ATSDR Report Confirms Glyphosate Cancer Risks

EPA Official Once Tried to Kill It, But the Nearly-Dead ATSDR Report Returns to Haunt Monsanto and Bayer

This week a public health agency of the U.S. Department of Health and Human Services (DHHS), the Agency for Toxic Substances and Disease Registry (ATSDR), released the long-awaited Draft Toxicological Profile for Glyphosate. And, it supports and strengthens the 2015 cancer assessment of another health agency, the International Agency for Research on Cancer (IARC).

For many years, glyphosate—more widely known under its trade name, Roundup—has been the world’s most widely used herbicide. GMO products from Monsanto—now owned by Bayer—like Roundup-Ready corn, soybeans, wheat, and cotton are specifically designed to be grown using glyphosate-based herbicides, which has driven its use and sales to about 300 million pounds annually in the US, with about 90% of that on farm fields, and about 10% on non-agriculture uses like lawns, gardens, golf courses, and parks and playgrounds.

According to documents made public during court proceedings, the EPA Pesticide Office worked with Monsanto to hold back this ATSDR report for several years while it promoted its “no cancer risk” position.

In April 2015, now-retired EPA Pesticide Office official, Jess Rowland told Monsanto’s Dan Jenkins, “If I can kill this I should get a medal”, according to internal Monsanto emails that have now been made public. Jenkins subsequently emailed his Monsanto colleagues: “I doubt EPA and Jess can kill this; but it’s good to know they are going to actually make the effort now to coordinate due to our pressing and their shared concern that ATSDR is consistent in its conclusions w EPA.”

Others in the EPA Pesticide Office were also helpful, according to a courthouse news report. EPA official Mary Manibusan was asked by Monsanto if she knew anyone at ATSDR working on chemical tox profiles. “Sweetheart, I know lots of people so you can count on me,” Manibusan told Monsanto's Jenkins, who said, “We’re trying to do everything we can from having a domestic IARC occur with this group. May need your help,”. That exchange was in June 2015, just three months after the IARC meeting on glyphosate.

Manibusan's LinkedIn shows a long stint with EPA, followed by two years with industry consultant Exponent (2015-2017), then in the Trump Administration she is back in government, first FDA (Dec 2017-Mar 2019), and now recently returned to EPA as a Division Director (Apr 2019-present). On LinkedIn, Manibusen has endorsed Rowland for his skills in "government" and "environmental awareness", paths from which both seemed to have strayed.

Unfortunately, ATSDR did temporarily back-burner its work on glyphosate, agreeing instead to pay close attention to the EPA’s 2015 cancer assessment, according to court documents (see US RTK Monsanto Papers).

Now, almost exactly two years later, the ATSDR’s Draft Toxicological Profile for Glyphosate has risen from the near-dead and is out for public comment. At an impressive 257 pages, the fully-referenced report comprehensively covers both the cancer and non-cancer health effects linked to glyphosate exposure, the latter including adverse effects on reproduction and child development, and organ toxicity. Bravo, ATSDR!

A pattern is emerging: non-industry experts (Zhang et al 2019) and health agencies IARC and ATSDR are finding a link with glyphosate and cancer; whereas, regulatory agencies are lining up with Monsanto and Bayer that it does not cause cancer, even when reviewing the same scientific evidence (see ATSDR Fig 2-13). 

More on the regulatory loopholes and failures in my companion blog, "Regulatory Failures = Superweeds and Glyphosate Cancers"

Glaring Cancer Risks

The most important thing to realize about the ATSDR report is how clearly it lays out the vast array of scientific evidence linking both pure glyphosate (rodent studies) as well as formulated glyphosate-containing products (in human epidemiologic studies) like Roundup as they are sold on the shelf to cancer. Here are the major takeaway points:

  • The association with non-Hodgkin Lymphoma (NHL) cancer risk is stronger when the study is adjusted for more days of glyphosate use, longer study latency period (time since exposure for cancer to develop), which strengthens the confidence in the results;
  • All three meta-analyses (evaluation of many studies together) show a statistically significant link with NHL cancer (positive confidence intervals), with narrow confidence intervals which strengthens our confidence in the link with cancer. This is true even of the Monsanto-sponsored analysis (Chang and Delzell, 2016), which is almost exactly the same result as the conclusions of the World Health Organization’s global cancer experts (IARC 2016).
  • Most studies are strongly positive for cancer (non-Hodgkin’s Lymphoma), even if they are not statistically significant at a 95% confidence level. All would likely show a statistically significant link with cancer at a 90% confidence level.

Since the ATSDR analysis, a more recent meta-analysis that includes the most recent update of the Agricultural Health Study (AHS) along with five case-control studies reported a statistically significant 45% (95% CI: 1.11–1.91) increase in risk of NHL cancers people most highly exposed to glyphosate-based herbicide products (Zhang et al 2019). (See image below, ATSDR Fig 2-4)

For Multiple Myeloma, ATSDR reports similar findings across available studies (See ATSDR, Fig 2-5, p 87).

Given ATSDR’s lucid and comprehensive review of the available epidemiology information, which evaluates the glyphosate-based products that people are exposed to under real-world conditions, it seems to me that anyone not affiliated with the chemical industry or defending the chemical industry would reasonably conclude that these products are linked to cancer.

