Page Banner

United States Department of Agriculture

Agricultural Research Service

Arsenic in Drinking Water - What's Safe?
headline bar

Eric O. Uthus

A recent Grand Forks Herald article featured a release by an organization called the Natural Resources Defense Council (NRDC) stating that North Dakota water contains too much arsenic. The article also noted, however, that all North Dakota towns are currently in compliance with the Environmental Protection Agency (EPA) interim maximum contaminant level (MCL) for arsenic of 50 parts per billion (ppb) set by the 1976 Safe Drinking Water Act. In 1996, EPA was directed to propose a standard for arsenic in drinking water. This proposal will be released next month. The NRDC wants this proposal to be 3 ppb.

In 1995, the cost to the American taxpayers to comply with a MCL of 5 ppb was estimated by the EPA to be 620 million dollars per year and 2.1 billion dollars per year for a MCL of 2 ppb. Does current evidence warrant the cost of drastically reducing human arsenic exposure? Probably not. Although arsenic in high doses has been associated with an increased risk of cancer, there are no human studies capable of determining whether arsenic in drinking water at or below the current MCL results in an increased risk of cancer. Thus, risk assessment must rely on extrapolation from exposure at higher intakes, well over 50 ppb, such as that found in an epidemiological study in Taiwan.

The NRDC and EPA apparently have decided to use the Taiwan study to calculate the risk for arsenic-induced cancer although a 1994 report, partially funded by the National Science Council, showed that the Taiwan data were too fraught with uncertainty to be used as a basis for the agency’s revision. There are many reasons for the Taiwan data being an unrealistic measuring stick for deriving an MCL for US drinking water. For example, the population supposedly affected by arsenic toxicity in Taiwan had poor socioeconomic status and were generally undernourished. Their intakes of nutrients that help prevent some cancers, such as protein, methionine, vitamin B12, folic acid, zinc and vitamin A, were considerably lower than that typically found in the US. The selenium intake of the people drinking well water containing 200 to 1000 ppb arsenic was similar to that (11 µg/day) in areas of China with endemic Keshan disease, a heart disorder associated with selenium deficiency. High arsenic can make selenium deficiency worse. Because selenium is known to have anticancer properties it is possible that the arsenic toxic effect in Taiwan was more a response to inadequate selenium. In the US, we do not have to worry about this because selenium intakes range from 98-224 µg/day.

We need data that directly show whether low doses of arsenic have any direct effect on the susceptibility to cancer and other diseases, especially since the body has the ability to handle arsenic in low amounts. Until then, data from animal experiments are useful. For example, in one study done by the U.S. Food & Drug Administration, rats ate diets containing arsenic concentrations as high as 250 ppm. After two years on these diets, the rats had no increase in tumors. Other studies with laboratory animals have shown that arsenic actually inhibits cancer.

Although the mechanism of action through which high amounts of arsenic can result in cancer is not known, one hypothesis suggests that arsenic causes DNA to be poorly methylated. This could change how our genes, made of DNA, control the increase in cells in tissues; an uncontrolled increase is cancer. Research done at the Grand Forks Human Nutrition Research Center suggests that a certain amount of arsenic is essential or beneficial for laboratory animals. Feeding rats a diet extremely low in arsenic results in poor methylation of DNA. This indicates that there is a needed amount or threshold value of arsenic – if arsenic is decreased below this value, harm could result.

A 1988 report of the EPA stated "Limited animal evidence suggests that arsenic might be an essential nutrient, raising the possibility that reducing the level of arsenic below some critical level (as yet unspecified) might result in a decrement of health in some way (as yet unknown)." Interestingly, by reanalyzing the Taiwan data and using data that were previously not obtainable, one major study showed that skin cancer associated with exposure to arsenic in trace concentrations (up to 10 ppb) in drinking water was lower in comparison to exposure to water with undetectable concentrations of arsenic; this supports the idea that a threshold does exist.

Perhaps arsenic should be reduced in drinking water, to 20-25 ppb, but not to 3 ppb. A drastic reduction would not only cost considerable money, it may result in the stated reason for lowering the arsenic: it may be harmful to your health.


Last Modified: 10/23/2006
Footer Content Back to Top of Page