Why I don’t trust the regulation of pesticides. A blog by Guy Watson.

147 pesticides I was assured were safe in the 1970s, based on the “best science available” at the time, have subsequently been banned, as risks to users, the environment or the public have emerged. The “best science available” was proven wrong, and history suggests that regulation has repeatedly underestimated the risks posed by pesticides. Yet the underlying assumptions used in assessing the toxicity of pesticides have not changed.

The “cocktail effect” might explain some of the failures. Toxicological assessments almost always look at toxins in isolation, despite the fact that synergistic or “cocktail effects” (whereby two toxins can create effects together greater than the sum of their individual toxicities), were first proven in the 1960s and are now well established (see references below). To quote a recent EU environmental report “There is strong evidence that chemicals with common specific modes of action work together to produce combination effects that are larger than the effects of each mixture component applied singly”. With so many chemical toxins in our environment the possible combinations are almost infinite, making realistic assessment impractical. This is not an excuse for ignoring them; the uncertainty should at least be acknowledged.

A second possible reason for science so persistently getting it wrong is the assumption that the dose determines the poison; ie. that toxic response is proportional to the dose received. Hormones don’t work in this way; their action, through time and site specificity, is much more subtle. Many pesticides are known endocrine (hormone system) disruptors, so we really shouldn’t be surprised to find effects at minute doses, often below those considered “safe” by our regulators.

“Absence of convincing results shouldn’t be taken as evidence of safety”

It’s easy to design an experiment to determine whether a chemical kills or damages a rat (and by extrapolation poses danger to humans) if the effect is quick (normally 3 months), shortlived and in isolation. If the effect is slow, things get much harder. If it is complicated by interactions with other chemicals, environmental factors or disease, things get progressively more complex with each interaction until assessment becomes practically impossible; the experiment needs to be larger, costs more and is less likely to deliver convincing results.

Absence of convincing results has too often been taken as evidence of safety. Should we really be surprised that neonicotinoids, applied to oilseed rape and present in sufficient concentrations throughout the plant to kill aphids and flea beetles, months later damage the nervous systems of bees feeding on the flowers? Perhaps we should be surprised if they didn’t. Yet the experiments and best scientific evidence cited in the approval process missed this subtlety; the bees didn’t die fast enough.

“The regulatory approach to pesticides has not fundamentally changed, so I am not reassured of its safety today”

History suggests that regulation has repeatedly underestimated the risks posed by pesticides. The regulatory approach has not fundamentally changed, so I am not reassured of its safety today. Around two thirds of fruit and vegetables contain detectable pesticide residues, even after washing. 98% of those are within legally acceptable “maximum residue levels” (MRLs) set by the EU and monitored in the UK by the Expert Committee on Pesticide Residues in Food (PRiF), which reports to HSE, so should we worry? Are those MRLs safe?

According to the gov.uk website “People eating small amounts of pesticide residues in their diet are not at risk, provided that intakes are below the safety limits that are set by expert committees when the pesticide is approved”. Rather less reassuring but rather more realistic is the statement that closely follows: “On the best science available, no harm will come to people who consume an amount of pesticide that is below the safety limits for that pesticide”. Based on the evidence of history and on common sense, I believe there can be no absolutely “safe” level for pesticides (especially endocrine disruptors); only degrees of risk which may or may not be deemed acceptable.

My understanding of “best science available” is that residue levels are deemed safe unless an experiment shows them to be dangerous. We live in a world full of potential toxins, so in most cases pragmatism suggests we must accept the risks and get on with it. But unlike asbestos or particulates from diesel engines, pesticides are specifically selected and designed to disrupt life process at incredibly low doses. Many act as endocrine disruptors, are known to interact synergistically (the cocktail effect) and to be the cause of the sort of long term effects slow to manifest themselves (eg. cancers, birth defects) often missed by the typical short term toxicity trials. For known endocrine disruptors and nerve poisons, I suggest that the burden of proof should be on pesticide manufacturers to prove safety, rather than environmentalists to prove danger. For those with the time there are references below; you decide.

“Keep eating fruit and veg”

As a footnote, although I would always encourage people to eat organic fruit and veg when they can, I agree with gov.uk’s comment that “The risk to health from eliminating fruit and vegetables from the diet would far outweigh the risks posed by possible exposure to pesticide residues”. Don’t let the fear of pesticides stop you eating fruit and veg, organic or otherwise. Hopefully the day will come when we don’t need to choose.

Guy Watson

Suggested reading if you have the time and interest
Cocktail effect in carcinogenesis. Particular reference to Atrazine and DDT. Fascinating window into the complexities of how organisms get and fight cancer. Fascinating but heavy going.

The pesticide certification process.

Basic background stuff, mostly sensible and in layman’s terms.

Both reports cast the current prevailing practice of evaluating toxins in isolation into serious doubt.

Brings into serious question the prevailing assumption used in pesticide approval, that toxicity is proportional to dose; particularly for endocrine disruptors. An easy read summary of the science.

The cocktail effect - the facts

  • Risk assessments for pesticides only look at their effects in isolation. But the average diet exposes people to a multitude of these chemicals at the same time – and research suggests they could become more potent when combined. This is known as the “cocktail effect”.
  • 320 pesticides are allowed in nonorganic farming. Their residues are found in 58% of all UK fruit and veg, with many containing multiple residues.
  • Organic farmers are permitted to use 15 pesticides, derived from natural ingredients including garlic, citronella and clove oil. These can only be used under very restricted circumstances. In reality, the vast majority of organic farmers have no need even for these, and if used the amounts are small.
  • Although there have been concerns about a possible synergistic effect between pesticides for some time, there was no strong evidence, and in 2002 the Food Standards Agency concluded “the risk to people’s health from mixtures of residues is likely to be small”. However, a recent study in France has shown 5 common pesticides to be more toxic to human cells when combined than alone.
  • Adjuvants are chemicals added to pesticides to improve their efficacy, eg. by making them stick to plants. Because adjuvants aren’t named as active ingredients, they aren’t subject to the same risk assessments. But research has found that adjuvants might make the world’s most widely sold weedkiller, glyphosate, many times more toxic.
  • Bees feeding from wildflowers growing next to neonicotinoid-treated crops will consume a cocktail of fungicides and insecticides. The fungicides could act synergistically, “making the insecticides up to 1000 times more deadly”.
  • Leading toxicologists have concluded that a substance-by-substance approach to risk assessment is too simplistic and “in danger of underestimating the risk of chemicals to human health and to the environment”.
  • There has been some progress: in 2016, the EU developed software for assessing the risk from exposure to multiple pesticides. However, there is still a long way to go before the cocktail effect is fully understood.

Have you heard of the cocktail effect? Join in the conversation on Facebook and Twitter.