3. EU regulators “disappear” birth defectsThe regulators’ response to Carrasco’s study suggests that they are in no hurry to take on board the findings of independent science. At Commissioner Dalli’s request, the German Federal Office for Consumer Protection and Food Safety, BVL, produced a written response to the study.44 This was not published but was only sent to the EU Commission.
BVL’s statement is anonymous. Though common with such items of grey literature, this operates against the public interest as no one can be held accountable a decision that could significantly affect public health. There is no way of knowing whether the people who wrote it are even qualified scientists, let alone if they have industry interests.
BVL’s conclusion can be summarised as: no action is needed on glyphosate. It tries to isolate Carrasco’s study, implying that it is the only one to find problems. BVL cites Germany’s 1998 draft assessment report (DAR) on glyphosate, which it says showed “no evidence of teratogenicity” (ability to cause malformations/birth defects).45
The DAR is a crucial document underlying glyphosate’s current EU approval. It is Germany’s summary and report on the dossier of studies submitted by industry in support of glyphosate’s approval. Based on this DAR, along with EU member states’ comments and a peer review of the dossier by the EU Commission’s ECCO scientific panel, the EU Commission’s health and consumer division DG SANCO approved glyphosate for 10 years in 2002. DG SANCO’s final review report on glyphosate acknowledges developmental abnormalities found in the industry studies but dismisses their importance by saying that they are confined to “maternally toxic doses” (see Section 3.1, below).46
BVL’s response to Carrasco was followed by a response from industry. Employees of Monsanto and Dow, two major manufacturers of glyphosate herbicides, published a letter in the same journal that published Carrasco’s original study.47 The Monsanto/Dow letter was published back-to-back with Carrasco’s response.48
Monsanto/Dow take the same line as BVL, claiming:
Glyphosate does not cause adverse reproductive effects in adult animals or birth defects in offspring of these adults exposed to glyphosate, even at very high doses.49
But both BVL’s and Monsanto/Dow’s claims are misleading, as we show below.
3.1. Industry’s own studies show that glyphosate causes malformationsGermany’s DAR concludes from the industry dossier of studies, “Glyphosate does not cause teratogenicity”. But Germany immediately goes on to qualify its conclusion, saying that higher doses of glyphosate caused “reduced ossification and a higher incidence of skeletal and/or visceral [internal organ] anomalies” in rats and rabbit foetuses.50 In reality, at odds with Germany’s reassuring conclusion, the details of the DAR contain convincing evidence of glyphosate’s teratogenicity.
Germany adds that in the industry studies, glyphosate given at high doses reduced the number of viable foetuses produced by rats and rabbits.51 Decreased numbers of viable foetuses are often consistent with increased incidence of malformations, as many mal-developed foetuses are spontaneously aborted.
The skeletal “anomalies” found in these early industry studies are consistent with Carrasco’s findings. But Germany dismisses them on the claimed grounds that the doses at which the effects were found were so high as to be toxic to the mothers (maternally toxic doses).
Germany here makes an assumption common among regulatory authorities – that foetal abnormalities found at maternally toxic doses are irrelevant to human risk assessment. The reasoning is that poisoning of the mother with any substance can affect the development of the foetus and lead to birth defects and therefore such malformations may not be a direct effect of the chemical in question on the foetus. So malformations in foetuses found at dose levels that are considered toxic to the mother are dismissed as irrelevant and the substance under examination is not classed as a developmental toxin or teratogen.
But this assumption is debated in the independent scientific literature. Paumgartten (2010) says that in cases of maternal toxicity, it is not possible to know whether an effect on the embryo is only due to maternal poisoning or due to a direct action of the chemical at doses that also adversely affect the mother. In the latter case, the chemical would be a developmental toxin.52
Even industry is actively discussing the relationship between maternal toxicity and birth defects. It was the subject of a recent workshop held by the industry-funded group, the International Life Sciences Institute.53
As yet there is no scientific consensus around the issue. The confusion is made worse by the poor design of standard industry chronic toxicity tests, which use so few animals that unrealistically high doses of the chemical have to be used in an attempt to obtain statistical significance in non-lethal effects.54 In fact, the doses for chronic two-year toxicity tests are derived from, and are only slightly below, the acute poisoning dose. So poisoning effects are common in such tests, which often miss more subtle effects.55
Thus, virtually all chronic tests commissioned by industry have an escape clause: “Perhaps the dose was so high it poisoned the animals.” This escape clause is frequently used by the rapporteur Germany in its DAR on glyphosate.
