Abstract
ALL LIFE 2023, VOL. 16, NO. 1, 2167872 https://doi.org/10.1080/26895293.2023.2167872 REVIEW Use of the concept ‘environmentally relevant level’ in linking the results of pesticide toxicity studies to public health outcomes a,b c Robin Mesnage and Charles Benbrook Gene Expression and Therapy Group, Faculty of Life Sciences & Medicine, Department of Medical and Molecular Genetics, King’s College b c London, London, UK; Buchinger Wilhelmi Clinic, Überlingen, Germany; Heartland Health Research Alliance and Benbrook Consulting Services, Port Orchard, WA, USA ABSTRACT ARTICLE HISTORY Received 8 September 2022 Evidence is growing that human exposures to pesticides are contributing in a myriad of complex Accepted 7 January 2023 ways to chronic disease. Regulatory and public health agencies have struggled for years with the definition of acceptable exposure thresholds. At the same time, scientists are trying to design stud- KEYWORDS ies so that a chemical is delivered at ‘environmentally relevant levels.’ The aim of this review is to: Environmentally relevant; (1) explain the many factors that must be taken into account in determining environmentally rele- risk assessment; pesticides; vant levels or doses; (2) improve the ability to properly translate results from laboratory studies into exposure human-health risk assessment; (3) enhance opportunities to compare results across studies using dif- ferent experimental designs, organisms and routes of exposure. We found that confusion over the relationship between concentrations, dosing levels, regulatory thresholds and ‘safe’ exposure lev- els is common. We provide recommendations to scientists and authors, peer reviewers and journal editors in the hope of advancing understanding of how to design, carry out, interpret and explain the real-world significance of both old and new lines of scientific inquiry. Introduction as thalidomide or diethylstilbestrol, and environmen- tal contaminants including organochlorine pesticides, Evidence is growing that human exposures to chem- plasticizers, so-called ‘forever chemicals’ (poly, etc.) icals are contributing in a myriad of complex ways and heavy metals (EEA 2013). Arecentcaseisthe to chronic disease etiology, as well as reproductive ongoing debate over the oncogenicity and genotoxi- and developmental problems (Landrigan et al. 2018; city of glyphosate-based herbicides (GBHs) and their Mesnage and Zaller 2021). The likelihood that chem- active ingredient glyphosate. The US Environmental ical exposures will cause, accelerate or exacerbate dis- Protection Agency (EPA), the European Food Stan- ease in human populations is difficult to predict with dards Agency (EFSA) and the International Agency for the current battery of toxicity tests required by reg- Research on Cancer (IARC) have reached divergent ulators. Pesticide risk-assessment tools and methods conclusions on whether exposures to GBHs increase have largelyfailedtotakeadvantage of rapidprogress cancer risks (Benbrook 2019). in mechanistic studies, epidemiology and genomic Regulatory and public health agencies have strug- sequencing (Wang and Gray 2015;Benbrooketal. gled foryears with howtoincorporate newtools and 2021). information when carrying out or updating risk assess- The quest for consensus over whether, and under ments, revisiting acceptable exposure thresholds and what circumstances a given chemical increases the risk deciding whether to alter risk-mitigation requirements of an adverse impact on public health is often delayed (Bopp et al. 2019;EFSA 2022). The gap is growing by contradictory evidence and/or alternative expla- between the risk-assessment status quo in industry and nations of observed experimental results (Mebane regulatory agencies and the cutting-edge methods and et al. 2019). Historical examples include drugs such CONTACT Robin Mesnage robin.mesnage@buchinger-wilhelmi.com, robin.mesnage@kcl.ac.uk Gene Expression and Therapy Group, Faculty of Life Sciences & Medicine, Department of Medical and Molecular Genetics, King’s College London, 8th Floor, Tower Wing, Guy’s Hospital, Great Maze Pond, London SE1 9RT, UK Buchinger Wilhelmi Clinic, Wilhelmi-Beck-Straße 27, 88662 Überlingen, Germany; Charles Benbrook cbenbrook@hh-ra.org Heartland Health Research Alliance and Benbrook Consulting Services, Port Orchard, WA, USA © 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 R. MESNAGE AND C. BENBROOK analytical systems currently deepening understanding evolve beyond dichotomous analyses hopefully rele- of how and why exposures to toxic substances trig- vant to, but too often not protective of all exposed ger reproductive problems, developmental anomalies populations to more nuanced appraisals of the degree or chronic disease. of risk likely to stem from routine exposures taking Scientists not affiliated with pesticide manufactur- into account a number of factors that are not typi- ers often strive to design experiments in ways that cally assessed in pesticide toxicology testing and risk enhance the relevance of reported results in protecting assessments. public health (Mesnage and Séralini 2018;Goumenou In an eor ff t to make laboratory studies more rele- et al. 2021). Indeed, an increasing number of stud- vant to human-health risk assessments and regulatory ies are designed to mimic real-life exposure scenarios, decisions, scientists are trying to design studies so that taking into account to one degree or another numer- achemicalisdelivered in vivo to a test organism, or ous confounding and complicating factors including within an in vitro cell assay system at ’environmen- multiple routes of exposure, bioaccumulation, simul- tally