Controls are Sweet!

In science you always.


Always use controls.

That is the very basis of science, for without a control running next to your experiment, you have no isolated variables, no conclusions that can be drawn from it, and no theories that it can support.

So when I was reading the Ethicurean, as I regularly do, I was simply flabbergasted at this post: Mercury in HFCS. Apparently, a research paper came out proclaiming that high fructose corn syrup (HFCS)-containing products had detectable levels of mercury. The explanation given was that HFCS is made using alkali soda, from plants that use mercury in the process of synthesizing it. (Except this has been for the most part phased out)

I took a look at the paper, and the first thing that I noticed was that it was not a peer-reviewed study. So this has not passed through the rigors of experimentation, review, re-testing if needed, and publication in a scientific journal.

Anastasia already tore into the paper, beating me to the punch.

The second thing I noticed was that the paper was mostly an argument about how unhealthy the American Diet is, and the big calorie baddie is HFCS.

We live in a truly global food system. Our system typically is geared more toward producing lots of cheap calories, and then selling those calories to consumers, than it is toward meeting other goals like reducing fossil fuel use or producing food that is healthy.

Such a rapid transformation in the American diet raises important questions: What are the potential health impacts of HFCS consumption? What exactly is HFCS and where does it come from? And what additional risks to consumers may stem from the industrialized processes by which HFCS is made and used?

The ‘study’ itself consisted of taking samples of foods that contain high-fructose corn syrup and testing them for levels of mercury. Those foods and the brands that made them were in the report, and they found that some of the foods had detectable levels of mercury in them. What levels? Parts per trillion. These are really low levels. Drinking water has a limit of 2 parts per billion, which means that you can have 100 times as much mercury in drinking water as is in these foods. The tap water you use to make your oatmeal might have more mercury than the oatmeal itself.

Their conclusion was that the mercury came from HFCS, and made a list of recommendations about what should be done about this sciency conclusion. But wait, did they actually prove that the mercury came from the corn syrup?

No. In order to demonstrate this scientifically, you have to have controls. There’s no telling where the mercury came from without isolating the variables. Were the oats sucking up mercury in the soil? Were the cows that produced the chocolate milk they tested the actual source of the highest mercury levels they found in their survey?

Of course, our survey was just a snapshot in time; we tested only one sample of each product. That is clearly not sufficient grounds to give definitive advice to consumers on specific products.

In other words, our efforts were never intended to take the place of full-scale safety testing by the FDA.

At least they were honest about that part. Except on the next page they proceeded to give advice based on this ‘study.’

Marion Nestle was right on the money with her assessment, dismissing it not only as uncontrolled semi-science, but also irrelevent considering the ever-present nature of mercury in the environment.

This seems like quite sensible advice, but how worried should we be about mercury in HFCS? I agree that mercury in any form is unlikely to be good, but I have no idea whether such low levels do measurable harm.  For one thing, these studies did not compare the amounts of mercury found in HFCS to those typically found in foods that do not contain HFCS.  My guess is that most foods contain low levels of mercury because mercury is prevalent in air, water, and soil, especially around coal-burning power plants.  Also, soft drinks are the major sources of HFCS in American diets, but these were found to be relatively free of mercury.  This is puzzling.

The fact that the foods with the highest amounts of HFCS had zero detectable mercury suggests that there is not a positive correlation between HFCS and mercury.

Look, folks, if we are to make any rational progress on food safety, nutrition, and on the health of the environment, we need to base our decisions on solid scientific research, conducted properly, independently, with basics like replicated trials and control groups to compare to your experimental groups, and finally, the research must pass peer review. To really have an impact, it should be replicated by other laboratories and make predictions that can further be tested by new research.

Otherwise, it’s Science By Press Release.

For the record, I’m a big fan of Green Chemistry – changing industrial chemical processes to be more environmentally-friendly. If there are chlor-alkali plants that still use mercury somewhere, they should change their methods of synthesizing it, at least because it will reduce demand for mercury, and eliminate the concomitant emissions. But instant sweetened Oatmeal is not the right rallying cry. It reminds me of when Steven Milloy claimed that Fluorescent Replacement lights were toxic superfund-sites-in-the-home waiting to happen. Same deal, different politics.

And I’m with Anastasia and Marion that focusing on Coal power pollution would be a more effective means of reducing our exposure to mercury. There is no such thing as clean coal.


4 thoughts on “Controls are Sweet!”

  1. But aren’t you just attacking a stunt study without even addressing the original study it was calling attention to? The first study, published in Environmental Health, was peer reviewed, and so presumably was a controlled rigorous scientific study. The “study” you’re arguing against is no more than a press event by an anti-HFCS group done to insure the first true study didn’t go unnoticed.

    There’s plenty of articles on this, making the distinction clear. Here’s but one:


  2. Thanks for both of your comments. I find it interesting that you both sort of agree and disagree with me and each other. I’ll put up another post discussing this in the context of the Dufault et al. study.


Comments are closed.