Contact an expert who can comment on the study
The complete study
Summary of the study
Consumption recommendation data at a glance
Frequently asked questions and answers about the new study
What can consumers do to limit their exposure?
Some of the contaminants found in farmed salmon
How the study's farmed salmon consumption recommendations were determined
Background information about world salmon production and consumption
Several small, peer-reviewed pilot studies on farmed salmon
About the study's authors
FOR IMMEDIATE RELEASE
For more information:
First Global Study Reveals Health Risks of Widely Eaten Farm Raised Salmon
Science Study Suggests Sharp Restrictions in Consumption
Albany, New York A study published this week in a leading scientific journal found significantly higher levels of cancer-causing and other health-related contaminants in farm raised salmon than in their wild counterparts. The study, published in Science and by far the largest and most comprehensive done to date, concluded that concentrations of several cancer-causing substances in particular are high enough to suggest that consumers should consider severely restricting their consumption of farmed salmon.
The majority of salmon served in restaurants and found on grocery store shelves is farmed rather than wild. In most cases, as detailed in the study, consumption of more than one meal of farmed salmon per month could pose unacceptable cancer risks according to U.S. Environmental Protection Agency (EPA) methods for calculating fish consumption advisories.
The Pew Charitable Trusts, one of the U.S.ís largest philanthropies, sponsored the study. Pew has sponsored major research on fisheries including a number of widely reported recent studies on the deterioration of the marine environment.
Whereas earlier studies have analyzed anywhere from 8 to 13 salmon samples from individual salmon farming regions, the current study analyzed fillets from about 700 farmed and wild salmon produced in eight major farmed salmon producing regions around the world and purchased in 16 large cities in North America and Europe. The studyís authors, six U.S. and Canadian researchers representing fields from toxicology to biology to statistics, selected salmon samples to be representative of the salmon typically available to consumers around the world.
The researchers found significantly higher concentrations of contaminants in farmed salmon versus wild. In particular, four substances that have been well studied for their ability to cause cancer PCBs, dioxins, dieldrin, and toxaphene were consistently and significantly more concentrated in farmed salmon as a group.
Among the studyís conclusions, salmon farmed in Europe were generally more contaminated than farmed salmon from North or South America. Farmed salmon purchased for the study from supermarkets in Frankfurt, Edinburgh, Paris, London, and Oslo were the most contaminated and triggered consumption recommendations of one-half to one meal per month based on U.S. EPA consumption advisories for these contaminants. A meal was considered to be an eight-ounce portion.
Farmed salmon purchased from supermarkets in Los Angeles, Washington, D.C., Seattle, Chicago, New York, and Vancouver triggered a recommendation of no more than two meals per month.
There was slightly more variation in fish purchased in North America than those purchased in Europe. While farmed salmon purchased for the study in New Orleans and Denver were generally least contaminated triggering a recommendation of about 3 meals per month farmed salmon purchased in Boston, San Francisco, and Toronto triggered the more stringent consumption recommendations of the European-purchased fish.
"Ultimately, the most important determinant of risk has to do with where the fish is farmed not where it is purchased," said Dr. David Carpenter, an author of the study and Director of the Institute for Health and the Environment at the University at Albany. "And because itís a global market, itís hard to be sure what youíre getting."
According to Carpenter, "Just because Europeans have the most contaminated farmed salmon, this doesnít mean American consumers shouldnít be concerned."
With very few exceptions, farmed salmon samples tested significantly exceeded the contaminant levels of wild salmon, which could be consumed at levels as high as 8 meals per month. Even the least contaminated farmed salmon, from Chile and the state of Washington, had significantly higher levels of PCBs, dioxins, and dieldrin than wild salmon.
Contamination Likely Related to Feed
The Pew-sponsored study concluded that the contamination problem is likely related to what salmon are being fed when theyíre on the farm. While wild salmon eat a diverse buffet from small aquatic organisms like krill to larger fish, farmed salmon are fed a concentrated and high fat mixture of ground up fish and fish oil. And since chemical contaminants a fish is exposed to during its life are stored in its fat, the higher fat "salmon chow" passes along more of these contaminants to the farmed salmon.
