Adam Shriver is Director of Wellness and Nutrition at the Harkin Institute for Public Policy and Citizen Engagement. This essay was first published on his Substack newsletter, Canary in a Cornfield.

Iowa’s high cancer rates are understandably generating a lot of headlines. And for better or worse, a lot of groups are working overtime to offer their own “interpretations” of some of the various reports on the cancer crisis that have come out. Since there are many opportunities to misunderstand these reports, I thought it might be valuable to clarify a few things in response to some of the claims I’ve been seeing.

One of the most egregious “misunderstandings” (I’m putting this in quotes because I think it’s an extremely charitable way of describing it) was the National Pork Producers Council’s response to a recent Yale University study.

Conflating studies of different populations

Researchers found that cancer rates were higher in areas in close proximity to large livestock feeding operations. Inside Climate News quoted the pork producers council’s response:

The Yale research is not as thorough as other analysis on the same industries, said Wendy Brannen, vice president of communications and marketing for the National Pork Producers Council, which represents more than 60,000 pork operations.

“There is rigorous, long-term research available that has followed real farmers for 30 years and found that Iowa farmers are significantly less likely to develop cancers than the general population, not more,” Brannen said in a statement that referenced a study published in March by the Iowa Cancer Registry.

Yale’s county-level comparisons, which the study’s authors described as “exploratory,” does “not refute” Iowa’s long-term data, Brannen said.

This statement contains numerous incorrect assertions, which could easily be found from reading even the abstract of the study, not to mention the study itself. The most egregious thing wrong with Brannen’s comments: the study cited had absolutely nothing to do with livestock feeding operations. It was specifically focused on pesticide applicators!

From the abstract of the 2019 paper Cancer incidence in the Agricultural Health Study after 20 years of follow-up, which was the main source of the Iowa Cancer Registry report finding that cancer incidence was lower among certain farmers:

The AHS cohort has been described in detail [21]. Briefly, from 1993 to 1997, 52,394 private pesticide applicators (IA and NC) and 4,916 commercial pesticide applicators (IA only) were recruited and completed enrollment questionnaires when they renewed their restricted-use pesticide licenses (82% applicator response rate). A total of 32,346 spouses of private pesticide applicators in IA and NC (an estimated 75% of spouses of married applicators) completed and returned enrollment questionnaires.

Needless to say, the finding that pesticide applicators had a lower rate of cancer than the general population does not in any way, shape, or form refute a study result that is specifically focused on how close one lives to concentrated animal feeding operations.

That claim is so wild that I’m amazed any organization would adopt it as a talking point, let alone one as well-funded as the National Pork Producers Council. A word to the wise: if you come for Yale in an academic argument, you probably should bring your A Game.

Misinterpreting the Agricultural Health Study

The pork producers quote stands out as one of the most absurd misinterpretations of the research. But there are other claims about the Iowa Cancer Registry report and the Agricultural Health Study that require a little more explanation in order to show why they are misleading. In particular, I’m thinking about claims suggesting those reports demonstrate that pesticides are not carcinogenic.

In order to explain, it’s first helpful to understand the structure of the cancer registry’s annual Cancer in Iowa Report. The report always has two parts: one part is a summary of the most recently released data regarding Iowa’s cancer rates (which tends to be from a couple years earlier) and the other is a discussion of a specific theme for each year.

The first part of the report is invaluable and is widely cited by almost everyone discussing Iowa’s high cancer rates (including us in our recent Environmental Risk Factors and Cancer report with the Iowa Environmental Council).

The second part is usually a somewhat narrow examination of a specific topic. A such it’s important to be clear on the scope of that examination and the claims involved. In the most recent report, the top line findings from the report were the following:

• Iowa Farmers in the Agricultural Health Study had 13% fewer cancers overall than expected compared to the general Iowa population after adjusting for age and sex. They had fewer cases than expected of cancers of the colon and rectum, lung, bladder, oral cavity and pharynx, pancreas, esophagus, larynx, liver, and tongue. However, they were diagnosed with more cases of prostate cancer and lip cancer than expected.
• Spouses of Iowa farmers in the study had 10% fewer cancers overall than expected compared to the general Iowa population after adjusting for age and sex, including fewer cases of colon and rectum, lung, bladder, pancreas, and cervical cancers. However, they were diagnosed with more cases of melanoma and thyroid cancer than expected.

This has resulted in some unfortunate interpretations and headlines, which imply that exposure to pesticides could reduce the risk of cancer, or at least doesn’t increase the risk.

Governor Kim Reynolds provided an example of this misinterpretation at a press conference in early April. (Thanks to Laura Belin for posting the video.)