Germany’s dismissal of the malformations found in industry studies on grounds of maternal toxicity is thrown into doubt by the findings of an independent study. Dallegrave (2007) examined the reproductive effects of Roundup on male and female offspring of Wistar rats treated with 50, 150 or 450 mg/kg of Roundup during pregnancy and lactation. The study found that these doses of Roundup did not induce maternal toxicity but did induce adverse reproductive effects on male offspring. Findings include a decrease in sperm number and daily sperm production during adulthood, an increase in the percentage of abnormal sperms, a dose-related decrease in the serum testosterone level at puberty, and signs of sperm cell degeneration during both periods. The study showed that Roundup is a reproductive toxin at non-maternally toxic doses.56
Even if we confine the argument to evidence generated by industry studies, Germany’s argument that glyphosate’s teratogenicity is confined to high, maternally toxic doses is untrue. The industry studies also found malformations at lower doses. This is made clear by Germany’s own summaries of the industry studies in the DAR and by the comments of the UK’s Pesticides Safety Directorate (PSD).
Our edited versions of Germany’s summaries of industry studies are presented below, along with the UK PSD’s comments and our own.
How are pesticides assessed for risk in the EU?
Risk assessment of pesticides in the EU is a long and complex process:
●●Industry submits a dossier of studies in support of its application for approval of a pesticide. The studies should fulfil the data requirements of the regulation in force.
●●The rapporteur member state reviews the industry dossier and compiles a draft assessment report (DAR).
●●The EU member states are invited to comment on the industry dossier and DAR.
●●A scientific panel of the EU Commission – formerly the ECCO Panel, now EFSA’s Panel on Plant Protection Products and their Residues (PPR Panel) – reviews the industry dossier and DAR, and writes an Opinion.
●●The EU Commission’s Health and Consumer Protection Directorate General (DG SANCO) compiles a review report, summarising the evidence on the pesticide.
●●A committee made up of representatives of DG SANCO and the member states, known as the Standing Committee on the Food Chain and Animal Health (SCFCAH) – Phytopharmaceuticals, meets to discuss the pesticide.
●●DG SANCO makes a proposal at a meeting of the SCFCAH to approve, reject, or conditionally approve the pesticide for certain uses.
●●If a large majority of SCFCAH members reject DG SANCO’s proposal, DG SANCO can change it or find a compromise.
●●The SCFCAH votes on whether to accept DG SANCO’s proposal.
●●In the event that a qualified majority vote is not achieved, the proposal passes to the European Council for a final decision.
The documents on which the current EU approval of glyphosate is based, including Germany’s 1998 DAR on glyphosate and comments of member states, are not readily available to the public or seemingly even to the EU Commission’s regulators, DG SANCO. DG SANCO told the authors of this report that it was unable to supply the DAR and referred the request to the German government office BVL, which only supplied it after a delay of several weeks. Even then, BVL withheld part of the DAR. In contrast, DG SANCO’s 2002 review report on glyphosate is publicly available.
Suresh (1993)
Submitter company: Feinchemie57Germany’s summary: This study on the teratogenicity of glyphosate in rabbits found that the total number of foetuses with major visceral anomalies was high in all treatment groups, including the low-dose level of 20 mg/kg, and was significantly increased at the 500 mg/kg (highest dose) level. The percentage of foetuses with dilated heart was significantly elevated at all dose levels. Skeletal variations, anomalies and malformations were found but there was no clear dose-response pattern. There was a dose-related increase in the occurrence of an extra 13th rib in all the glyphosate-treated groups and in the highest dose group this was statistically significant.
