Something tastes bad…

When I first read reports of mercury in corn syrup, I was first shocked, then concerned, then skeptical. Janet (qualifications unknown) at Ethicurean described the source of the mercury (in fact, she was picked up by Huffington Post):

How did the heavy metal get in there? In making HFCS — that “natural” sweetener, as the Corn Refiners Associaton [sic] likes to call it — caustic soda is one ingredient used to separate corn starch from the corn kernel. Apparently most caustic soda for years has been produced in industrial chlorine (chlor-alkali) plants, where it can be contaminated with mercury that it passes on to the HFCS, and then to consumers.

First of all, I’m no particular fan of corn syrup; it tastes nasty and I avoid it. However, I also avoid added sugar or rice syrup or any other sweetener because I eat enough calories without them. Various types of foodie have been railing against HFCS for a long time, but I haven’t actually be able to figure out why. Instead of saying “HFCS is bad” we should be saying “processed food is bad”. Any special link between obesity and HFCS was broken in December with a comprehensive review in the American Journal of Clinical Nutrition (see press release in Newswise). The other argument against HFCS is that we are growing too much corn, but this is a sidestep at best. If people really cared about the amount of acres taken up by corn, they’d be saying “eat less meat” instead of “eat less corn syrup” (see the ISU Extension fact sheet about corn syrup for the uses of corn – ironically, you can’t get both ethanol and HFSC from a given bushel). Seriously, if you don’t like the stuff, then don’t eat it – but it’s helping no one to spread falsehoods and exaggerations.

Ok, back to mercury. While I’ll be the last person to say that the FDA is doing the best job in keeping us all safe, or that food processing conglomerates aren’t out to get a profit no matter what, American capitalism does have some protective effects. I’d wager that the Corn Refiners Association knew about the possible contamination source long ago and has done their best to remove or reduce it (which is exactly right, according to the CRA press release) simply to avoid future boycotts and lawsuits. Some commenters on Janet’s post were also skeptical, along with Marion Nestle on her Food Politics blog.

Marion points out that the study used no controls, and I heartily agree. The researchers should have obtained multiple brands of approximately equivalent foods (vanilla flavored yogurt for example), tested for mercury, and looked for any statistically significant differences between those that contain and those that do not contain HFCS. Without this comparison, the result that “nearly one in three” of the products contained detectable amounts of mercury is meaningless. Some amount of mercury is in everything we eat, processed or not, perhaps the result of decades of coal burning. Of course, there’s many more details to consider…

Some of their conclusions are good, like better food oversight and cleaning up chlorine plants, but, overall, the report Not So Sweet: Missing Mercury and High Fructose Corn Syrup from the IATP (Institute for Agriculture and Trade Policy) is propaganda. It’s full of inflammatory language like:

Just published in the peer-reviewed scientific journal, Environmental Health, is the bombshell that commercial HFCS appears to be routinely contaminated with mercury. It turns out the contamination isn’t so much accidental as newly recognized, given the fact that much HFCS has been made and continues to be made using “mercury-grade” caustic soda.

The full text of the peer-reviewed study Mercury from chlor-alkali plants: measured concentrations in food product sugar is available from Environmental Health, but it only contains the study on HFCS itself (not of processed foods). The writing style is too conversational for a scientific paper, but it is better than the IATP report. According to the abstract:

The [HFCS] samples were found to contain levels of mercury ranging from below a detection limit of 0.005 to 0.570 micrograms mercury per gram of high fructose corn syrup. Average daily consumption of high fructose corn syrup is about 50 grams per person in the United States.

Perhaps that consumption estimate is a little low. Let’s use the estimates reported in Not So Sweet: “American 19- to 30-year-olds consume about 60 grams of HFCS per day. For 12- to 18-year olds, HFCS consumption is about 70 grams”. Worst case scenario, a “heavy user” may consume 39.9 ug (0.0399 mg) per day (if all 70 g of HFCS were produced with mercury cells), according to this data.

Before we panic (or write condescending blog posts), we should know: how much mercury is in HFCS today, what form of mercury is it, how much mercury is in various foods, and how much of the mercury in food products is from HFCS compared to other ingredients?

The data in the Env. Health paper is from 2005. Why is it just now being published? The CRA says HFCS production methods have changed since this data was collected, so it would be irresponsible to make policy based on it. The authors said they were unable to secure HFCS from the sources as they did in 2005 – but couldn’t they get the samples from the food processors that buy the syrup? It feels like they just gave up (or that they knew a newer data set might prove their conclusions wrong).

The form of mercury matters because the different forms are absorbed into the body differently. According to the DoE Risk Assessment Information System’s page on mercury:

Gastrointestinal absorption of inorganic salts of mercury from food is <15% for mice and about 7% for humans (Goyer 1991). Organic mercury compounds (methyl- and phenylmercury) have been shown to be readily absorbed (>80%) by humans and animals following oral exposure (ATSDR 1989, Goyer 1991).

In other words, measuring the total mercury isn’t as useful as it seems. If the mercury in HFCS is the type that accumulates in fish, then we have cause to worry. If it is inorganic mercury, (as we would expect from  the mercury cell process) then the danger is minimized to a worst case scenario 0.0028 mg effective dose of mercury per day.