One could argue how strong the link is, under what exposure conditions, or other aspects, but anyone should be able to see that there is a link. Yet, regulatory agencies in the US and worldwide all seem to have adopted Monsanto’s position that there is no link, largely by dismissing studies that don’t meet the 95% confidence standard—a threshold that is not required by law and is rejected by experts: "It is ludicrous to conclude that the statistically non-significant results showed ‘no association’, when the interval estimate included serious risk increases"  (See recent article by over 800 signatories published in Nature 2019). In fact, the only two government agencies that seem to be able to provide an accurate scientific assessment of Monsanto’s carcinogenic products are the non-regulatory health agencies, ATSDR and IARC.

In response to the ATSDR report, a recently-retired EPA official from the Pesticide Office, Tina Levine, was quoted as saying, “There does seem to be an effect. It’s not a large risk [of NHL] but it’s there based on what ATSDR has found… That said, consumer exposures are likely to be very different from professional applicators”. Her comments are consistent with the conclusions of the EPA expert Scientific Advisory Panel that rejected EPA’s “not likely to be carcinogenic” classification, with some Panelists favoring the stronger classification of “suggestive evidence of carcinogenic potential” (see SAP report, p. 22).

So, how come the EPA Pesticide Office had all the same studies as IARC and ATSDR, but couldn’t find the cancer risk? My previous blog details the split among EPA experts—with the Science Office finding cancer risk, and the Pesticide Office finding none. In addition to many other failings, EPA and other regulatory agencies state that they are only focused on exposures to the general population from food, and specifically are not accounting for routine occupational exposures to pesticide applicators and others (despite the fact that the EPA Pesticide Office is charged with protecting everyone, including farmworkers and pesticide applicators). 

Non-cancer effects of glyphosate and formulated products

ATSDR identifies scientific evidence in rodent studies and some human epidemiologic studies linking glyphosate with developmental delays, gastrointestinal effects including nausea and vomiting, kidney and liver toxicity, and eye irritation (ATSDR summary data p. 3-5).

Long-term ongoing health studies conducted by the U.S. National Cancer Institute of over 20 thousand pesticide applicators report that human exposures to glyphosate-containing products is linked to an elevated risk of wheezing, chronic bronchitis, and allergic asthma (ATSDR, Table 2-5, p. 36; Ag Health Study)

Developmental risks reported in studies of farm families identify a link between parental use of glyphosate and an increased risk of neural tube defects, miscarriage, preterm delivery, and small for gestational age (ATSDR Table 2-5, p. 40-41). Although many of these are not statistically significant at a 95% confidence level, they are significant at a 90% confidence level, which should be confident enough to generate regulatory action to protect families wishing to conceive.

ATSDR summarizes a number of animal studies of early-life exposure to glyphosate-based products that report developmental effects including testicular lesions, decreased sperm production, elevated abnormal sperm, decreased testosterone, and skeletal malformations (ATSDR, p. 14).

ATSDR report likely underestimates risks

ATSDR set a Minimum Risk Level (MRL) of 1 mg/kg-day based on a second-hand report from EPA of a Monsanto-sponsored rabbit study that reported gastrointestinal effects following exposure to pure glyphosate. ATSDR used a standard 100-fold uncertainty factor to extrapolate from an animal study to humans, and to account for inter-human variability (for details, see ATSDR Appendix A).

Despite a mountain of good work, the ATSDR report will almost surely underestimate health risks for a number of reasons, including but not limited to:

Raw data and original studies not provided to ATSDR - Importantly, ATSDR didn’t have access to the animal studies (ATSDR, p. 12). Instead, it unfortunately had to rely on EPA Pesticide Office summaries (called Data Evaluation Records or Reports, DERs) of Monsanto-sponsored studies. That’s two steps removed from having access to the raw data, and two layers of unreliability in my opinion. It is also a reason why the epidemiology data is so important to take seriously.

Formulated Products More Toxic than Pure Glyphosate - Ongoing cell-based toxicity tests of the U.S. National Toxicology Program show that the glyphosate-based formulated products including lawn and garden products are more harmful than pure glyphosate, some by over 100-fold (NTP 2018).

Studies in young rodents conducted by the Ramazzini Institute reported that early-life exposure to both pure glyphosate and formulated Roundup mixtures resulted in modifying the gut microbiome in ways that may impair gastrointestinal health or have other long-term adverse health impacts (Mao et al 2018).

However, most of the rodent studies evaluated by ATSDR were on the pure glyphosate, which ignores all the chemicals in the formulated products like Roundup that people are exposed to in the real world, and that are likely to be more toxic. Again, a good reason to take the epidemiology studies more seriously, since they are real-world exposures.

Let ATSDR know what you think: ATSDR is soliciting public comments on its draft Toxicological Profile for glyphosate. Comments may be submitted via Regulations.gov until July 8, 2019 (Docket ID: ATSDR-2019-0001).

About the Authors

Jennifer Sass

Senior Scientist, Federal Toxics, Health and Food, Healthy People & Thriving Communities Program

Join Us

When you sign up you'll become a member of NRDC's Activist Network. We will keep you informed with the latest alerts and progress reports.