The NOEL (no observable effect level, the highest dose tested that did not produce an adverse effect) for maternal toxicity was 20 mg/kg bw/d [body weight]/day, based on the fact that possibly treatment-related deaths occurred in the higher dose groups. With regard to visceral malformations, the study’s author concluded that the NOEL was less than the lowest dose of 20 mg/kg bw/d.58
UK’s comment: “The increased incidences of abnormalities … are of concern, particularly the heart effects which are also reported in other rabbit studies with glyphosate… The interpretation of this finding must rely on comparison with historical control data. If the typical incidence [of malformations] is approximately 5 fetuses per group then there is no concern. However if this is a very rare finding in control animals and the concurrent controls for this study are typical then there are concerns regarding the potential fetotoxicity of this source of glyphosate.”59
Our comment: With regard to this study, even industry is telling Germany that glyphosate is toxic at 20 mg/kg bw/d, if not lower. Germany however explains away the findings on the grounds that the actual number of foetuses with dilated heart was small, that there was no increase in foetuses with heart dilation in the mid-dose over the low-dose group, that almost no other soft organ malformations occurred, and that the consequences of this heart malformation are “equivocal”. Together, those arguments lead them to conclude that the low dose of 20 mg/kg bw/d and even the mid dose of 100 mg/kg bw/d were NOELs.
An objective evaluation of this study would conclude that the low dose of 20 mg/kg bw/d is not the NOEL, or, as it is usually called today, the NOAEL (no observed adverse effect level). In this study, 20 mg/kg is the LOAEL – the lowest level at which an adverse effect was found. Statistically significant teratogenic effects were found at this dose. As no NOEL was found in this study, Germany should have demanded that further tests be done to establish the NOEL, with the highest dose set at 20 mg/kg bw/d and lower doses added to try to establish a true NOEL.
Germany’s comment that the number of foetuses with abnormalities was small merely identifies a shortcoming of industry studies. Larger numbers of test subjects are always preferable. If the number of animals used in the study is small, any effect will only be seen in small numbers of animals.
Germany’s dismissal of the heart malformations on the grounds that no other types of soft organ malformations were found is not consistent with the current state of knowledge in developmental biology. Many toxic agents target a specific organ (known as “organ specific” effects) or have one specific effect. In light of this, Germany has no basis for arguing that the heart malformations are not important because malformations were not observed in other tissues. Also, Germany’s argument that the heart dilation malformation has unknown consequences and can therefore be dismissed is scientifically and clinically indefensible.
Germany’s expectation of a proportional dose-response pattern in skeletal malformations is also not supported by current knowledge of developmental biology. There is no evidence in the scientific literature demonstrating that toxicity must always be proportional to dose, increasing as the dose increases. Toxicologists now recognize that dose-response relationships can be complex, especially when the endocrine system is involved. Toxic effects can be found at low doses but not at higher doses, and different toxic effects can be found at different doses.60 61 62 63
Industry toxicologists ignore these scientifically established facts. They only test unrealistically high doses and extrapolate effects to low doses, wrongly assuming a linear dose-response relationship. They also wrongly assume that there is a threshold dose below which there is no toxicity. In short, they fail to gather data from almost every area of the dose-response graph.
Germany has wrongly dismissed the hard data in this study, which clearly indicate the toxicity of glyphosate.
Germany repeatedly tries to explain away the finding of malformations in industry studies by referring to historical control data instead of focusing on comparison of the experimental and control groups within the study under consideration. When a study shows clear differences between experimental and control groups, instead of concluding that the study demonstrates the toxicity of glyphosate, Germany compares the experimental group to control data from other sources. Such control data will have wide variability, the range of which will overlap with the values reported for the glyphosate-treated groups in the study under consideration. Based on this overlap, Germany concludes that there is no evidence of toxicity, since the experimental results are within the range of normal variability. This conclusion is not valid, because the variability within the control data gathered by Germany is artificially large, due to the fact that the studies from which those data are drawn have been done under a range of conditions.