The studyís results confirmed this possibility when it found higher contaminant concentrations in salmon feed from Europe than feed from North and South America, a result roughly consistent with contaminant levels in European and American salmon.
Consumption Advisories and Recommendations
Given the overall contaminant levels found, if these were locally caught fish instead of fish purchased commercially EPA and many state consumption advisories would suggest that consumers restrict their consumption of farmed salmon to an average of no more than one meal per month. However, consumers need to be aware that in some cases even that could exceed advised contaminant exposure levels. EPAís consumption advisories use acceptable lifetime risk levels to identify the maximum number of fish meals per month that can be safely eaten.
"If anything, the study conservatively estimates the health risks from the contaminants in farmed salmon," said the University at Albanyís Carpenter. The EPA fish consumption guidelines donít take into account exposures people have to the same cancer-causing substances from all other sources in the environment. "They assume," said Carpenter, "that fish consumption is the only source of exposure people have to these substances; and we know thatís not true." "Also," according to Carpenter, "the recommendations only consider the risk of cancer and donít take into account the neurological, immune, and endocrine system effects that have been associated with these contaminants."
Consumers interested in knowing whether salmon is wild or farmed should be aware that the word "Fresh" on the label does not mean the salmon is wild-caught from the ocean. And any salmon labeled "Atlantic" in the U.S. is almost always farmed. Salmon labeled "Atlantic" in other countries is most likely farmed. The authors recommended that governments require clear and prominent labeling of farmed and wild salmon as well as the country of origin of all farmed salmon.
The authors also said their results strongly reinforced the recommendations of a July 2003 National Academy of Sciences report on dioxins in the food supply which called for reducing dioxin levels in animal feed such as fishmeal.
Since contaminants build up in the fatty tissue of the fish, the authors point out that consumers may be able to reduce their consumption of contaminants in farmed salmon by following the recommendations of many state governments and the federal government to remove as much skin and visible fat as possible. However, it is difficult to determine how much of the contaminant load can be removed in this way.
In assessing the human health risks of consuming farmed salmon, the authors of the study used U.S. EPA consumption guidance for PCBs, toxaphene, and dieldrin covering locally caught fish rather than U.S. Food and Drug Administration (FDA) standards for these substances governing commercially-sold fish because EPAís recommendations are based on health effects only. While FDA is the agency that actually regulates contaminants in commercial fish, unlike EPA FDA does not have consumption standards for toxaphene in fish, and the agencyís standards for PCBs and dieldrin werenít set using purely health-based criteria. According to Dr. Barbara Knuth of Cornell University and one of the studyís authors, "Because the FDA regulatory levels take into account factors such as effects on the food production system, they were never designed to consider exclusively human health risk-which was the only concern we were looking at in this study."
"Plus," said Knuth, "the health and diet information and the technology FDA used to help set the regulatory levels for PCBs are 20 years out of date. We can detect PCBs at much lower levels today; new studies provide more information about the health risks associated with these substances; and people eat more fish today." In fact, both EPA and FDA have agreed that FDA levels are inappropriate for setting fish consumption advisories (see last paragraph on p. 1-5 in the EPA National Guidance for Fish Advisories).
Knuth said, "Itís this vast difference in the approach of the two agencies that explains why farmed salmon with these levels of contaminants could trigger such restrictive consumption recommendations based on EPA methods, but is still allowed to be sold legally in the U.S. by the FDA."
The annual global production of farmed salmon has increased 40 times during the last two decades making inexpensive salmon available to consumers year-round. Between 1987 and 1999, salmon consumption increased at an annual rate of 14% in the European Union and 23% in the U.S. Since 2000, over half of the salmon eaten globally has been farmed, coming primarily from fish farms in Northern Europe, Chile, and Canada.