If you read the bullet points by themselves, without any context, you might reach that conclusion. However, it’s important to note that neither the authors of the Cancer in Iowa Report nor the authors of the Agricultural Health Study have suggested pesticides are not carcinogenic. In fact, the very next bullet point in their recommendations is the following:

Iowa farmers and spouses reported lower rates of smoking and alcohol use than the general population, which likely contributes to the lower incidence of many cancers. The “healthy worker effect”—the tendency for actively working people to have fewer illnesses—may also play a role.

So right there in the report, there is an explanation for why farmers’ cancer rates were lower. It has more to do with lifestyle factors than environmental risks.

Using a narrow data set to answer a broad question

When considering the broader question of whether exposure to various pesticides increases the risk of cancer, it’s also important to note that the agricultural health studies (and specifically the ones quoted in the Iowa Cancer Registry report) are just one small subset of available data to consider.

As seen above in the cohort description, the Agricultural Health Study tracks farming populations in Iowa and North Carolina over many years. This project employs excellent researchers who rely on state-of-the-art methodology for their analysis, and as such is an incredibly valuable source of information. Even so, if you want to ask a broader question of “is chemical X linked to a higher rate of cancer”, you wouldn’t limit yourself to only one study (or one set of studies).

For example, if you were studying whether alcohol might be linked to Iowa’s high cancer rates, you wouldn’t limit yourself to only looking at studies tracking bartenders in Iowa and North Carolina (or Wisconsin for that matter, if you enjoy your Wisconsin drinking jokes).

You would not want to limit yourself in that way because there is no such thing as a perfect, 100 percent bullet-proof methodology for studying cancer. Every study has weaknesses, even if many of them have strengths. The Agricultural Health Study findings are no exception to this, and there has been a lot written about some of their methodological limitations.

This is not, in my mind, a criticism of those studies—though some people occasionally frame it that way. It’s simply how science works! You often need multiple studies looking at a problem from many different angles in order to get a full picture of the problem.

The best methodology for an overall evaluation of whether various pesticides are linked to cancer would take into account the agricultural health studies alongside all of the other relevant peer-reviewed and credible sources of information about the association between various exposures and changes in incidence of cancer.

The Iowa Cancer Registry’s most recent report zeroed in on a set of studies that tracked farmers up until the year 2015. Those studies are indeed interesting and worth examining in detail. But it would be a huge mistake to jump from those studies to the conclusion that the most common pesticides used in Iowa have no relationship with cancer. And none of the researchers involved make that jump—even if some industry representatives (and headlines) have done so.

Final thoughts and takeaways

Regarding the question of whether any of Iowa’s pesticides are linked to higher incidences of cancer, I want to note a few things:

First, you will not be surprised to learn I think people should read the Harkin Institute and IEC’s Environmental Risk Factors and Iowa’s Cancer Crisis report. We look at the top three pesticides used in Iowa (atrazine, acetochlor, and glyphosate) and include over 100 citations of articles in our analysis of pesticides. In contrast, the recent Cancer in Iowa report has eighteen references total. Again, that is not a criticism—it just reflects the different aims of the two reports.

Of course, the number of citations by itself doesn’t prove that an analysis is correct, which is why I would highly recommend reading the report for yourself to make your own judgments. I think the quality of the report is evident to all who read it.

Second, as noted above, the studies cited in the Iowa Cancer Registry’s report looked at tracking data up until 2015. But more recent Agricultural Health Study papers have found associations between specific pesticides such as atrazine and cancer rates. So it would not be accurate even to say that those studies have found no relationship, if one were to look at the totality of the research.

Third, Agricultural Health Study researchers don’t limit themselves to only citing their own studies. As good scientists, they cite other literature on the relationship to cancer during reviews of the literature.

Fourth, the Agricultural Health Study is currently in the process of updating their methodology for quantifying non-occupational pesticide exposures. I look forward to learning what the new methodology reveals.

This post has focused mostly on misinterpretations of the Iowa Cancer Registry’s Cancer in Iowa Report. But I’ve also seen our own report cited incorrectly (or at least questionably) a few times as well.

One example: in a recent podcast interview with GOP candidate for governor Zach Lahn (which was pretty interesting overall), the host said the following at around the 18:36 mark:

I was talking to a senator the other day who was a retired farmer and you know, he pointed to the Harkin’s study or whatever that study was and said, oh, well, farmers have 13 percent less cancer than the rest of the population.

Pretty clearly, he was not referring to our study! It sounds like he was talking about the Cancer in Iowa Study, which of course had a very different takeaway message from our report, as noted above. I’ve also heard some other erroneous attributions to “the Harkin study” in recent weeks and as someone extremely proud of the report, I must clarify that it was The Iowa Environmental Council and Harkin Institute report, because they deserve so much credit for making the report as fantastic as it is.

The big takeaway should be fairly obvious: don’t assume that people (especially interested parties with profits on the line) are accurately summarizing reports about cancer in Iowa. Read the reports, or at least their executive summaries, for yourself.