Germany’s practices might be overlooked if the effects found were marginal and if other studies with similar findings did not exist. But neither condition applies to these industry studies on glyphosate, which consistently show malformations. Significantly, the independent studies cited in this report do not rely on “historical control data” to explain away findings.
It is clear from the UK’s PSD’s comment on the teratogenicity studies that it had not seen the historical control data and so was not prepared to discount the possibility that glyphosate was teratogenic and toxic to foetuses.64
Brooker et al., 1991
Submitter companies: Monsanto/Cheminova65Germany’s summary: This study looked at the effects of glyphosate on pregnancy in rabbits, at doses of 50, 150, and 450 mg/kg bw/d. It found a significant increase in embryonic deaths in all the glyphosate-treated groups compared with controls. However, a comparison with historical control data showed that the incidence in the control group was untypically low. Also, a clear dose-response relationship was not shown. On the other hand, an increase in late embryonic deaths at the top dose level (450 mg/kg bw/day) was also found in another study on rabbits.
There was concern about the more frequent occurrence of foetuses with heart malformations in the high dose group, but the incidence was in the range of historical background data. However, anomalies of the heart have been described in other rabbit teratogenicity studies with glyphosate, too. Thus, a possible effect on the occurrence of visceral anomalies remains equivocal.66
UK’s comment: “The increased levels of embryonic death/post-implantational loss at all dose levels are of concern, as are the reports of heart defects… a more robust argument should be presented before these findings can be dismissed.”67
Our comment: Again, Germany uses historical control data and an inappropriate model for toxicity dose-response to explain away malformations of the heart in a glyphosate-exposed group. Again, by taking this position, Germany appears to be acting against the public interest by ignoring or dismissing findings of glyphosate-induced teratogenicity and foetotoxicity.
Bhide and Patil (1989)
Submitter companies: Barclay/Luxan68Germany’s summary: This study examined teratological effects of glyphosate in rabbits at doses of 125, 250, and 500 mg/kg bw/d. At the high dose, two females aborted. There was no evidence of foetotoxic and teratogenic effects up to and including the mid-dose group. But the high-dose group had a decreased number of viable foetuses per litter and the number of non-viable implants (non-development and death of embryo) increased. The number of visceral and skeletal malformations was increased in the high-dose group.69
The study’s authors do not mention whether a statistical analysis was performed.
UK’s comment: “Another study with equivocal evidence of heart defects.”70
Our comment: The data shows that dose-dependent increases in lung and kidney malformations were found across all glyphosate-exposed groups. Increased heart malformations were found in all exposed groups. Increased skeletal (rudimentary 14th rib) malformations were found in the mid-dose and high-dose groups.
Germany incorrectly claims that the teratogenic NOAEL is the mid dose of 250 mg/kg bw/d. In reality, there are evident increases in most of the defects, even at the lowest dose of 125 mg/kg bw/d. The authors of this study do not provide an analysis of statistical significance and groups of only 15 animals were used, making statistical significance difficult to establish. But it is more accurate to say the mid dose, possibly even the low 125 mg/kg dose, is the LOAEL. Testing the effects of lower, realistic doses requires far larger animal groups if an increase in toxicity compared with the unexposed control group is to be reliably detected.71 72
At the very least, this study should have been repeated with a larger sample size and lower doses. Effects should have been examined thoroughly by allowing full gestation and pup development.
Anonym. (1981)
Submitter company: Alkaloida73Germany’s summary: This oral feeding study examined teratological effects of glyphosate in rats and rabbits. Vital details were either not recorded or poorly described, so the study was only considered as supplementary information. No malformations were recorded, but there were more foetal deaths at the two upper dose levels (50.7 and 255.3 mg/kg bw/d).74 It is difficult to understand why an increase in foetal deaths would occur at doses far below those at which foetal effects were found in the gavage [force-feeding via stomach tube] studies. Thus it is doubtful whether this effect is related to glyphosate.75
UK’s comment: “Though this study is questioned [by the rapporteur, Germany] for showing evidence of fetotoxicity at lower doses than other studies, the study by Brooker (see above) may also indicate fetotoxicity at 50 mg/kg bw/d.”76
Our comment: Germany here again appears to show a bias towards considering low-dose findings as non-treatment-related and irrelevant – seemingly because it cannot accept that oral feeding may result in different exposures and effects than gavage. But the UK’s PSD points out that another study supports this study’s findings.
