In 2004 the EPA and FDA published new guidelines suggesting that pregnant women (and those who might become pregnant) limit their consumption of fish to 12 ounces (340 g) per week due to concerns about mercury exposure.
These guidelines, which were only intended for pregnant women for the sake of their developing children, have quickly become an accepted fact among the mainstream media, the medical establishment, and the general public. They have also been indiscriminately extended to other populations, including men and women who are not pregnant. I frequently have patients that tell me they’re limiting seafood consumption due to concerns about mercury, and it seems like there’s a new study or media story on this topic every week.
But is it really true that eating fish increases the risk of harm from mercury—either for the developing children of pregnant women, or anyone else? Should pregnant women be limiting their consumption of fish for this reason? And is it possible that that pregnant women might actually be harmed by following the FDA/EPA guidelines?
Are you limiting fish intake due to concerns about mercury? Find out why you might be doing more harm than good.
I answer these questions in detail below. This article will be longer than usual; it’s an important topic and I want to give it the attention that it deserves. For the time challenged among you, here are the key points:
- There’s no evidence that pregnant women or children (or anyone else) will be harmed by eating more than 12 ounces (340 g) of the most commonly consumed varieties of fish.
- In fact, there’s a lot of evidence that eating less than 12 ounces a week of fish could cause significant harm to both pregnant women and young children.
- Most pregnant women need to eat more fish, rather than less.
That’s the summary. Now let’s look more specifically at 5 reasons why you shouldn’t take conventional advice about mercury in fish at face value.
#1: Studies of mercury exposure in fish consistently ignore the important role of selenium
We’ve known about the role of selenium in preventing mercury toxicity for at least 45 years, with the first research report on this topic appearing in 1967. (1) Since then several studies have shown that selenium consistently and predictably counteracts the adverse effects of mercury exposure. (2)
How does selenium do this? Exposure to mercury is harmful because it deactivates special selenium-dependent enzymes—called selenoenzymes. Since the brain consumes nearly 25% of the oxygen we breathe, it continuously produces oxygen by-products that can damage the fats and proteins that make up the brain. Selenoenzymes are extremely important in the brain because several of them prevent oxidative damage while others actually reverse it.
In the past, researchers thought selenium was protective because it binds to mercury and prevents mercury from harming other molecules. This led to the mistaken idea that mercury causes harm in the body until selenium binds it. But our current understanding is almost the reverse: it’s not that selenium prevents mercury toxicity by binding to mercury, but that mercury interferes with selenonzyme function by binding to selenium. In fact, mercury cannot cause harm until it occurs in high enough amounts to inhibit a significant percentage of selenoenzyme activities. Mercury is only harmful because it binds to selenium and prevents it from performing its vital roles in the brain.
As long as you are eating fish that contains more selenium than mercury, the amount of selenium in the body will always be in plentiful excess of mercury. That means that these essential selenoenzymes are never inhibited to a meaningful degree. Fortunately, the vast majority of fish most people consume have more selenium than mercury. The exceptions are pilot whale, shark, tilefish, king mackerel and swordfish.
Unfortunately, the well-documented protective effect of selenium is consistently ignored in both the medical community and the media when reporting on potential harms from fish consumption. This is almost certainly causing harm, as it has led to advising pregnant women and young children to eat less fish, when we should instead be telling them to eat more.
#2: The evidence suggesting harm from consuming seafood is weak, and doesn’t apply to the fish most people consume
There’s no question that mercury is a potent neurotoxin, and that significant exposure can lead to problems in both children and adults. For example, catastrophic pollution from a chemical plant in Minamata Bay in Japan caused severe toxic effects in the local population during the 1950s and 1960s. (3)
But does that mean that eating seafood which contains mercury is harmful? This idea comes primarily from a study performed in the Faroe islands, which are located approximately halfway between Norway and Iceland, and another performed in New Zealand. (4, 5)
The Faroes study is problematic for several reasons. First, it examined the neurological and developmental effects of maternal consumption of pilot whale meat, (the source of the bulk of their total mercury exposure), and found that eating ocean fish actually protected against the adverse effects that were noted. Even if the blood levels of mercury observed in this study were associated with harm in developing babies (which is debatable, as I argue below), such low exposures were not expected to be harmful to adults and no harms were observed in adults. This is because adults have plentiful reserves of selenium throughout their bodies and in their brains that help protect their brain selenium levels from being depleted. It is only when these reserves are depleted that oxidative damage can begin to occur. Developing babies are more vulnerable because they have no stockpiles of selenium to protect against mercury.
