Glossary

• Action Level
  The Food and Drug Administration has set an "Action Level" of 1 part-per-million for mercury in fish. Legally, the FDA is empowered to prohibit the sale of any fish that exceeds this level. Activist groups complain that this Action Level is a "toothless" regulation, but the FDA itself has put it in its proper perspective, writing that its mercury Action Level "was established to limit consumers' methyl mercury exposure to levels 10 times lower than the lowest levels associated with adverse effects." In this light, it's a good thing that the FDA isn't removing fish from the market because they contain just one-tenth the amount of mercury that might be a cause for concern.
  
  
• Benchmark Dose Lower Limit
  Before the EPA divided by ten to reach its "Reference Dose" for mercury, it discerned a "Benchmark Dose Lower Limit" (BMDL) of 58 micrograms per liter (µg/L) of blood from a selective reading of existing science. [see "Faroe Islands"] After a 10-fold margin of safety is factored in, the resulting EPA "Reference Dose" is by far the most restrictive in the world. The World Health Organization, the British Food Standards Agency, the United Kingdom, and the Canadian government all have much higher "safety" limits for mercury than ours. But this is much ado about nothing: Recent data from the Centers for Disease Control and Prevention indicate that no one in the U.S. tests anywhere near the BMDL.
  
  
• “Detection” and “Reporting” Limits
  A “Minimum Detection Limit” (MDL) is the smallest amount of a substance that laboratory equipment can measure under controlled circumstances. The mere fact that something is “detected,” however, is not sufficient evidence that the substance is present. Eyedroppers attract dust, after all. And it’s difficult to clean lab glassware sufficiently to completely rule out the presence of contaminants.
  
  Consequently, most reputable labs also specify a “Minimum Reporting Limit” (MRL), a higher threshold than the MDL. (This is also sometimes referred to as a “limit of quantification.”) Results above a lab’s MRL are considered reliable, and appropriate to report publicly.
  
  The flawed Institute for Agriculture and Trade Policy (IATP) study which found trace amounts of mercury in high fructose corn syrup samples used a lab that reported its results all the way down to its MDL. Practically all of IATP’s results would fall between the MDL and a typical MRL. Accordingly, most of IATP’s results should never have been reported publicly at all.
  
  
• Faroe Islands
  The Faroes are a chain of Danish-governed islands in the North Atlantic Ocean, where the native population eats a diet heavy in fish and pilot whale. Because this whale meat is unusually high in mercury and other pollutants, many scientists believe that mercury measurements of this population are misleading. Like the EPA, many activist groups claim that the Faroe Islands study is iron-clad evidence that mercury in fish is a clear health danger. But researchers studying the Faroe Islanders have consistently refused to release their data to the global scientific community, calling their pronouncements into question.
  
  The best evidence of health risks from mercury in the Faroe Islands study is the slightly lower-than-normal performance of some children on a psychological exam called the Boston Naming Test -- a test designed for use with elderly adults. The test has a "standard deviation" (think of it as a plus-or-minus margin of error) of between four and six I.Q. points, but even the worst-performing Faroe Islands children tested were only off by around 1/10 of one point.
  
  In 2004 the lead Faroe Islands researcher wrote in The Boston Herald that "fish consumption does not harm Faroese children. In the contrary [sic], the fish consumption most likely is beneficial to their health." Nevertheless, the U.S. Environmental Protection Agency relies primarily on this study to calculate its extraordinarily conservative "Reference Dose" for mercury consumption.
  
  
• Methyl mercury (MeHg)
  When "elemental" mercury (like the slick, silvery stuff of thermometers) meets microorganisms in water or soil, it can be transformed into methyl mercury. The name is taken from the "methylation" that occurs when methane-producing bacteria do the job. The dominant scientific view is that methyl mercury is formed deep in ocean waters where sediment traps mercury. Most scientific surveys have found that the level of methyl mercury in ocean fish has remained constant (or even declined) during the past 25 to 100 years.
  
  
• Minamata Bay
  In the 1950s, 111 people from Minamata City, Japan were poisoned after they ate fish contaminated by a nearby industrial mercury spill. These fish were artificially spiked with up to 40 parts per million (more than 20 times the highest average levels found in U.S. fish). And then, as now, the Japanese ate far more fish in their diet than Americans ever have. In a similar case in 1965, 120 residents of Niigata, Japan were also poisoned by industrially-contaminated fish. According to University of Rochester School of Medicine toxicologist Dr. Thomas Clarkson, these Japanese poisonings are the only clinical cases anywhere in the scientific literature that document acute mercury poisoning from fish.
  