In Not So Sweet, the question of how much mercury ends up in food products that contain HFCS is answered (sort of). Their results are discussed by ChemRisk, “a leading scientific consulting firm” in a report they made at the behest of the CRA, along with a comparison of these values with other foods:

More than two-thirds of the samples analyzed by IATP had no detectable level of mercury at all. In the remaining sample, most of these were at or near the limit of detection. The average concentration for the 17 samples with detectable levels was only 128 parts per trillion (ppt). EPA sets limits for mercury in drinking water at two parts per billion.

It is well known that small amounts of mercury are broadly present in our environment. For example, Health Canada reported in 2003 that the concentrations of total mercury in steak ranged from 420 to 1,800 parts per trillion (ppt); fresh pork contained 1,100 to 1,500 ppt; organ meats (liver and kidney) contained over 2,100 ppt; and lamb contained 290 to 2,300 ppt of total mercury. (Dabeka et al, 2003) For the sake of reference, one part per trillion is equal to one drop of water spread out into 26 Olympic-size swimming pools. (Washington Suburban Sanitary Commission, 2009)

That same study by Health Canada looked at mercury in seafood, finding amounts that ranged from 40,000 ppt in fresh or frozen marine fish to 148,000 ppt in canned fish. Other foods, such as canned mushrooms, had 5,100 to 16,000 ppt total mercury, grapes had 180 to 590 ppt, blueberries 210 to 640 ppt, rice 570 to 1,800 ppt, raisins upwards of 700 ppt, and shelled seeds up to 1,000 parts per trillion (ppt).

Unfortunately, Dabeka, et al. isn’t available for free. The numbers reported by ChemRisk do match numbers I found elsewhere when researching this post.

Without controls in a properly designed experiment, we do not know if the mercury found in the items they tested is due to HFCS or if it is due to other ingredients. There are many ingredients that are common to a variety of processed foods. The ChemRisk report states:

IATP assumes that the total mercury they detected in a questionably small sampling of consumer foods is primarily the result of high fructose corn syrup; an assumption that has not been properly tested or validated. In fact, the authors do not attempt to characterize whether there may be mercury in any other ingredients contained within the consumer products tested, even while the recipes for the items studied may have had multiple sources of potential contamination.

Normally my suggestion for health and safety is simple: eat as little processed food and as few animal products as possible. Even that general message of moderation won’t work when it comes to mercury. Unfortunately, mercury is all around us. It would be nice to get kids to cut back on sweets, and it would be nice if the mercury cell HFCS refining process was changed, but the real problem is elsewhere. I have to question the ethics of any organization the leads us on a wild goose chase.

Coal fired power plants are the single largest emitter of mercury into the atmosphere. If you really care about children ingesting mercury in their food, write letters to your congressmen demanding that they act to reduce mercury emissions from existing plants (the technology exists) and to prevent new coal fired plants from being built. Encouraging China to do the same is another matter entirely.

One upcoming source of mercury in the environment is CFC light bulbs. They won’t be anywhere near the level of pollution from coal fired plants, but we should be conscious of the mercury in the bulbs. According to EcoGeek, some places are now offering recycling. Contact your city leaders and ask for CFC recycling in your area.

Renee Dufault, Blaise LeBlanc, Roseanne Schnoll, Charles Cornett, Laura Schweitzer, Lyn Patrick, Jane Hightower, David Wallinga, Walter Lukiw (2009). Mercury from chlor-alkali plants: measured concentrations in food product sugar Environmental Health, 8 (1) DOI: 10.1186/1476-069X-8-2


  1. Thanks for commenting, Renee. I hope I’ve done a reasonably good job of presenting the facts (particularly those that have been left out by other sources). My main goal here is to convince people to follow the links and learn about things for themselves rather than accepting everything verbatim.

    Regarding your comment, there were 8 other authors on that paper. Why couldn’t one of them, or one of their grad students, continue the work? If there was a credible public health threat, someone would be willing to do the work. I can’t imagine that the tests for mercury are all that expensive, assuming you know someone with the right spectrophotometer.

    With all due respect, it seems irresponsible to present work from 2005 knowing that people are going to assume it applies in 2008. The abstract doesn’t mention when the samples were obtained, and we all know that the majority of people only read abstracts. Of course, none of the bloggers, NGOs, or media reporting on your paper mention when the samples were collected.

    I have to ask – did you know how your work would be presented by the IATP? Your paper, as far as I can tell, is sound (although I wasn’t so excited about the the discussion and conclusion where HFCS was singled out for additional testing over other ingredients). The problem is how the IATP sensationalized everything.

    If I was you, and didn’t know in advance about the IATP, I’d be quite angry at them for misusing my results. Any thoughts about a press release ala“ rel=”nofollow”>Barney Gordon?

  2. Wow, that was a pretty fast review of the mercury claim, Anastasia. The thing that I noticed that stuck way out was the lack of controls. The mercury was assumed to come from the HFCS and nothing else. No non-HFCS foods were tested. And the milk products were way higher… how would normal standard (or even raw and organic) milk fare under the same examination?

    All that could be reliably concluded from the Not So Sweet study is that there were part-per-trillion levels of mercury in 30% of the foods studied. Assigning sources takes something a bit more scientific.

Comments are closed.