Tasker, E.J. and Rodwell, D.E. (1980)
Submitter companies: Monsanto and Cheminova77
Germany’s summary: This teratogenicity study in rats found a higher number of foetuses with malformations at the highest dose level (3500 mg/kg bw/d), but this was within the range of historical control data and was not considered to be due to glyphosate treatment. Specifically, there were more foetuses with unossified sternebrae (bones of the sternum/breastbone) in the high-dose group. While this effect was considered to be due to the glyphosate treatment, it is “rather a developmental variation than a malformation.”78
UK’s comment: The UK PSD does not comment on this study.
Our comments: Germany once again resorts to historical control data in order to conclude that there is lack of evidence of teratogenicity. Given the findings of malformations from glyphosate treatment in several other studies, this is unjustifiable.
Germany’s decision to redefine unossified sternebrae as a “variation” rather than a malformation is scientifically unjustifiable and at odds with other authorities. Unossified sternebrae in the rat are clearly defined as a skeletal deformity in The Handbook of Skeletal Toxicology.79
3.2. Glyphosate’s “pattern” of teratogenicity dimissed by EU expert panelThe UK PSD’s overall conclusion supports the teratogenicity of glyphosate: “Taken in isolation, none of the findings in these rabbit teratology studies would be clearly of concern. However, overall there is an indication of a pattern.”80
The PSD ended by asking Germany to make available the historical control data. It is unclear whether the PSD ever saw this data or, if it did, how it responded. Certainly, the data has not been placed in the public domain for scrutiny by independent scientists.
The teratogenicity question then passed to the EU Commission’s ECCO scientific review panel. The panel noted “the incidence of heart malformations”, but dismissed them on the grounds that they were “within the range of the historical control data”.81 It is unclear whether the ECCO Panel saw the historical control data or merely accepted Germany’s conclusion. No details are given of the previous studies from which the historical control data were derived or how the figures were analyzed. The experimental animal species, experimental design, identity of the researchers and laboratories, and purity of the substance tested, are unknown. There are significant variations between different formulations of glyphosate: glyphosate produced in the 1970s will not be the same as formulations produced in later decades. But none of these variables can be checked because the historical control data does not appear to be in the public domain.
The historical control data that enabled the ECCO Panel to dismiss glyphosate’s teratogenicity must be added to the large pile of grey literature supporting pesticide approvals that cannot be evaluated by the public or independent experts.
The use of historical rather than concurrent controls adds variables to an experiment that aims to control variables, obscures the teratogenic effects of glyphosate, and biases any conclusion. This is why the use of historical control data is controversial.82 83 The practice should not be allowed in evaluating animal toxicological and other studies for pesticide approvals.
Valid control groups for an experiment are animals of the same strain and age, in the same environment, which are studied at the same time as the exposed (experimental) animals. In addition, the manner in which the animals are examined and evaluated, and the data recorded, must be the same. “Historical control data” fail to meet these criteria. It appears that they are being used as a smokescreen to hide glyphosate’s teratogenic effects.
Clearly, only after findings emerged showing glyphosate’s teratogenicity did Germany and the ECCO Panel introduce the artifice of historical control data as a way of calling into question the scientifically-proper controls. In this way, the differences between exposed and unexposed animals were buried in the variability within the historical control data.
If such practices were uncovered in an independent scientific study, they could be considered scientific fraud. In this case, we do not even know who perpetrated this act, which has placed public health at risk.