Second, the “adverse effects” observed in this study were extremely subtle. The magnitude of the effects noted were on the order of one finger tap slower than an unaffected child in a test which measured how many times the child could tap in ten seconds. (Tap one finger as fast as you can on your computer keyboard or mobile device and you’ll get a clear idea of just how subtle this effect was.) What’s more, the children who had the highest whale meat exposures were primarily from rural areas, whereas the unaffected kids were from urban areas. It is possible that factors other than mercury exposure may have differed between the two groups (e.g. socioeconomic status, home environment, having to wake up earlier and drive a long distance to the test site, etc.), and those factors may have influenced the results instead of their mercury exposures.
Third, over 85% of the seafood consumed by pregnant women in the Faroe Islands study was pilot whale meat. Earlier in the article I mentioned that pilot whale meat is one of the few species of seafood that contains more mercury than selenium, and thus would be expected to cause harm. Pilot whale meat is also much higher in other environmental toxins like cadmium, PCBs, and dioxins. A lot of toxins can accumulate in a 5,000 lb. whale during its 45–60 year life at the top of the oceanic food chain.
The New Zealand study was also problematic. Its results were highly dependent on whether or not a single child was included or excluded. (6) That child had by far the highest blood mercury content, but also was a high achiever. Including that child made the study results insignificant, but excluding the child made the study results significant. When the results of a study depend significantly on a single subject, that throws the findings into doubt. In addition, selenium deficiency was quite common in New Zealand at the time this study was performed—they were one of the most selenium-deficient nations on earth at that time—and the take-out “fish and chips” consumed by New Zealanders in the late 1970s included fish such as sharks—which like pilot whale, are one of the few species of seafood which contain far more mercury than selenium. (7)
What these studies tell us is that consumption of seafood that is high in mercury and low in selenium by pregnant women is potentially harmful for their children, especially if overall selenium intakes are low. They do not indicate that consumption of commonly eaten seafood with more selenium than mercury cause harm to pregnant women, children, or anyone else.
#3: Studies of pregnant women consuming seafood show benefits, not harm—as long as the fish contain more selenium than mercury
There are four major studies that have evaluated the effects of maternal mercury exposure from seafood on subsequent child development. I discussed two of them above, and showed why they do not apply to most people who eat fish. That leaves two studies: one performed in the Seychelles Islands (northeast of Madagascar), and another in the UK. (8, 9)
The Seychelles and UK studies found no adverse effects from consuming seafood. On the contrary, the UK study indicated substantial benefits from increasing maternal fish consumption, and noted neurological and developmental impairments among children whose mothers had avoided fish consumption. These studies are better indicators of the effects of seafood consumption for pregnant women, because outside of a few areas like the Faroes, most seafood that is commonly consumed in the U.S. and around the world contains far more selenium than mercury.
At least one US study confirms the beneficial impact of maternal seafood consumption. Researchers examined the effects of prenatal mercury exposure in a group of women living in Manhattan at the time of the World Trade Center disaster. The women who ate more seafood did have higher levels of mercury in their umbilical cord blood, but that did not translate into worse outcomes for their children. On the contrary, consumption of seafood during pregnancy was associated with significant benefits in motor development and verbal and total IQ. (10)
Once again, we see that the effects of consuming seafood that contains less mercury than selenium are not comparable to those of consuming seafood with more mercury than selenium.
#4: Advising pregnant women (and everyone else) to reduce their fish consumption is NOT harmless
If eating fish only exposed us to mercury without offering any benefit, we wouldn’t be having this conversation. Unfortunately, that’s far from the truth.
In the UK study I mentioned above, while the researchers did not find harm in consuming more than 12 ounces (340 g) of fish per week, they did find that consuming less than that amount was associated with significant impairments of communication skills and verbal IQ at six and eighteen months of age, and continued on throughout their adolescence (these children are now ~20 years old). This study was by far the largest (14,000 mother-child pairs), best designed (multiple evaluations performed and repeated throughout life), and best controlled for other factors such as socioeconomic, educational, and a host of other potentially pertinent factors. They found the worst effects were observed in children whose mothers ate no seafood at all during their pregnancy (about 12% of the study population).
