EFSA and IARC: apples and oranges?

March 15, 2016 at 6:07 pm | Posted in Comment | Leave a comment

Can EFSA and IARC’s assessments of the carcinogenicity of glyphosate truly be non-comparable?

Soybean Crop - Barbara Eckstein - Flickr - w1080 - border

A month ago, we were speculating as to whether the few extra studies to which the European Food Safety Authority (EFSA) had access but the WHO International Agency for Research on Cancer (IARC) did not were really able to account for their difference in opinion with regard to glyphosate’s carcinogenicity. We thought it unlikely.

Since then:

  • IARC has cancelled a clear-the-air meeting with EFSA because they were unhappy with how EFSA has been presenting IARC’s review of the carcinogenicity of glyphosate (see letter 12 February 2016).
  • A German study has found suggestion that human exposure to glyphosate is not only ubiquitous, but human exposure often greatly exceeds levels permitted in water.
  • Re-approval of glyphosate for use in the EU has been delayed, with several EU Member States rebelling against EU Commission plans to relicense the compound.
  • A group of researchers has published a detailed critique of EFSA’s approach to assessing the carcinogenicity of glyphosate and the organisation’s criticisms of IARC’s approach.
  • A different group of researchers published a consensus statement voicing concerns about “over use of glyphosate-based herbicides and risks associated with exposures”.

This is an awkward situation. The dependence of European agriculture on glyphosate makes it difficult to ban, but if it is carcinogenic, EU pesticides laws require its use to at least be heavily restricted.

This puts tremendous pressure on any scientific agency which assesses health risks posed by the compound, making it much more likely that any review of the science will be heavily politicised.

And the result is two authoritative bodies in dispute about the carcinogenicity of the world’s most heavily-used herbicide.

Others (including some listed above) have already commented on the relative transparency and scientific robustness of the two agencies’ respective processes and EFSA’s extra vulnerability to politicised scientific opinion in this regard.

Here, we will briefly consider another issue: EFSA’s specific choice of objective in assessing the carcinogenicity of glyphosate.

The difference is in the question, not the data. Last month, we expressed scepticism about EFSA’s claim that the difference in opinion between the agencies about the carcinogenicity of glyphosate came down to EFSA analysing a few confidential studies which IARC could not see.

In our discussion of this claim, we probably gave too much space to a bad argument; however, scientific agencies need to be held to account on their use of scientific research, and it was instructive for a better understanding how weight-of-evidence analysis and research synthesis in risk assessment actually works (i.e. if you have a large body of evidence which strongly suggests one thing, then if a small number of studies are going to overturn that they will have to be extremely strong).

We concluded that the few extra studies which EFSA deemed to be so important are probably irrelevant in explaining why the agency reached a different conclusion to IARC, and that a more plausible explanation therefore lies elsewhere.

One of these explanations is flagged by Bernard Url, Executive Director of EFSA, in a letter to Dr Christopher Wild, Director of IARC. In this letter, Url says EFSA operates “in a specific regulatory context and that the scope and objectives of the EFSA and IARC hazard assessments are different and not directly comparable.”

Url is essentially saying that, when they are assessing glyphosate’s carcinogenicity, EFSA is engaged in a qualitatively different project to when IARC assesses glyphosate’s carcinogenicity, such that the results cannot be compared and any disagreement is somehow superficial and the whole discussion at cross-purposes.

To put it another way: that in spite of appearances to the contrary, while IARC has been dealing in apples, EFSA has been dealing in oranges.

It sounds like an odd defence. Up to this point, everyone has been labouring under the impression that IARC is convinced glyphosate is carcinogenic in animal studies and EFSA is convinced that it is not. But now we are being asked to believe that, in fact, both EFSA and IARC can be right at the same time, because they are talking about different things.

As far-fetched as that is, it might be true. EFSA has different definitions of strength of evidence for carcinogenicity, uses different statistical tests and places different emphasis on historical data. Most significantly, it also looked at a different substance: the pure formulation only, rather than the commercial formulation which IARC included in its assessment.

So perhaps one can legitimately argue that EFSA and IARC are engaged in non-comparable projects. The problem is, the question then needs to be asked: which of the two approaches is fit for purpose? And this raises more awkward questions than it answers:

  • If EFSA’s focus on the cancer hazard posed solely by the active ingredient rather than the real-world formulations of the herbicide (including e.g. adjuvants to make glyphosate more effective) results in it failure to address real-world concerns about its use, then how is it performing a useful societal function?
  • On a related point, what is the value of a scientific assessment which asks a question which few stakeholder in the process actually care about?
  • If EFSA, in its assessment of glyphosate, is engaged in a process which is not comparable with IARC’s approach to identifying carcinogens, why has it not been making this absolutely clear all along, rather than variously suggesting its approach is superior to IARC’s, or their results contradict IARC’s?

If EFSA’s approach really is non-comparable to IARC’s, then all the fuss over whose methods are better has surely done little more than spread confusion and potentially undermined the authority of another scientific agency over an issue which is, in fact, irrelevant to the debate.

If nothing else, at least this saga emphasises the value of getting the question right before embarking on an analysis of the scientific evidence.

Leave a Comment »

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Blog at WordPress.com.
Entries and comments feeds.

%d bloggers like this: