Sign up to receive email updates, action alerts & health tips from EWG. [Privacy]

Myth 1: Animal tests of pesticides don't predict human cancer risks

Pesticide Industry Propaganda: The Real Story: Myth 1: Animal tests of pesticides don't predict human cancer risks

May 1, 1995

Myth #1: Animal tests of pesticides don't predict human cancer risks because:

  • the doses tested are so high that "everything causes cancer," and
  • animal results are irrelevant to humans, because "mice are not little men."

Here's The Real Story Behind Myth #1

Myth #1: Animal tests of pesticides don't predict human cancer and other health risks because:

Animal studies are the public's first line of defense against toxic substances. Major public health disasters have been avoided or minimized, because regulators acted on the basis of animal studies. For example, DDT was banned due to problems first identified in animal tests.

As plainly stated by Dr. David Rall, former director of the National Institute of Environmental Health Sciences, "Animal studies must serve as a primary tool of prevention. Epidemiology studies, while valuable, often provide information 25 years too late."

In other cases regulators ignored the results of animal studies, causing great human suffering. Workers were not protected from asbestos until after lung cancer cases in workers were linked directly to occupational exposure to the substance. Evidence that asbestos caused cancer in animals was suppressed by the manufacturer for at least 15 years. Animal evidence was also ignored with the fertility drug DES, which was not banned until the daughters of women who took it developed a rare vaginal cancer. High dose animal testing is used by every public health agency around the world, from EPA to European bureaus to the World Health Organization-- and even by industry when it likes the results (for example when these studies prove the safety of drugs, cosmetics, or other pesticides).

Animal studies accurately predict risk for humans. Extrapolating from mice to men is logical because rodents and humans are remarkably similar genetically (Rall et al. 1987). It is not surprising, therefore, that all known human carcinogens have also been shown to cause cancer in experimental animals. Most scientists agree that it is prudent to assume the reverse is also true and that chemicals clearly causing cancer in animals present human risks (NRC 1993a). The same is true for chemicals that cause birth defects in humans; they all cause birth defects in animal studies (Kimmel et al. 1992). In fact, current animal testing protocols, particularly for cancer and subtle multigenerational effects, underestimate human risk (NRC 1993a). People are exposed to pesticides from conception through death. In contrast, animals are exposed typically beginning at 8 weeks (roughly equivalent to 5 years of age in the human), and ending at two years (roughly equivalent to age 65 in the human). One study designed to better understand this shortcoming found that rats fed the carcinogens N-nitrosodiethylamine (NDEA) or N-nitrosodimethylamine (NDMA) for two and one-half years had seven times the cancer incidence compared with rats fed NDEA or NDMA the standard two years required by the EPA (Peto et al. 1991).

A substantial body of evidence points to dramatically increased cancer rates when experimental animals are dosed in the womb and as neonates. A major study of 1,040 animals found a six-fold increase in cancer incidence when exposure began at three weeks, as compared to 20 weeks of life (Gray et al. 1991). Another review of animal studies on 22 chemicals found that more cancers were produced, and were produced earlier in life, when animals were exposed from conception and during weaning (McConnell 1992).

Most chemicals do not cause cancer, even when tested at very high doses. To discredit animal tests of pesticides, industry fixates on the Maximum Tolerated Dose (MTD) (the highest dose that can be fed to an animal without causing tissue damage), claiming that virtually "everything causes cancer" at such a high level. But in fact most chemicals are not carcinogenic even when tested using maximum tolerated doses; of the hundreds of chemicals tested by the National Cancer Institute and the National Toxicology Program, 68% proved carcinogenic when selected for testing because of their suspected cancer-causing potential. When chemicals were tested on the basis of potential high human exposure, only 22% caused cancer in high-dose tests, suggesting that about one-fifth of all environmental pollutants may cause cancer in high-dose animal tests (Rall 1994, Fung, et al. 1993). Most chemicals that cause cancer at high doses also cause cancer at low doses. A review by the National Toxicology Program found that only 6% of all chemicals analyzed caused cancer at the high dose only (Rall 1994). While several alternative theories have been advanced, mainstream scientists still agree that there is no dose of a carcinogen that does not increase the risk of cancer (Portier et al. 1994). This is particularly true in the current environment where people are exposed to scores of carcinogens each day, each one adding to the cancer risk of the other.

Several important animal studies have tried and failed to identify a so-called "threshold", or safe dose. Recently, using extremely low doses on over 4,000 rats, researchers were unable to find a dose of N-nitrosodiethylamine or N-nitrosodimethylamine that did not significantly increase cancer rates (Peto et al. 1991).

Industry further complains that the government overreacts to reports of rodent tumors and tries to ban any chemical so implicated. In fact, the opposite is true. According to the Office of Technology Assessment, most rodent carcinogens are not regulated and few are banned (OTA 1987). Of the more than 90 pesticides found to cause cancer in animal studies, the vast majority continue to be used on food crops. (See Note 1).