If the foundation behind what we consider "safe" is a risk calculation that produces a number expressed in ppm or ppb, how confident are we in the variables used the calculation? Let's look at the calculation used to derive the Maximum Contaminant Level Goal (MCLG) for drinking water.
The MCL is based on the concept that drinking water should contribute no more than 20% of the contaminant entering into the receptor. The Drinking Water Equivalent (DWEL) is a calculation that is based on the maximum amount of the contaminant being consumed through drinking water only.
So, if we assume that the human receptor will receive a dose of arsenic from other sources, what would the "safe" level of arsenic need to be for drinking water?
If arsenic comes from no other source other than drinking water, the DWEL calculates the "safe" level as:
- 0.01 mg/L
Not to complicate this, bu arsenic is a carcinogen so when we look at the risk of arsenic in drinking water, it is based on non-cancer DWEL (0.01 mg/L) and 1 in 10,000 cancer risk (0.002 mg/L) as follows:
Since these posts are about the reference dose (RfD), I am going to ignore the cancer risk for now.
The formula for the DWEL is:
- (0.0003 x 70)/2 = 0.0105 mg/L or 0.01 ppm or 10 ppb
The MCLG is therefore 20% of that amount, or 0.20 x 0.01 = 0.0021 mg/L. However (there is always a however) because arsenic is a carcinogen, the MCLG is set at "zero." The MCL is set at 0.01 which the EPA believes is a reasonable concentration that a water system can treat down to based on current technology and be considered "safe" Notice how the MCL is set at the DWEL.
For arsenic, we want to see zero but that's not feasible for all drink water suppliers, so an acceptable level of arsenic is deemed to be 0.01 ppm.
So back to the original dead horse statement/question:
If the foundation behind what we consider "safe" is a risk calculation that produces a number expressed in ppm or ppb, how confident are we in the variables used the calculation?
For a non-carcinogen, to accept the MCLG one must accept the DWEL. To accept the DWEL, one must accept the parameters and the values used. There is little controversy regarding the weight of an adult/child nor on the amount of water consumed, they are averages based on what we have found to be true in the population. So the only parameter in question, the only value that is paramount is the RfD.
If the "safe" level is to be derived then the RfD must be as bullet proof as possible. This, is where the fun begins. The RfD is:
An EPA estimate, with uncertainty or safety factors built in, of the daily lifetime dose of a substance that is unlikely to cause harm in humans. (1)
Let's look at perchlorates for example:
Perchlorate was discovered by the [EPA] in private drinking water wells in a portion of the Deep Creek area of western Spokane County in early 2006. Since that discovery, EPA has sampled additional wells and continues to find perchlorate at low levels across the study area. The perchlorate levels range from non-detected to 3.2 micrograms per liter (ug/l). These levels are not considered an immediate health risk.
For the EPA to make that claim, it had to be based on something...
There are no federal or Washington State drinking water standards for perchlorate. However, EPA uses a 24.5 ug/l perchlorate drinking water equivalent level (DWEL)1 as a preliminary remediation goal (PRG).
You can see why I beat this dead horse. If the clean up goal is based on the DWEL and the DWEL is based on an "EPA estimate, with uncertainty or safety factors built in, of the daily lifetime dose of a substance that is unlikely to cause harm in humans." that "estimate" better be solid.
For perchlorates, is it solid?
EPA completed its draft toxicological review of perchlorate in 2002, and proposed a reference dose (RfD) of 0.00003 milligrams per kilogram per day (mg/kg/day) based primarily on studies that identified neurodevelopmental deficits in rat pups.So 0.00003 mg/kg-day was derived based on a "maternal exposure to perchlorate" that saw "neurodevelopmental deficits in rat pups" The EPA then finds:
Subsequently, NAS [National Academy of Science] reviewed the health implications of perchlorate, and in 2005 proposed an alternative reference dose of 0.0007 mg/kg/day based primarily on the Greer et al, 2002 study.
Why the increase in what was deemed an acceptable dose (0.00003 to 0.0007)? Instead of rat pups, we now had data on humans.
During that study, 37 human subjects were split into four exposure groups and exposed. Significant decreases in iodide uptake were found in the three highest exposure groups. Iodide uptake was not significantly reduced in the lowest exposed group, but four of the seven subjects in this group experienced inhibited iodide uptake. The RfD proposed by NAS [0.0007 mg/kg-day] was accepted by EPA and added to its integrated risk information system (IRIS) in 2005.
So based on one study [Greer] involving 37 human subjects, the RfD is calculated at 0.0007 mg/kg-day based on Iodine uptake.
Is that enough to mandate a cleanup level for a site contaminated with perchlorates? Well, whatcha got that's better? You see, that's all we have to go on. Small studies, most of the time based only on animals. How confident are we that 10 ppb of arsenic is safe? That 5 ppb of perchlorates is safe? If you factor in just how low the dose is that caused the adverse health effect, and then take in the fact that the dose is made even less due to uncertainty, the level we deem is safe is very protective of public health.
That's how it works. That's how we beat that dead horse, The question now becomes, how certain are we of the certainty of the RFD.
Next post: RfD and Consensus...Part 3