Flip-flopping headlines — Part Two

The body of evidence

Women have been eating fish and delivering healthy babies for ages. The findings in the latest Lancet study were not at all surprising, despite the confusing, tentative and unclear media reports.

Why? Because the body of the highest-quality and most careful research has never shown that the amount of fish — and the amounts of methylmercury in the fish — eaten by American pregnant women and children put them or their babies at risk. And even among populations eating ten or more times the fish Americans eat, scientists have found no credible evidence of neurotoxicity, let alone anything remotely resembling brain damage, developmental delays, retardation or learning disabilities among the children.

Of course, this is considerably different from what we’ve heard for years. The media and alarming government announcements have been warning expectant mothers of possible neurological dangers to their unborn babies and children because of methylmercury and advising them to limit their fish consumption.

When I first heard the FDA advisories many years ago, my heart stopped. I’d been in neonatal intensive care and pediatric emergency triage for much of my nursing career and had been following the American College of Obstetricians and Gynecologists and most health professionals in encouraging women of childbearing age to eat 2 to 3 servings of a variety of fish each week to help ensure healthy babies and pregnancies. Had fish suddenly become more dangerous to eat? Or, worse, had we been giving bad information for decades that was putting mothers and babies in harms way? Thus began a multiple-year research project, reading hundreds of studies and position papers, pages of transcripts of FDA scientific meetings, interviewing toxicologist across the country, and countless sleepless nights as I reviewed and re-reviewed the evidence to make sure that the information was as accurate and sound as possible, and that I hadn't missed something. The more I read, the more concerned I became — not by anything frightening demonstrated in the evidence but because the science was so different from the scares in the media. Understandably, most healthcare professionals don’t have the time to devote to an exhaustive review of the evidence but lamentably few have gone to the original studies to see for themselves what the evidence really shows. Even doctors on television with the opportunity to help clarify the science for consumers haven’t, as seen in a local spot discussing the latest study.

The only cases in the scientific literature of mercury poisoning from fish and subsequent neurological problems — a point confirmed by Dr. Thomas Clarkson, a toxicologist at the University of Rochester School of Medicine — were the result of an industrial mercury spill in Minamata Bay, Japan in the 1950s, which resulted in fish with methylmercury levels 40 to 1,000 times higher than the fish Americans and most people around the globe eat. These tragic poisonings first pointed out that at extremely high exposures, mercury was a neurotoxin and might affect the developing fetuss.

Methylmercury — an organic compound produced by methane-generating bacteria deep in bodies of water as they ingest inorganic mercury — is nothing new. It’s always naturally been in fish and in human bodies. In fact, despite recent fears, the trace levels people have been exposed to have been dropping for centuries, as evidenced in tests on 200-year old fish samples at the Smithsonian Museum and of 550-year old Alaskan mummies, both of which found levels multiple times higher than they are today.

Scientists took concerns arising from those heartbreaking earlier poisonings seriously, however, and launched the most comprehensive, state-of-the-art methylmercury studies ever conducted in the world to make sure that the fish we eat is safe. These are the primary sources for the most indepth and careful information on methylmercury exposure risks. Funded by the National Institute of Environmental Health Sciences, these studies — in the Faroe Islands, a Danish territory in the North Sea, and the Seychelles, an island group off the east coast of Africa — have exhaustively followed nearly 2,000 children and after 15 years have been unable to find any manifestations of harm among the children.

The first, the Seychelles study, is the most careful and exhaustive, double-blind, longitudinal methylmercury exposure study ever conducted on mothers and children. The Seychelles people have the highest per capita consumption of fish in the world, typically 12 fish meals each week, according to the Agency for Toxic Substances and Disease Registry’s Dr. Christopher DeRosa. We eat the same fish varieties, but average only one serving a week. The methylmercury exposures of Seychelles children are 10 to 20 times higher than among U.S. children, yet every time the researchers have tested the children, they’ve found no detectable adverse effects. The researchers used numerous global tests and have evaluated over 57 endpoints for the children’s neurocognitive, language, memory, motor, perceptual-motor, and behavioral functions and followed them for more than 14 years now.