Taking all these industry studies together, there is enough evidence to require regulators to apply the precautionary principle and withdraw glyphosate from the market.
3.3. Industry and regulators failed to disclose glyphosate’s teratogenicityThe evidence discussed above shows how EU regulators reasoned their way from evidence of glyphosate’s teratogenicity in industry’s own studies to a dismissal of these effects in DG SANCO’s 2002 final review report.84
Taken together, the industry studies and regulatory documents on which the current approval of glyphosate rests reveal that:
●●Industry (including Monsanto) has known since the 1980s that glyphosate causes malformations in experimental animals at high doses
●●Industry has known since 1993 that these effects could also occur at low and mid doses
●●The German government has known that glyphosate causes malformations since at least 1998, the year it submitted its DAR on glyphosate to the EU Commission
●●The EU Commission’s expert scientific review panel has known since 1999 that glyphosate causes malformations
●●The EU Commission has known since 2002 that glyphosate causes malformations. This was the year its DG SANCO division published its final review report, laying out the basis for the current approval of glyphosate.
The public, on the other hand, has been kept in the dark by industry and regulators about the ability of glyphosate and Roundup to cause malformations. In addition, the work of independent scientists who have drawn attention to the herbicide’s teratogenic effects has been ignored, dismissed, or denigrated.
3.4. Germany set misleading “safe” level for glyphosateThe central purpose of any pesticide risk assessment is to establish the Acceptable Daily Intake (ADI), a level of exposure that is deemed safe for humans over a long period. The ADI is calculated from the industry tests in the dossier. The level that should be used to set the ADI is the highest dose at which no adverse effect is observed (NOAEL), which is also lower than the lowest dose that is toxic (LOAEL). This level should be selected from “the most appropriate study in the most sensitive species”, as the glyphosate rapporteur Germany notes.85
Germany set the ADI for glyphosate at 0.3 mg/kg bw/d.86 This ADI was accepted by the European Commission in its final review report.87
However, we argue that this is incorrect. Germany indulges in some creative manipulation of data to arrive at this level. It begins by excluding certain studies from the ADI process:
●●Germany excludes mid-term studies on the grounds that only long-term studies should be used to set safe chronic exposure levels.88 This enables it to avoid using the rabbit teratogenicity studies, which were mid-term.
●●Germany claims that the most sensitive species for chronic exposure is the rat. This gives it another reason to exclude the inconvenient rabbit teratogenicity studies, which found significant adverse effects at lower doses than the rat studies.
Based on this biased selection of data, Germany cites as its starting point for working out the ADI a LOAEL of 60 mg/kg bw/d from a two-year rat study by Suresh (1996), which found significant toxicity at that level.89 This is said to be the lowest dose at which toxicity was observed. Germany then identifies the highest NOAEL below that level: 31 mg/kg bw/d, in a study by Lankas (1981). It implies that this is the figure from which it calculates its ADI (though also, confusingly, denies that it bases the ADI on any single study).90 The ADI is derived by dividing this figure by 100, to allow a safety margin. Applying this 100-fold safety factor, Germany arrives at an ADI of 0.3 mg/kg bw/d.
However, we argue that Germany should have begun the ADI process using the LOAEL of 20 mg/kg from the 1993 Suresh rabbit teratogenicity study, which is three times lower than Germany’s chosen LOAEL of 60 mg/kg bw/d.91 92
To sum up the difference between these two studies:
●●The study Germany uses to set the ADI: Suresh’s 1996 chronic toxicity study on rats found statistically significant toxicity at 60 mg/kg bw/d (the LOAEL).
●●The study Germany ignores in setting the ADI: Suresh’s 1993 teratogenicity study on rabbits found statistically significant toxicity at 20 mg/kg bw/d (the LOAEL).
Germany relegates the inconvenient 1993 Suresh study to setting the acceptable operator (applicators’) exposure level (AOEL). It argues that it is a mid-term rather than long-term experiment and therefore more suitable to setting an applicators’ exposure level.93
We believe that Germany’s reasoning would not stand up to independent scientific scrutiny. Germany’s failure to take into consideration the worrying rabbit teratogenicity studies means that its ADI ignores the problem of the teratogenic effects of glyphosate – as shown even in weak industry studies.
