STATS at George Mason University

Media coverage - what you needed to know

The only information the public received about Hites et al. was from the news media, which ran most of its coverage of the research on the day it was published in Science magazine (January 9, 2004). The question is whether the public got the right kind of information to make sense of the claim farmed salmon posed a risk to their health. As this was, fundamentally, a story about numbers and what they mean, the reader needed to have the following information.

Regulatory limits
The FDA limits for PCBs in fish — or if outside the United States, the limits set by Canada Health, the World Health Organization, the European Union, the Food Standards Authority (United Kingdom) etc; the EPA limits for PCBs in fish; the PCB levels in the salmon analyzed by Hites et al. Without these numbers, the reader has nothing to go on except adjectives. How high is high? Many newspapers, for example, used some variant on “within FDA limits” to describe the PCB levels in the salmon analyzed by Hites et al; but what did that really mean — just within or well within?

Why the EPA limits differ from the FDA’s (and WHO’s, EU’s, etc)
The discrepancy between the EPA and FDA limits would also lead most readers to ask how each organization could have a different estimation of the same risk, and which estimation is the most accurate.

The increased risk of getting cancer according
to the EPA’s limits for PCBs in fish

Given that the point of covering the salmon study was to warn the public about an apparent health risk, one would expect journalists to tell readers that consuming eight oz of raw salmon per month containing PCBs with levels similar to the average found by Hites et al. produces an increased risk of cancer by 1 in 100,000 over 70 years.

Without this information, it is impossible to decide whether to change one’s diet or whether to dismiss the risk as too small to worry about. Moreover, it is unconscionably alarmist to warn about “cancer-causing contaminants”— as many news stories did — and then fail to state what the risk of getting cancer is.

The EPA’s lower risk estimation for non-cancer health risks
Given that the study’s authors were quoted in many news stories on the developmental and neurological risks associated with PCBs in farmed salmon, one would expect journalists to point out that the EPA’s estimation of risk for such health effects is four times lower than it is for cancer.

The risk context
In discussing the potential risks from eating farmed salmon, it is vital that those risks are put into a broader context by noting the relative benefits of eating salmon, the presence of PCBs in other foodstuffs, and the risks from other health problems related to diet such as heart disease and obesity.

The controversy over the claim that PCBs are probable carcinogens
Finally, any discussion of the risk of getting cancer from PCBs in fish should address the controversy over the EPA’s belief that PCBs in fish are probable carcinogens in humans. In many ways, this is the key element in the story, the very reason why we are being told about the study in the first place — we might get cancer from PCBs in farmed salmon!

And yet, as we have seen, neither PCBs in fish, nor chronic exposure to PCBs in industrial settings, have ever been shown to cause cancer in humans. Instead, the EPA has inferred this probability from tests where animals have been repeatedly dosed with massive amounts of PCBs, a method of risk assessment disputed by many scientists.

about staff publications email contact search contribute
Salmon home
email this page to a friend

STATS home