The second study is the source behind most claims of evidence for concerns. It deserves a close look, as healthcare professionals and consumers might be surprised by what it actually found. The Faroe Island study reported an association between prenatal methylmercury exposure as measured in maternal cord blood (blood present in the umbilical vessels at the time of delivery) and “subtle and small neuropsychological changes” in the children. However, the FDA Food Advisory Committee and scientists at methylmercury scientific meetings held in 2002 voiced concerns about this study and its conclusions. Even lead researcher Dr. Philippe Grandjean from Odense University in Denmark admitted to the FDA Committee that testing children’s development proved “complicated to do the same way every time” and, while a mercury effect appeared in the children examined by one neuropsychologist, no effect was seen when they adjusted for different examiners. They also didn’t find meaningful correlations between any developmental delays and methylmercury levels as measured in the mother’s hair.

The significance of the “subtle” changes the researchers used to base their conclusions were also questioned by the FDA because, while a considerable number of developmental tests were done on the children, only one at age 7 — the Boston Naming Test (developed to help diagnose aphasia, difficulty comprehending spoken or written language or in speech expressions, in elderly adults after strokes, brain trauma, or dementia) — suggested any risk. They reported a 0.25% difference in mean scores (a mere 0.1 point) between children of mothers eating no fish and those eating the most. Not only is this well below the standard deviation of 4 to 6 points for this test, but the test itself was not designed to be used in children and its developers, neuropsychologists Drs. Edith Kaplan and the late Harold Goodglass, noted problems using it in children.

As the children grew older and the Faroe Island study was not double blind, the researchers were able to find only one test to base claims of subtle effects of methylmercury: the Brainstem Evoked Potential. This test uses electrodes placed on the test subjects’ heads to measure the electrical signals from the brain in response to auditory stimuli. Grandjean and colleagues found only a 1/100th of a millisecond difference between the controls and the Faroe children. Their questionable interpretation of this as “worrisome” evidence of irreversible developmental damage is as good as the evidence gets to support fears of dangers in today’s fish consumptions for Americans.

For more than 30 years since the earlier poisonings, University of Rochester researchers have closely followed populations eating large amounts of fish, including Samoans, Peruvians and residents of the Seychelles, and not found any associated adverse effects in adults or children. Other smaller studies, including those in New Guinea and Canada, have also reached null results. And a small New Zealand study was found by a number of scientific reviews, including the WHO and National Academy of Sciences, to be too rife with problems to be used.

While there are more detailed analyses of the available research available, even this brief overview is probably more than most consumers and healthcare professionals have encountered. While the evidence should have been reassuring, that’s not what consumers have been hearing for years. Instead, women have been frightened into believing that they and their little ones are in actual danger (or "at risk") if they cross a “safe” level of exposure.

What is never explained is that this level, called the reference dose, does not represent a cut-off where any actual risk has ever been evident. It is a threshold representing arbitrary EPA safety cushions that are many times beyond where any danger has ever been shown. The EPA’s safety cushion is the most restrictive in the world and most other scientific agencies here and around the world examining the same evidence, including the FDA, have established minimum exposure levels multiple times higher.

To arrive at their levels, the EPA took a level where there was no observed effect at all in the most sensitive of the population with a lifetime of exposure — a methylmercury level nearly ten times that found in American women — and added another ten-fold safety cushion to that.

When the U.S. Centers for Disease Control and Prevention’s National Health and Nutrition Examination Survey (NHANES) studied the mercury levels among American women of childbearing age and young children, it found not a single woman or child had values anywhere near unsafe levels. Its November 2004 report, for example, found only that about 6% of women were at or above the EPA’s reference dose, but none were anywhere near actual danger.

To determine if it was even possible for American women to eat enough fish to put their newborns at risk, the U.S. Department of Agriculture’s Human Nutrition Information Services analyzed the diets of American women of childbearing years using several surveys and factored in endless possibilities — such as heavy fish consumption, eating fish varieties with the highest methylmercury levels, repeatedly eating the same fish such as canned tuna, and the amounts of methylmercury in a range of commercial fish samples. After 100,000 iterations, they found it was inconceivable for a mother to eat so much purchased fish she’d put her baby anywhere near harm’s way. Dr. James Heimbach, former associate administrator with the HNIS, reported at the FDA scientific meetings that American women “simply are not exposed to levels of methylmercury that would place the newborn children at risk.”