This should not come as a surprise. A large number of studies indicate that lower intake of long-chain omega-3 fats (found in fish) during pregnancy is associated with growth retardation, delayed or suboptimal depth perception, lower scores in tests which measure neurodevelopment, deficits in fine motor skills, speed of information processing in infants, and irreversible deficits in the release of key neurotransmitters like serotonin and dopamine. (11)
What’s more, it’s likely that both the benefits (from consuming more seafood) and harms (from consuming less seafood) that were observed in the UK study would be amplified if it was repeated here in the US. The ratio of omega-6 to omega-3 fats in the US is significantly higher than it is in the UK. A high intake of omega-6 fats limits the conversion of short-chain omega-3 fats into their longer-chain—and far more important, from a health perspective—derivatives like EPA and DHA. This makes it even more necessary for pregnant women (and others) in the US to obtain these beneficial long-chain omega-3 fats directly from food. And guess what the only significant dietary source of these fats is? That’s right: seafood.
Finally, in addition to the data I mentioned above linking higher intake of omega-3 fats to better neurological and developmental outcomes in children, there is evidence that fish consumption is beneficial for adults. Studies have shown that even modest fish consumption (1-2 servings of cold-water, fatty fish like salmon per week) are associated with a 36 percent decrease in deaths from heart disease, and a 17 percent reduction in deaths from all causes. (12) For more on the benefits of eating fish, see my article “Is Eating Fish Safe? A Lot Safer Than Not Eating Fish!”.
#5: Ignoring the protective effects of selenium underestimates the risk of eating some fish
Although most ocean fish contain far more selenium than mercury, the selenium status of freshwater fish is much more variable. Mercury tends to accumulate in fish in lakes where selenium availability is limited. That leads to a double-whammy where some freshwater fish have relatively high mercury levels along with low selenium levels. This may explain why adverse effects were observed in a study of freshwater fish consumption in Finland, a country notorious for its poor selenium status, and was so low that it became the first nation to add selenium to its fertilizers. (13)
It’s important to note that even when total mercury levels in fish are below the recommended safety limit, consuming these high mercury, low selenium fish may still cause harm to consumers. In other words, it’s the ratio of selenium to mercury in the fish (known as the “selenium health benefit value or; HBVSe”)—rather than the absolute amounts of either—that determine whether it is safe to eat.
As the evidence above indicates, aside from the warnings to avoid shark, swordfish, tilefish, and king mackerel (varieties which can contain far more mercury than selenium), the EPA/FDA advice to limit fish consumption during pregnancy is not only unfounded, it is potentially harmful. From a public health perspective, pregnant women should be eating more fish—not less.
There is one reason, however, that I’m concerned about advising people to increase their fish intake: the environmental and social impact. To meet even the EPA/FDA guidelines for maternal fish intakes—which as I’ve argued in this article, are likely too low—would require a doubling of per capita fish intake in the US. Yet ecologists have warned for years that the exploitation of global fisheries at the current rate is not sustainable without massive conservation efforts. (14) Many fisheries are either completely exhausted or in significant decline, and pressure to increase fish consumption may exacerbate this problem if it’s not done in a sustainable manner.
This is a big problem—the elephant in the room when it comes to discussions about the health benefits of seafood consumption. As with many similar predicaments, there is no easy solution. I’m continuing to research the issue and will report back soon. In the meantime, I suggest using guides such as those published by the Marine Stewardship Council and The Monterey Bay Aquarium to choose only the varieties of fish that are being harvested sustainably. Fortunately, many of the most sustainable choices (e.g. sardines, anchovies, mackerel, wild-caught salmon, etc.) are also the fish that are highest in the beneficial long-chain omega-3 fats.
Now I’d like to hear from you. Have you been limiting fish consumption due to concerns about mercury? If so, how has the information in this article affected you? Are you concerned about the environmental impact of eating fish? Share your thoughts with me in the comment section below. (And please share this article with your friends and family, especially women that are pregnant and/or have young children.)
I’d like to thank Dr. Nick Ralston for his generous guidance and contributions to this article.