In our view, the Suresh 1993 LOAEL of 20 mg/kg bw/d should be the starting point for the ADI and for the applicators’ AOEL. The 1993 Suresh study from which this LOAEL is derived found no NOAEL (no observed adverse effect level). In other words, even the lowest dose produced adverse effects.94 95 So Germany should have insisted on further tests to establish the NOAEL, using 20 mg/kg as the highest dose.
If this LOAEL of 20 mg/kg were used, then, following the same procedure as Germany, the highest NOAEL below this dose from Germany’s approved list of studies is 10 mg/kg.96 Applying the 100-fold safety factor, this would give a more objectively accurate ADI of 0.1 mg/kg bw/d, one-third of the ADI suggested by Germany.
Interestingly, one of the industry applicants, Feinchemie, suggests a far lower ADI than Germany or us: 0.05 mg/kg bw/d. This is five times lower than the ADI suggested by Germany and accepted by the Commission. Feinchemie bases its suggested ADI on its 2-year rat study, which found a NOAEL of 5.5 mg/kg bw/d.97
Feinchemie’s suggested ADI is consistent with the NOAEL of the 1996 Suresh study, which Germany used to derive the LOAEL but ignored to set the ADI. The NOAEL of that study was 6.3 mg/kg bw/d, which would give an ADI of 0.06 mg/kg bw/d, close to Feinchemie’s proposed ADI of 0.05 mg/kg bw/d.
It is ironic that industry asked for stricter – and more scientifically justifiable – safety standards than the rapporteur, Germany. In contrast with Feinchemie’s proposed low ADI, however, Monsanto asked for an ADI of 1.75 mg/kg bw/d, the highest of all the industry-suggested ADIs.98
3.5. What the ADI should be – according to independent studiesIf a manufacturer of glyphosate says the ADI should be five times lower than the one suggested by Germany and accepted by the Commission, what do independent studies say it should be?
Two high-quality mammalian toxicity studies show that glyphosate’s LOAEL should be even lower than that proposed by Feinchemie (which in turn was lower than that proposed by Germany):
●●A study on rats showed that a Roundup formulation was a potent endocrine disruptor and caused disturbances in reproductive development when the exposure was performed during the puberty period. Adverse effects, including delayed puberty and reduced testosterone production, were found at all dose levels, including the LOAEL of 5 mg/kg. The dose-response relationship was clear.99 One of the critical failures of regulatory toxicity tests is to ignore important developmental windows such as puberty. This study helps to fill that knowledge gap.
●●A 75-day study on rats showed that Glyphosate-Biocarb (a Brazilian formulation) caused damage to liver cells in a dose-response manner, including at the LOAEL of 4.87 mg/kg. According to the authors, the findings suggest that the damage to liver cells was “irreversible”.100
Both studies use a species (rats) and an exposure route (oral) approved by EU regulators and industry.
No dose below these two LOAELs was tested in these studies, so the true NOAEL is lower – by how much, no one knows. But the NOAEL could reasonably be assumed to be 2.5 mg/kg bw/d. Applying the usual 100-fold safety margin results in a scientifically defensible ADI of 0.025 mg/kg bw/d. This is over ten times lower than the Germany’s ADI, which is currently in force. The MRL (safe level in food) should be correspondingly revised downward.
Of course, all assumptions need to be tested, and not even independent science has explored the full picture of Roundup and glyphosate’s toxicity. Studies should be carried out immediately to determine the true NOAEL and ADI for glyphosate and Roundup, using the most comprehensive, up-to-date scientific knowledge. These studies would involve:
●●testing for more effects
●●using lower, more realistic doses that will allow accurate determination of the NOAEL
●●using larger numbers of animals to ensure sufficient statistical power to reliably detect effects from realistic doses
●●dosing during vulnerable developmental windows
●●extending study time-frames to allow mid- and long-term effects to show up, instead of killing the test animals before disease has a chance to develop. Industry test animals are killed at the human equivalent of about 60 years old, so many effects of the chemical tested are missed.101
In addition, research should be done that determines glyphosate levels in food and feed imported into the EU. Finally, based on independent (not industry) data from North and South America, an assessment should be carried out of the increase in glyphosate use, and therefore exposure, that would be expected to occur if glyphosate-tolerant GM crops were allowed to be grown in the EU.