  
• NHANES
  The National Health And Nutrition Examination Survey (NHANES) is a massive health-census conducted by the Centers for Disease Control and Prevention (CDC). NHANES measures everything from Americans' height and weight to our cholesterol and immunization histories. And along the way, NHANES surveyors have also measured the amount of mercury in the bloodstreams of thousands of U.S. residents. It's the analysis of this scientific data that leads to conclusions about how many adults (and unborn children) are supposedly "at risk" because of mercury exposure. But these levels are always measured against the Environmental Protection Agency's artificially low "Reference Dose," which has a ten-fold margin of safety built in. When activist groups report that 7.8 percent of women in the NHANES survey tested above the EPA's "safe" level for mercury, they usually forget to add that this "safe" level is still 1,000 percent lower than the threshold for any actual health risks.
  
  
• Part-Per-Trillion
  One part per trillion (or one 1,000,000,000,000th) is a scientific measurement of an extremely weak concentration—so small that it couldn’t be reliably measured a decade ago. One part per trillion is equivalent to a single drop of water in 26 Olympic-size swimming pools.
  
  There is no existing scientific study showing that a concentration of any substance at the one-part-per-trillion level is harmful to human health.
  
  
• Reference Dose (RfD)
  Based on a controversial and suspect scientific study [see "Faroe Islands"], the U.S. Environmental Protection Agency (EPA) determined that the lowest "safe" lifetime average dose of mercury in the human bloodstream was 58 micrograms per liter. [see "Benchmark Dose Lower Limit"] But the agency then applied an "uncertainty factor" to its calculations, lowering that number ten-fold. The resulting "reference dose" of 5.8 micrograms per liter is a ridiculously low number and a meaningless measure of whether or not Americans' health is at risk. Even so, only a small percentage of American women test above this level. The EPA has indicated that it may soon review the Reference Dose (and perhaps lower it still further), opening the door to pointless warning labels and nuisance lawsuits against grocers, restaurants, and fishermen.
  
  
• Scientific Control
  Scientists use “controls” in order to eliminate alternate explanations of experimental results. The classic example is a clinical trial of a new medication: Half of the participants will receive the medication, while a similar group of patients take an inactive placebo. This is the only way to be sure that results (whether good or bad) can be attributed to the drug being studied, rather than other factors.
  
  The deeply flawed 2009 Institute for Agriculture and Trade Policy (IATP) study which found trace amounts of mercury in high fructose corn syrup samples did not include a control group. An appropriate control would have been the mercury testing of grocery products that did not contain any high fructose corn syrup.
  
  By failing to even consider the possibility that tiny, trace levels mercury might be uniformly common in products found on grocery store shelves, IATP rendered its conclusions scientifically meaningless.
  
  
• Seychelles Islands
  The Seychelles, a group of islands in the Indian Ocean near Madagascar, are the site of the world's most comprehensive study of the effects of fish-borne mercury on human health. After following a group of Seychelles Islanders for fifteen years, the researchers reported no negative health effects from the mercury in the local diet -- even though the Seychelles residents eat an average of 12 fish meals every week. Ironically, the U.S. Environmental Protection Agency refuses to acknowledge this Seychelles Islands study, because it showed no danger from mercury. The EPA is in the business of calculating health risks, and a study that gives fish a clean bill of health isn't of much use to the agency.
  
  
• Uncertanity Factor
  After the EPA determines the level of a toxic substance that could theoretically be harmful to people, the agency applies various Uncertainty Factors to slice that level down to an even smaller number, known as a "Reference Dose." For instance, it's common to use an Uncertainty Factor to account for differences that may arise when using animal studies to predict a chemical's behavior in humans, or to account for how different people might process the same chemical in different ways. For each category of uncertainty, EPA regulators pick a number that represents just how "uncertain" they are about its impact. In the case of methyl mercury (the mercury in fish), the EPA applied three Uncertainty Factors -- each with a value of 3 -- and then arbitrarily added 1 to reach a round number of 10. And voila! The "Benchmark Dose Lower Limit" (BMDL) for mercury, which is 58 micrograms per liter of blood, was divided by the Uncertainty Factor (10) to get a "Reference Dose" of 5.8 micrograms instead. Activist groups and some elected officials have deliberately confused the public by suggesting that this lower number is the gateway to a danger zone. But it just isn't so.