Cocktail of denial

A long-awaited report, commissioned by the UK Food Standards Agency, on risks from mixtures of chemicals in food and the environment has been published. Its reassuring conclusions have already been confounded by a disturbing toxicity study on 2,4-D.

Many consumers and public interest groups believe that long-term, low-dose exposure to mixtures of pesticides in food, water, and the environment is contributing to disease in our society. This concern led the Food Standards Agency (FSA) to commission a report(1) from an independent expert committee – the recently published Working Group on Risk Assessment of Pesticides and similar substances (WiGRAMP) report – to investigate whether or not the current approvals system takes this factor adequately into account.
    Currently, pesticides are assessed for approval one at a time. Despite the widespread use of thousands of different pesticide products, and evidence for multiple-chemical ‘body burdens’ in human tissue, there is no requirement for testing in combination before new pesticides are put on the market.
    The approvals system which has, since the 1950s, allowed high numbers of pesticides into the environment without this safety check, would not be able to sustain public or political support if it can be proven scientifically that combinations of pesticides can have a synergistic effect, that is, a greater effect than those ingredients on their own. This would necessitate testing any new pesticide in combination with all those currently in use, to check the effect, which would be prohibitively time-consuming and costly. Toxicopathologist Vyvyan Howard has described how, to test the commonest thousand toxic chemicals in unique combinations of three, it would require at least 166 million different experiments (and this disregards the need to study varying doses)(2).
    Broadly there are four possible outcomes if chemicals are mixed. If they have simple similar action on the body’s organs, the effect is additive: it is as though the dose was just being increased. If they have simple dissimilar action, they may be acting on different target organs in the body, as well as having an additive effect, so symptoms and effects may be worse. If they potentiate each other (have synergy), their effect will be unpredicatably greater than additive. It is this effect which has the most serious implications for our current risk assessment system.
    There is also the possibility that they have an antagonistic effect. This means that certain pesticides in combination actually decrease the toxic effects of each. The metabolism of laboratory animals adjusts to repeated toxic doses as a survival tactic. Regulators never explicitly state that this is what is also occurring in the human population, as this would be unacceptable to the public. However, one of the studies in this report comments on the ‘protective effect of aldrin and chlordane against the acute toxicity of the carbamate carbanolate(3)’.
    The conclusions of this report by the Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT) are based on a review of 164 toxicological studies. Almost all of them are interpreted to indicate that effects are caused by a simple similar or dissimilar action, or antagonistic effects. Where potentiation has been found (even within certain elements of a series of experiments, or for certain combinations only) cautionary comments are inserted suggesting that synergy is not proven, or that it is difficult to interpret the results, or that the study design is inadequate, or that the applicability to humans is dubious.
    These caveats go beyond appropriate scientific rigour and scepticism. They are based on a convention in toxicology which is fundamentally incorrect and has no place in modern pesticide approvals systems. Again and again, studies showing potentiation are dismissed because the shape of the dose-response curve has not been established. The only curve illustrated in this report shows a linear incline indicating that, the higher dose, the greater the effect(4).

Greatest effect at low doses
However, it is a well known toxicological phenomenon that the dose-response curve can be non-linear, erratic, or an ‘inverted U-shape’, the latter indicating that the pesticide exerts its most toxic effects at lower doses. For example, a new study(5) has found that the greatest reproductive harm to mice was caused by water contaminated with the lowest doses of a herbicide containing a mixture of mecoprop, dicamba and 2,4-D. The greatest decrease in live pups born occurred at a dose of only 0.039 parts per million of 2,4-D, about one seventh of the United States Environmental Protection Agency’s recommended limit for drinking water.
    One of the authors, Warren Porter, believes that there is a specific, very low level at which the chemicals have most effect, and at higher doses, an unknown protective response may occur, over-riding the effect of the chemicals(6).
    A significant aspect of this study was the deliberate use of a low dose equivalent to those to which we can all be exposed, at levels typically found in water or food contamination. Effects found in these studies have crucial significance in assessing health risks from mixtures, the task of the WiGRAMP. However, the Committee chose to leave such studies out. ‘With some exceptions, these have not been included in this review as they generally provide little information regarding the nature of any interactions that may occur between the individual constituents as they have not concurrently investigated the effects of those constituents alone(7)’.
    A close analysis of the report reveals a further bias. Where findings are ambiguous, and it is not clear whether increased effects are due to an additive or synergistic action, the Committee always favour the former possibility. Even in a study(8) which clearly demonstrates the potentiation of chlorpyrifos by pyrethrins, and a solvent, the Committee emphasise that ‘the effects of other combinations did not differ from those expected on the assumption of concentration additivity’. 

Precautionary principle missing
Very few changes have been made to the conclusions and recommendations in the final report, despite a major public meeting organised by the FSA to discuss the draft version in February, and subsequent written submissions to COT. There is still no reference whatsoever to the precautionary principle, or for its application to the approvals process. Despite the volume of evidence of data gaps amassed in the report, the Committee recommends that risk assessments of mixtures should only be carried out for certain combinations. PAN UK urges the government to take a leading role in identifying potentiation possibilities, not leaving it to a limited number of independent research scientists to tackle this vast field.
    Delaying tactics over implementing the precautionary principle are also evident in the recommendation that ‘further work be undertaken … to characterise both the nature of, and dose-response relationships for, combined actions of pesticides.’ But experiments with mixtures on laboratory animals in this report indicate that it is impossible for toxicologists at post mortem to identify which injury was caused by which chemical, or to establish the mechanisms underlying the interactive responses(9). To suggest this programme of work, rather than recommending a reduction of public exposure to pesticides – especially when a system which registers them one at a time has been recognised to be inadequate – does not well serve public health. (AC)

References
1. Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment, Risk assessment of mixtures of pesticides and similar substances (WiGRAMP report), September 2002.
2. Howard V, ‘Synergistic effects of chemical mixtures – can we rely on traditional toxicology? ‘The Ecologist, Volume 27, No. 5, September/October 1997, 
3. WiGRAMP report, op cit 1, 8.176, p. 165.
4. WiGRAMP report, op cit 1, Fig 7.2, p. 90.
5. Cavieres M, Jaeger J and Porter W, Developmental toxicity of a commercial herbicide mixture in mice: I. Effects on embryo implantation and litter size, Environmental health perspectives, Vol. 110, No. 11, November 2002.
6. New Scientist, 121, September 2002
7. WiGRAMP report, op cit 1, 8.182, p. 166.
8. Axelrad J, Howard CV and McLean WG, Interactions between pesticides and components of pesticide formulations in an in vitro neurotoxicity test, Toxicology, Vol. 173: 259-268, 2002.
9. WiGRAMP report, op cit 1, 10.21, page 200.

[This article first appeared in Pesticides News No. 58, December 2002, page 18]