This science-based assessment of glyphosate and Roundup will allow the EU to establish a credible policy that protects EU citizens. Until that assessment is complete, the EU should apply the precautionary principle and withdraw glyphosate herbicides from sale in the EU.
3.6. Does current risk assessment protect the public?The current system of pesticide risk assessment in the EU is not transparent or easy to understand. Those who make the effort to study it will see that it is open to manipulation and abuse. In risk assessment, it is the details that count. The conclusions that are drawn depend heavily on how data is selected – what is included and what is left out. This is clear from the above discussion of the approach Germany used in justifying its incorrect conclusion that glyphosate does not have teratogenic or foetotoxic effects. Particularly revealing was Germany’s exclusion of the rabbit teratogenicity studies in setting the ADI for glyphosate.
Industry also has room for manoeuvre in discussions of how toxins behave in the human body. For example, industry uses broad arguments to claim that toxins are broken down in the liver, or do not cross the placental barrier in pregnant women. Even cases of clear harm can be minimized. An anonymous scientist critic said, “There are many tricks that are used. If all insects in a field are killed for a full year, this is not a problem, because they will come back next year. A regulator told me that with the current system of risk assessment you could get any chemical approved, including DDT.”
Even the underlying assumption of risk assessment, that there is a “safe” level below which a toxic pesticide is not toxic, is questionable. Many compounds accumulate in the body. Some toxins, particularly endocrine disrupting chemicals, are more potent at low doses than higher doses. People and species vary in their susceptibility to toxins and individuals at different stages in development and maturation and at different stages of biological cycles. Even the latest independent science has only begun to explore the true effects of chemicals on vulnerable groups such as developing foetuses, infants and children, the elderly, and immune-compromised people.
In addition, the industry tests carried out for risk assessments mostly look for a narrow range of gross effects. These include tissue and structural changes, such as malformations and tumours, which tend to occur at the high doses that industry tests use. But these tests often miss functional changes (effects on how the body’s organs and systems function), which tend to be seen at lower doses and more closely reflect effects from real-life human exposures. These functional changes are important because they can lead to more severe and difficult-to-reverse disease conditions. In other words, they perform a signalling role in predicting serious health problems. Independent scientific literature, which is not tied to OECD test designs, has been more effective than industry science in finding these functional effects – but it has hitherto been virtually ignored in pesticide risk assessments.
Other aspects that have not been adequately examined in existing risk assessments are the impact of the individual’s existing body burden of toxins and synergistic effects that are not seen when the compounds are tested in isolation.
For these reasons and more, some scientists and policy-makers advocate reforming the risk assessment of pesticides – for example, by increasing the use of hazard analysis. Hazard analysis stipulates that if a pesticide has certain hazardous qualities, it should automatically be rejected (“hazard cut-off”). This differs from the current risk assessment approach, which assumes that even when a hazard exists, the risk can be managed. The new pesticide regulation 1107/2009 contains some “hazard cut-off” criteria. For example, a pesticide cannot be approved if it is carcinogenic, mutagenic, a reproductive toxin, persistent in the environment, bioaccumulative, or an endocrine disruptor (apart from specific uses, such as in closed systems).
These are positive developments. But industry, together with the German Federal Institute for Risk Assessment (BfR), which is involved in the registration of pesticides in Germany, is lobbying to prevent the new system of hazard analysis and cut-off criteria gaining a foothold in Europe and to keep the existing system of risk assessment.102 103
For the sake of public health, it is vital that they do not succeed.