Let’s talk about those industry funded studies. You know the ones. The ones you hear about from anti-GMO folks in the comment section of any story about GMOs. According to those folks, the whole scientific consensus on GMO safety is based on industry funded studies. They say that the only studies that show that GMOs pose no different risks than conventionally bred crops were all bought and paid for by Monsanto. That makes the consensus suspicious right? It would if there weren’t many independently funded studies with similar results.
Complaints about industry funded studies show an ignorance of the literature and may indicate a lazy desire to dismiss inconvenient evidence in order to preserve predetermined ideological commitments. It’s plain old confirmation bias and motivated reasoning run amok.
Monsanto is a medium sized company ($57.43B). Is it really possible that they’ve manipulated tens of thousands of scientists performing thousands of studies for three decades with no whistleblowers? Could Monsanto’s power have resulted in a scientific consensus that has been bent completely to their will? In comparison, fossil fuel behemoths Exxon Mobil ($394.83B), Chevron ($215.45B) and BP ($150.07B) (total: $760.35B) have been completely stymied in their efforts to buy a scientific consensus on climate change. Let’s put aside the fact that this line of thinking just doesn’t make sense. Instead, let’s take a look at the evidence and unravel some of the pretzeled logic often employed to dismiss the weight of that evidence in support of the scientific consensus on GMOs.
Let’s start with the European Union (EU). Politicians in the EU are generally not friendly to GMOs and they wanted to be very careful about them. So they ponied up €200 million over a decade to look into the matter. The resulting studies are neatly summarized in A Decade of EU Funded GMO Research [pdf].
This new publication presents the results of 50 projects, involving more than 400 research groups and representing European research grants of some EUR 200 million. This figure brings the total Commission funding of research on GMO safety to more than EUR 300 million since its inception in 1982 in the Biomolecular Engineering programme.
…The main conclusion to be drawn from the efforts of more than 130 research projects, covering a period of more than 25 years of research, and involving more than 500 independent research groups, is that biotechnology, and in particular GMOs, are not per se more risky than e.g. conventional plant breeding technologies.
In addition to the EU studies, scientists around the world have been building on our knowledge of GMOs. Biofortified’s GENERA project (in progress) has amassed a list of over 1000 studies on the safety of biotechnology, about 1/3 of which have independent funding. Not all of the studies are supportive of the position that GMOs are no riskier than their conventionally bred counterparts, but the vast majority support that proposition.
Let’s look at a type of papers that are of particular value to non-scientists: literature reviews (video) and meta-analyses are a great way for getting a sense of the weight of the evidence on a given topic. They help us avoid single study syndrome and keep us from missing the forest for the trees. Here are four of these type of papers from the Biofortified list of studies with independent funding.
- Assessment of the health impact of GM plant diets in long-term and multigenerational animal feeding trials: A literature review
The aim of this systematic review was to collect data concerning the effects of diets containing GM maize, potato, soybean, rice, or triticale on animal health. We examined 12 long-term studies (of more than 90 days, up to 2 years in duration) and 12 multigenerational studies (from 2 to 5 generations). We referenced the 90-day studies on GM feed for which long-term or multigenerational study data were available. Many parameters have been examined using biochemical analyses, histological examination of specific organs, hematology and the detection of transgenic DNA. The statistical findings and methods have been considered from each study. Results from all the 24 studies do not suggest any health hazards and, in general, there were no statistically significant differences within parameters observed.
Honey bees (Apis mellifera L.) are the most important pollinators of many agricultural crops worldwide and are a key test species used in the tiered safety assessment of genetically engineered insect-resistant crops. There is concern that widespread planting of these transgenic crops could harm honey bee populations. We conducted a meta-analysis of 25 studies that independently assessed potential effects of Bt Cry proteins on honey bee survival (or mortality). Our results show that Bacillus thuringiensis (Bt) Cry proteins used in genetically modified crops commercialized for control of lepidopteran and coleopteran pests do not negatively affect the survival of either honey bee larvae or adults in laboratory settings. Although the additional stresses that honey bees face in the field could, in principle, modify their susceptibility to Cry proteins or lead to indirect effects, our findings support safety assessments that have not detected any direct negative effects of Bt crops for this vital insect pollinator.
