Talking tuna.

Talking tuna.

Talking tuna.

Health and medicine explained.
March 29 2005 7:23 AM

Talking Tuna

How much sushi and albacore is too much? Be careful who you ask.

Illustration by Mark Alan Stamaty

New research continues to probe the long-term, neurological ills of mercury, the fish-loving contaminant. Doctors in California warn that some of their affluent, fish-eating patients show signs of mercury poisoning. The California attorney general's office is skirmishing with the tuna industry, demanding that it post warnings to consumers about toxic exposure. Watchdog groups are threatening to sue the FDA over its latest consumer advisory. In other words, it's time to talk tuna again.

No one denies that mercury is a powerful toxin, particularly damaging to the brain development of fetuses and young children. But the two government agencies separately charged with protecting human health and protecting the environment have traditionally disagreed about how much contaminated fish is too much. The FDA, which regulates bought fish, has more expertise in nutrition and has tried to weigh the harms of toxins against the health benefits of eating fish, with their heart-healthy omega-3 fatty acids. The EPA, which oversees caught fish, by contrast has focused on the health impact of contaminants, including PCBs, dioxins, and metals like mercury. In other words, the agencies are thinking differently about risk. Last year, they split their differences and offered a joint advisory on fish contaminated with mercury. But the dispute is beginning to rumble again inside the agencies and in the press—which is a good thing, because last year's guidelines should be revisited.


The interagency compromise, which leans toward the FDA's position, is mired in contradiction. It sets these limits for women of childbearing age, pregnant women, and nursing mothers: No more than two meals (12 ounces) a week of most fish, including salmon, catfish, and canned light tuna; no more than six ounces a week of albacore (aka white) tuna; and no swordfish, king mackerel, shark, or tilefish at all. Spend a moment with this handy mercury calculator, however, and you'll see that an average woman could follow these rules to the letter and still exceed the EPA's cutoff for mercury—0.1 micrograms of mercury per kilogram of body weight per day—by a sizeable margin. For instance, a 140-pound woman who ate just one 6-ounce can of white tuna a week would be 30 percent over the EPA cutoff, defined as a daily exposure level that over time is unlikely to cause appreciable harm.

The FDA justifies these loosey-goosey guidelines by painting the EPA's traditional approach to exposure limits as overly cautious. In establishing the mercury reference dose in 2001, the EPA looked at studies of adverse effects of exposure in the Faroe Islands, the Seychelles Islands, and New Zealand, and applied some standard statistical calculations. Then the agency built in an uncertainty factor to account for the reality that information is incomplete, some people are more sensitive to mercury than others, and so forth. All this is customary practice. Still, David Acheson, head of the FDA's Office of Food Safety, says that the EPA's threshold should be treated not as a "bright line" but as a sort of general guide. (Translation: Forget the mercury calculator.)

Acheson is right to view public health as a balancing act. Stronger warnings about mercury could scare off fish eaters, leading them to gorge, say, on artery-clogging baloney and hotdogs. Pregnant women and babies aren't the only group with health considerations—think of their husbands, who might be at greater risk of the cardiovascular trouble that eating fish can stave off, and who might hear of a mercury warning and assume it applies to them, too.

It isn't easy being the FDA, or any agency with a broad public health mandate. Critics attribute the agency's apparent laxness to the influence of industry, especially the mighty tuna foundation. But there is no widely accepted model for quantifying trade-offs between the benefits of certain foods and the dangers associated with small levels of contamination. It's even harder when the benefits apparently apply to the population at large, while the risks seem to be concentrated among a certain group—in this case, developing fetuses, nursing babies, and young children. And even if the warnings about mercury specify that their target audience is childbearing women and little kids, it's reasonable for the agency to worry about how its messages will mutate as they ripple outward. As smart academics have pointed out, people perceive risks in subjective and often inaccurate ways, particularly when the risks are environmental, when they're associated with industries that don't seem so trustworthy, and when they involve babies.

That said, the FDA position on mercury and fish—and the EPA's decision to sign off on it last year, despite contradictory work by its scientists—seems more questionable the closer you look. The EPA's own mercury exposure limit may not represent a bright line, but what if it's too loose rather than too restrictive? Philippe Grandjean, a professor at Harvard's School of Public Health whose work in the Faroe Islands was important to establishing the agency's mercury reference dose, now believes that this exposure limit should be substantially lower than it is. His argument is based on a re-examination of the impact of imprecise measurements of exposure. The point isn't that Grandjean is clearly right. It's that for now at least, the EPA's standard is the best measure we've got, based on an objective, reproducible, peer-reviewed methodology. What justifies issuing public health guidelines that fail to keep many people within this limit?

What's more, while in general the effort to weigh health risks against health benefits is a worthy one, when it comes to mercury and fish, the apparent trade-offs are not as stark as they seem. Fortunately the kinds of fish that are highest in mercury are not necessarily the same as those that are highest in heart-protective fatty acids. Swordfish, for instance, is high in mercury but low in omega-3s. Salmon, on the other hand, is high in omega-3s but low in mercury. In other words, a savvy consumer can sometimes have the best of both worlds—choosing heart-healthy fish without piling up a plate of mercury. (The issue is more complicated for other toxins, such as PCBs, which do tend to be higher in the fatty fish that are high in omega-3s.) New research has also shown that mercury itself can increase an adult's risk of heart disease. The toxin's presence, then, may actually undermine the cardio benefits of omega-3s—which would mean that too much fish really is too much.

The FDA and the EPA should continue to wrestle over trade-offs. It's a plus that the National Academy of Sciences has agreed to weigh in too, because its involvement will advance the science of risk-benefit analysis. For now, though, the most rigorous and credible standard for mercury exposure that we have is the EPA's reference dose, however imperfect it may be. So for those of you who want to play it safe, here's the best available data on how much mercury there is in fish, and here's that calculator again. Bon appétit.

Amanda Schaffer is a science and medical columnist for Slate.