Ninety percent of the fish we eat is purchased and all fish legally sold in the U.S. must be under the FDA’s safe limits for methylmercury. Farm-raised fish, the second largest source of fish in our diet, contain especially low levels because they’re fed a special, mercury-free diet. The FDA found that the 25 most popularly-eaten fish average methylmercury levels of 0.12 ppm, about one-tenth of the legal limit. So by eating a variety there is no evidence we need to be concerned about subtle variations in individual samples and the FDA Committee said this strategy should give all Americans a 10 times or higher margin of safety/uncertainty when measured against the worst case possibility.

The FDA’s determinations for safe levels in food are universally-recognized as most accurate for safe human exposures in foods, as they not only weigh the body of evidence for health benefits and risks, but also test foods as they are actually consumed by the public; whereas the EPA measures environmental contaminations, including residues on skins and parts of foods that are not eaten. For the small numbers of us who enjoy recreationally-caught fish, the EPA monitors fishing waters and issues local fishing advisories when higher levels are found. So follow EPA advisories concerning safe fishing spots to minimize worry.

During the years the FDA was issuing health advisories spelling out dangers in increasing detail, the scientific community was objecting, noting the lack of credible scientific evidence to support such warnings, and that they were unnecessarily frightening consumers, especially taking advantage of the special vulnerability of expectant mothers. Multiple scientists warned that needless advisories could curb mothers and children from eating a nutritious food. Experts specifically cautioned the FDA Advisory Committee of the importance of not deterring consumers from canned tuna, the primary affordable, easy-to-prepare and most popular fish protein for most consumers. The latest warning, issued jointly by the FDA and EPA, was the most complicated ever, and among the details, it instructed women of childbearing age and children to limit fish to 2 servings a week of varieties low in mercury.

What was behind these advisories was not growing evidence for concerns but an intense campaign of political and public pressure led by special interests. (An enormously well-funded and well-organized collection of environmental and organic groups were working to alarm us about the safety of our food and environment in order to influence mercury emissions legislation. Many were aided by Fenton Communications, a for-profit public relations firm for activist
organizations whose employees also staff Environmental Media Services, which supplies media stories and spokespeople for Fenton clients. Its clients also include the American Medical Association and Center for Science in the Public Interest who were also visibly active in these efforts. The FDA mercury warnings were used as "proof" of a health threat.) For years, prior to each new mercury advisory, the FDA Commissioner and Secretary of Health and Human Services had been besieged by letters and repeated legal motions had been filed to block the FDA from issuing less stringent advisories. The public was inundated in the media with insupportable claims and frightening information about the threat of mercury in fish; legislation was introduced in states across the country; lawsuits were filed against fish businesses for exposing people to “known carcinogens or reproductive toxins;” and letter writing campaigns resulted in half a million letters to the EPA.

This situation is an example of the FDA and EPA capitulating and of politics finally trumping science. Sadly, the scientists’ original concerns about the government’s health advisories quickly appeared to be realized.

Understandably, pregnant women and new mothers are especially sensitive to health advice and have been most alarmed by these warnings. For many, fish seems too scary to bother to eat at all, never mind trying to make sense of the complicated advisories. A recent Harvard study found that after the well-publicized 2001 mercury advisory, pregnant women in eastern Massachusetts dramatically reduced their consumption of dark meat fish, canned tuna, and white meat fish, resulting in about 1.4 fewer servings of fish per month. Charles Santerre, Ph.D. of Purdue University reported that among the low-income women of childbearing age in his studies, about one-third of them hadn’t eaten any fish at all in the past year.

Carefully weighing the validity of scares and the possible consequences of frightening young pregnant women (and all consumers) is an important role for healthcare professionals. The health and well-being of women and their babies should ideally be paramount; advancing agendas should never take precidence. Not only have these scares added to the stress and anxiety of 4 million women who become pregnant each year, they’ve also discouraged them from enjoying a delicious, safe and wholesome food with plenty of nutrients for themselves and their babies. The expected fears that come from warnings, rather than being helpful or healthful, add to stress levels. Extremely high stress for expectant woment can contribute to increasing risks for preterm or low-birthweight babies, according to the March of Dimes.

As with most scares about our foods, when the body of evidence is carefully examined we find a lack of defensible evidence to support them. Looking at the evidence on the safety of fish and seafood, we find that the actual science hasn’t flip-flopped, the results of this latest study were not surprising and merely added additional support to the existing evidence negating concerns, and that there is no research to credibly justify claims that American babies have been born with mental deficits because their mothers ate or didn’t eat fish.

© 2007 Sandy Szwarc. All rights reserved.