Although scores of experiments have examined the ecological consequences of transgenic Bt crops, debates continue regarding the nontarget impacts of this technology. Quantitative reviews of existing studies are crucial for better gauging risks and improving future risk assessments. To encourage evidence-based risk analyses, we constructed a searchable database for nontarget effects of Bt crops. A meta-analysis of 42 field experiments indicates that nontarget invertebrates are generally more abundant in Bt cotton and Bt maize fields than in nontransgenic fields managed with insecticides. However, in comparison with insecticide-free control fields, certain nontarget taxa are less abundant in Bt fields.
There is one more literature review from the Biofortified list that I want to look at, but in the context of making an important point. There are lots of industry funded studies. The majority in fact. But there are also many independent studies. How can we judge if the industry funded studies are reliable? If the independent studies and the industry studies come to the same conclusions, then we can conclude that the industry studies are reliable. And that is exactly what we find.
Let’s parse out the findings of a 2007 literature review on human and animal toxicological/health risks studies for GM foods/plants and the 2011 follow up.
In is 2007 paper Toxicity studies of genetically modified plants: a review of the published literature, José Domingo found little to no evidence that genetically engineered crops posed significantly different risks than conventional crops, while sounding several cautious caveats and underlining that the body of literature seemed too scant for drawing confident conclusions. One statement of the abstract caught me eye. “Moreover, most published studies were not performed by the biotechnology companies that produce these products.”
The same author followed up in 2011 with a colleague and found a much larger number of studies, but also a shift in the proportion of industry funded studies. This wasn’t surprising since prior to 2006, companies hadn’t been publishing their test results (aside from in their patent applications) but an industry-wide push for transparency had changed that.
The main goal of the present review was to assess the current state-of-the-art regarding the potential adverse effects/safety assessment of GM plants for human consumption. The number of citations found in databases (PubMed and Scopus) has dramatically increased since 2006. However, new information on products such as potatoes, cucumber, peas or tomatoes, among others was not available. Corn/maize, rice, and soybeans were included in the present review. An equilibrium in the number research groups suggesting, on the basis of their studies, that a number of varieties of GM products (mainly maize and soybeans) are as safe and nutritious as the respective conventional non-GM plant, and those raising still serious concerns, was currently observed. Nevertheless, it should be noted that most of these studies have been conducted by biotechnology companies responsible of commercializing these GM plants. These findings suggest a notable advance in comparison with the lack of studies published in recent years in scientific journals by those companies.
José Domingo and Jordi Giné Bordonaba are certainly no cheerleaders for biotech crops. Yet, despite the increase in industry funded studies between their reviews of the literature in 2007 and 2011, they still find plenty of evidence to affirm their cautious stance towards the technology.
An even more robust review of the total literature published in 2014 is more conclusive in their findings: “The scientific research conducted so far has not detected any significant hazards directly connected with the use of genetically engineered crops.” In An overview of the last 10 years of genetically engineered crop safety research, the authors collected and evaluated 1,783 research papers, reviews, relevant opinions, and reports published between 2002 and 2012, a process that took over 12 months to complete. The review was published in Critical Reviews in Biotechnology and covered all aspects of GM crop safety, from how the crops interact with the environment to how they could potentially affect the humans and animals who consume them. And their conclusion?
The lead author, Alessandro Nicolia, an applied biologist at the University of Perugia in Italy, said in an interview: “Our goal was to create a single document where interested people of all levels of expertise can get an overview on what has been done by scientists regarding GE crop safety. We tried to give a balanced view informing about what has been debated, the conclusions reached so far, and emerging issues.”
Looking at the scientific literature about GMO safety, we find little difference between the results of independent and industry funded studies. What if we were a little more rigorous in our scrutiny? Johan Diels of led a team that did exactly that. The results were interesting, but not without some problems.
In a study involving 94 articles selected through objective criteria, it was found that the existence of either financial or professional conflict of interest was associated to study outcomes that cast genetically modified products in a favorable light (p = 0.005). While financial conflict of interest alone did not correlate with research results (p = 0.631), a strong association was found between author affiliation to industry (professional conflict of interest) and study outcome (p < 0.001).
