Jul 31 2015

GMOs and Making Up Your Own Science

Dedicated anti-science groups engage in a number of methods to maintain their propaganda upstream against the scientific evidence. It’s actually not difficult- people are generally very good at motivated reasoning. We can demonize or lionize anything.

Methods include dismissing scientific studies whose conclusions you don’t like, supporting low quality studies you do like, misinterpreting and distorting other studies, and of course cherry picking. Sometimes, however, dedicated activists seem to literally make up studies out of whole cloth, or ideological scientists perform dubious studies to create fodder for their side.

This week on the SGU we interview Kevin Folta (the show will be published tomorrow) about some of his experiences with anti-GMO activists who have no problem making up the science to advance their ideological agenda. The more I look into anti-GMO activism the more I realize that the anti-vaccine movement has nothing on them when it comes to pseudoscience. Their methods are identical. The only real difference is that anti-GMO propaganda is much more mainstream.

The latest episode concerns Dr. V.A. Shiva Ayyadurai, who recently published a study claiming that there are high levels of formaldehyde in GMO soy. I did a Goggle search for “GMO Soy Formaldehyde” and the first hit was this: GMO Soy Accumulates Carcinogenic Formaldehyde: Game-Changing Study. The anti-GMO crowd is making a lot of hay with their latest “proof” that all GMOs are evil.

Let’s take a closer look at Ayyadurai and his study. Kavin Senapathy at the Genetic Literacy Project has some more information. She notes that Ayyadurai has four degrees from MIT, but none in food science or genetics. This is not a quibble – it is very common for ideological scientists to be operating outside of their area of expertise.

Much more importantly, however, it turns out that Ayyadurai’s study is not an analysis of GMO soy, it is a computer model. That’s it. He made a computer model that predicts that GMO soy should accumulate formaldehyde using what is called “systems biology.”

Kevin explains on his own blog that such computer models are only as good as the inputs. If the assumed data is not robust and sufficient, computer models based on that data are worthless. It is also easy to manipulate the inputs to create a desired result.

In other words, Ayyadurai’s study is utter crap. Kevin writes:

“The bottom line is, corn is probably the most biochemically dissected plants in terms of composition. Soy too. There is no evidence ever published or otherwise reported in a legit place that shows a difference in formaldehyde between GM and non-GM varieties of anything. These authors could have tested their prediction, and maybe they did, but there is no evidence of formaldehyde ever reported.”

They never tested their predictions, which means their model has not been validated in any way. Further, the predictions made by their model go against what scientific evidence we already have.

I should further note that formaldehyde, while scary sounding, is a natural byproduct of metabolism and exists in many foods. Here is a table listing the amount of formaldehyde in common foods. It’s nothing to worry about – go ahead and eat those shitake mushrooms.

The paper itself is poor science. The authors do not spell out details critical to the analysis. Further, there are conflicts of interest galore (anti-science groups only seem to mention those when they apply to the government or corporations they don’t like). As Kevin reports:

First, what do we know about International Center for Integrative Systems?  Certainly sounds impressive!  Turns out it is a 501(c)3 non-profit organization, surprise, started by the lead author.  They have a variety of projects, including one that seeks to define standards for raw and organic food.  A little poking around the website and the agenda is showing.

Conflicts of interest are OK, apparently, if the interest is yours. We see this among anti-vaxxers as well, who seem totally unbothered by Andrew Wakefield’s blatant conflicts.

Kevin has offered to actually test the levels of formaldehyde in GMO soy. This would be another test of the computer model. Ayyadurai so far has refused to participate – he seems completely uninterested in testing his model directly.


Good skeptics should be concerned first and foremost with method. Skepticism is not a set of beliefs (or non-beliefs), it is dedication to using methods that are fair, valid, logical, and evidence-based when evaluating any empirical claim.

I understand that some people have legitimate claims about the food industry and Big Agriculture. Despite the fact that I commonly address anti-corporate pseudoscience, I do not consider myself a defender of corporations. I am trying to defend good science and skeptical reasoning. In fact I think that corporations generally have highly effective and well-funded departments dedicated to motivated reasoning and lobbying the government toward their interests. We need effective watchdogs and we need to call industries and corporations on their BS.

That is why pseudoscientific campaigns cause even more harm – they actual hamper effective industry watchdogs by tainting the enterprise with pseudoscience and conspiracy theories. They suck all the air out of the room, making it more difficult to have a good conversation about effective regulations, because we have to spend our time countering rank nonsense and misinformation.

When it comes to the issue of GMO I feel like we don’t yet have our head above water. Greenpeace, the massive organic lobby, and other anti-GMO groups have dominated the propaganda for too long, and they have effectively confused the public over GMO issues.

Skeptics, however, are leading the way toward change. As Kevin pointed out in the interview for this week, 10 years ago he would be the lone voice of reason in the comments section of an article on GMOs. Now an army of skeptical nerds will often get there before he does, effectively making the points that need to be made.

But we are not there yet.

10 responses so far