explain where site-specific exposures lie in the larger scheme of things. Use of benchmark dose (BMD) is an example of using more than one point to indicate the shape of the dose-response. If the data are sufficient, BMD analysis is the preferred method for derivation of MRLs. In consultation with a chemical profile manager, a health assessor could run a BMD analysis, if warranted, to verify the shape of the curve in the low dose range. This will often help provide the perspective community members seek, and it will help you decide which, if any, harmful effects might be possible. In some cases, consistent findings might be seen across studies. For other contaminants, findings might be more disparate. Consider these questions:

• Where does the NOAEL or LOAEL for the critical study fall in relation to other studies? Although the critical study will weigh most heavily in your analysis, it can be helpful to examine the similarity or disparity of doses and concentrations for other health endpoints from other studies. For example, you might find that many of the reported effect levels are in the same general range as the critical study, strengthening the evidence that in addition to the effects based on the critical study, other harmful effects might be possible at the doses or concentrations determined for the site. Remember, the critical study has been identified—after careful review of the scientific literature—as the best for developing protective health guidelines.
• If the health guideline is based on animal data, do any human data exist that shed more light on the issue? If extensive epidemiologic data are available for a contaminant, these data will likely have been reviewed and considered in the derivation of the health guideline for that contaminant, except for new studies that were not available at the time of health guideline development. However, available epidemiologic data may be able to verify the biologic relevance of the findings of animal studies. For example, an occupational study can show that exposure to a particular contaminant is associated with the same toxic endpoint seen in animal studies. This observed species concordance would provide greater weight to the available animal dose-response data used to evaluate human health effects. The exposure levels and associated outcomes, when available, can sometimes be used for comparison purposes with site exposures. Consider this example:
  • Community members are concerned about low levels (2 parts per billion [ppb]) of a particular contaminant in drinking water that they have been drinking for approximately 10 years. They believe leukemia rates are elevated. Two independent studies of community drinking water supplies with 100 ppb of the same contaminant revealed no elevated leukemia or any other cancers in populations exposed for 30 years. Health assessors can review the overall evidence of an association between the particular contaminant and cancer (based on NTP, IARC, EPA, and NIOSH cancer classifications), and examine the available epidemiological literature that appears relevant to the evaluation of the site population—but avoiding use of that comparison to draw a conclusion of likely or unlikely association of exposure with cancer incidence. When epidemiological data/studies are available, the health assessor should engage an ATSDR epidemiologist for assistance. Note that these studies are often very difficult to conduct and subject to many limitations. As such, the epidemiologist would be able to provide the health assessor with guidance on interpretation, recommendations, etc. Additionally, the epidemiologist might suggest that a cancer incidence analysis is warranted for specific cancer sub-types in the community.