No association was found between ﬁnancial conflict of interest (COI) and article outcome. The authors did find a correlation between “industry affiliation” and favorable study outcome. But realize how far we have moved the goal posts. We started with the proposition that we couldn’t trust any of the research because it was it was all paid for by the industry. But we found that’s not true. Now we have researchers looking into the matter and they can’t find a relationship between industry funding and favorable study outcomes. And there’s no connection there. What’s left is griping about the industry ties of some of the researchers. Before looking a little closer at that, let’s get one thing out of the way. When a company pays for a study, they are paying because they want to find out something. Fudging the data does not help them in their business. Such data manipulation would be generally be counter productive. That’s why it didn’t show up in the data.
Now let’s focus on the association between professional COI and study outcome. There are a few things that might explain that correlation besides a lack of independence. Half the sample was undeclared regarding COIs, leaving a very small sample to examine. In the professional COI category nearly 10% of the sample size was categorized as undetermined. Moreover, while 43 authors had COIs, 28 of the studies were compositional studies. These are nearly always funded by the companies and they are close to always favorable since finding substantial equivalence is very likely. Remove those 28 studies from the set of 43 with financial or professional COIs and the P values will shift towards insignificance.
Another thing to keep in mind, especially where compositional studies are concerned is that the company has already performed in-house studies. They are contracting independent scientists to confirm their findings. This is going to skew the results of the sample towards industry favorable study outcomes. This doesn’t mean the studies were suspect. They were just more likely to result in a favorable outcome to begin with. If the in-house study had an unfavorable outcome in compositional assessment or other tests, then that project would be stopped and it’s back to the drawing board for a new project. There is no need for follow up testing by outside independent researchers. That’s a big reason why so many studies in that sample will produce favorable results.
In written conversation, independent researcher and professor of horticultural sciences, Kevin Folta put it this way:
I think the other factor is that industry recruits independent experts to independently reproduce findings. They show in house that x+y=z. They then hand the test to a university, that finds x+y=z. If the test fails in-house, then it does not go for independent verification. That will skew statistics too, because the outcomes of the university-based tests have already been demonstrated. The reason the results frequently agree is because they are frequently correct.
On his blog, Folta looked at the question of industry funding at the University of Florida:
First, I went to an easy source at my university, the University of Florida. The Institute for Food and Agricultural Sciences (IFAS) publishes their financials every year. You can find this online here.
How much Big Corporation money did we spend? Not that much. It is buried somewhere in that “other sponsored funds” piece of the money pie.
Now wait, I can hear critics already screaming that “other sponsored funds” is almost 10% of the research dollars spent, and that’s a significant amount at a place like the University of Florida. So let’s use the record to break that down:
Yikes. Corporate sponsorship is a pretty small sliver of that pie. So about two percent of our funds come from corporate interests. For the anti-scientific critics out there, that’s about two dollars out of every hundred.
If we are bought and paid for, we’re bought really cheap and not paid well.
Elsewhere in that conversation, Folta said,
The other important thing to remember is that almost no university researcher is going to commit career suicide by fudging data, especially for some damn company. That’s why companies come here. If the results don’t agree with what they found, it means that they are not reproducible. That’s the answer they NEED to know, and why they go independent in the first place!
…Even if a university gets a building, how does that make an individual researcher change research, essentially torpedo a career? Even if a company finances a lab (which is rare) they want the real answer, not some fabrication. To suggest that we’re all somehow sellouts is insane. Show me where my data do not hold up.
I’m not saying that people shouldn’t pay attention to potential conflicts of interest or that we should take industry funded research at face value. What I am saying is that dismissing an entire body of research because it’s supposedly bought and paid for is foolish and lazy.
When you shout ‘Conflict of Interest’ before evaluating the evidence and analysis, it becomes an excuse for discounting inconvenient evidence. Asking about conflicts of interest should be a safeguard against getting snookered by weak evidence. Instead, it becomes an excuse for dismissing good evidence. Examining the soundness of the evidence must come first. Then you can decide whether questions of funding and loyalties are relevant. This is how you maintain a firm footing on solid ground. Use awareness of conflicts of interest to avoid motivated reasoning. Otherwise you are only fueling the fire of your own biases.
A previous version of this article appeared on REALFOOD.ORG on 13 February 2014.