Last week, Reuters reported on an alleged conspiracy to cover up data showing that there is no relationship between glyphosate (the main ingredient in Monsanto’s Roundup weedkiller) and cancer after all. The piece triggered a frenzy of media coverage and fresh attacks on the credibility of World Health Organization’s classification of glyphosate’s cancer risk. However, after the original documents that the article relied on surfaced, it was clear that the sensational headline doesn’t tell the whole story.

The World Health Organization made the news in March 2015 when its International Agency for Research on Cancer (IARC) classified glyphosate as “probably carcinogenic to humans.” The IARC found convincing evidence of glyphosate’s ability to cause cancer in lab animals and limited evidence that it causes cancer in people. That’s because it takes decades of observation to see cancer outcomes among pesticide users, so those studies are more rare than those involving lab animals.

Monsanto and its industry allies responded by launching an all-out attack on the IARC, questioning the integrity of the agency, even harassing the scientists who contributed to the assessment. Monsanto went so far as to hire “independent” scientists to conduct a separate cancer review while ghostwriting some of the studies themselves. All this effort to undermine the IARC classification isn’t just a matter of public relations. The company is currently battling lawsuits by individuals who have developed non-Hodgkin’s lymphoma (NHL) after using Roundup.

The Reuters article begins by saying that Aaron Blair, a retired scientist, had “something he wasn’t telling” the IARC while serving on its glyphosate assessment panel. At the time, Blair was also contributing to the U.S. National Cancer Institute’s Agricultural Health Study (AHS), a comprehensive study of tens of thousands of pesticide users and their families over decades. The AHS data did not suggest a correlation between glyphosate use and NHL, and Blair testified that this data might have influenced the IARC assessment had it been included (Reuters cites Blair’s deposition given in March as its source, although it did not provide a copy of the transcript).

Later in the article, the author concedes that the IARC could not have incorporated the AHS data into its assessment since the agency only includes published data in its assessments. Nevertheless, the Reuters article takes aim at Blair and at the IARC itself, quoting Monsanto as saying that the IARC “ignored” crucial data and that the National Cancer Institute decided to withhold the data from publication because of its findings. Other media outlets and industry groups have since jumped on the bandwagon to repeat these claims.

Reuters did not share Blair’s deposition, but veteran reporter Carey Gillam obtained it and made it available online. It shows that the Reuters narrative is sloppy journalism at best, and a deliberate mischaracterization of Blair’s testimony at worst.

The deposition documents a Monsanto lawyer drilling Blair for hours as to why he did not provide the data to the IARC, and why the study was not published in time to be included in the IARC assessment. Blair repeats ad nauseum that the data could not be used by the IARC because it was not published, and that there was no deliberate delay in publishing; rather, “scientific research takes time” and it is normal for studies to be delayed during the writing and review processes. Overall, Blair’s testimony provides no evidence that he tried to mislead the IARC, nor that the AHS was delayed for any nefarious reasons.

What Reuters left out is even more telling. For example, Blair says in the deposition that the AHS lack of correlation between glyphosate and NHL is not statistically significant. Additionally, the plaintiff’s lawyer questions Blair about another pending study that he is coauthoring using data on glyphosate users and cancer outcomes. The data suggest a statistically significant correlation between any amount of glyphosate use and NHL. Moreover, individuals in that data set who used glyphosate three or more times a year are over two and a half times more likely to develop NHL than those who did not.

Blair also “withheld” this study’s findings from the IARC because it was unpublished data. Yet the Reuters article is silent about it. Here we have a statistically significant  correlation between any  amount of glyphosate use and cancer, and Reuters (and later Monsanto) is hammering down on the dataset that finds no such correlation but is also not statistically significant. Once again, Monsanto is trying to manipulate the narrative on glyphosate’s cancer risks.

It is impossible to say how either study might have impacted the IARC’s cancer classification of glyphosate. New data will continue to emerge. For now however, the IARC’s cancer assessment remains the gold standard because it only considers independent, published data – not industry studies that are more likely to find results that are favorable to Monsanto.