relevant levels’ or ‘realistic doses’ or ‘acceptable taneous exposure to several chemicals and risk fac- (to an regulatory agency) rates’ (Figure 1). The ongo- tors unique to certain population groups and indi- ing and commendable eo ff rt to integrate new science viduals (e.g. microbiome health and genetic polymor- from many dieff rent fields and approaches into prac- phisms). Regardless of the approach or focus of pes- tical recommendations to promote public health will ticide risk-assessment science, progress in identifying benefit from a deepened understanding of what con- policy changes and regulatory interventions to pro- cepts like an ‘environmentally relevant dose,’ ‘realistic mote public health depends on: exposure level,’ and ‘relevant dose level’ really mean (Weltje and Sumpter 2017). • Accurate identification of chemical hazards, In this review, we explain how various standard • Understanding the magnitude of risks associated concepts and measures of herbicide levels, doses, regu- with variouslevels, timing,routesand durationsof latory benchmarks and exposures should be deployed, exposure in various population cohorts, including explained and related one to the other. We do not aim especially known, high-exposure cohorts, to present a systematic review on the topic, but we crit- • Recognition of risks likely to arise from alterna- ically commented a diverse set of laboratory toxicity tives to a chemical or chemicals subject to stricter studies designed to inform pesticide risk assessment. regulation and We focus on herbicides containing glyphosate, 2,4- • Realistic assessment of the likely effectiveness of dichlorophenoxyacetic acid (2,4-D) and dicamba as possible risk-mitigation measures. active ingredients. Use of these three herbicides has risen dramatically in recent years in areas where farm- A fundamental, but incomplete principle in toxi- ers have embraced herbicide-tolerant corn, soybean cology is that the dose makes the poison (Hayes and and cotton cultivars (Benbrook 2016). Rising herbicide Dixon 2017). In reality, the dose plus timing of expo- use usually leads to new routes of exposure and higher sures, tissues exposed, other exposures and stresses on body-burden levels as borne out in recent biomon- the organism, life stage, general health and the genetics itoring results (CDC 2022;Freisthleretal. 2022). of the exposed organism ‘makes the poison’ (Festing Section 2 discusses what an ‘environmentally relevant’ and Vesell 1987;Myers et al. 2009;Dallmannetal. dose or exposure level typically refers to. We describe 2016). This reality is why pesticide risk-assessment important regulatory benchmarks and thresholds used science is growing so much more complex and con- around the world and the role of biomonitoring data in tentious despite rapid advances in the underlying sci- linking toxicology study results to public health out- ences. Over the last 50 years, the focus of pesticide risk comes. Section 3 describes several common mistakes assessment hasbeenonjustafewdetectable, adverse in describing the relevance of experimental results to health outcomes in mostly rat and mouse studies in pesticide risk assessment and quantification of possible which treatment groups were not even administered public health outcomes, with special focus on cutting- the formulated chemical that people, and especially edge genotoxicity, epigenetic and microbiome studies. mixer-loaders and applicators, are exposed to. In the Last, section 4 provides recommendations to scien- decades ahead, pesticide risk-assessment science must tists and authors, peer reviewers and journal editors ALL LIFE 3 Figure 1. An increasing number of toxicity studies mention the use of environmentally relevant concentration(s)/dose(s). We analysed the number of PubMed abstracts and titles containing the words ‘environmentally relevant concentration’ or ‘environmentally relevant dose’ among the abstracts containing the word ‘toxicity’ (Performed on July 27st, 2022). in the hope of advancing understanding of how to the soil) in contrast to the delivered and absorbed design, carry out, interpret and explain the real-world doseofthe herbicideinatestorganism, or aper- significance of both old and new lines of scientific sonoranorganisminthe real world. Herbicidelev- inquiry. els in biomonitoring studies are static concentrations expressed usually as mg orµg of the herbicide per kilo- gram of the food, water, biou fl id, or tissue in which the Essential elements of an ‘environmentally herbicide is detected. Levels can charge quickly over relevant’ and/or ‘acceptable’ level of exposure time. Estimating exposure levels in an organism from biomonitoring results requires pharmacokinetic stud- Forthe sake of clarity, ourdiscussionhereinfocuses on ies that accurately account for a chemical’s intake and herbicide applications, herbicide exposures to humans rates of absorption, metabolism and excretion. occurring as a result, and associated risks of adverse Typical, real-world herbicide exposures occur consequences in people. The basic concepts apply to through four routes: food, drinking water and other all types of pesticides and chemicals, different organ- beverages, dermal absorption and inhalation. A her- isms and a variety of adverse health or ecological bicide dose level via any of these routes, or cumulative impacts. exposuresisafunctionof: Many authors describe their herbicide dose levels, or dosage regime as ‘environmentally relevant.’ On • The person’s bodyweight, what basisdotheymakethisassertion?Theytypi- • The amount of food and beverages ingested, cou- callydosoonthe basisoftheir belief that thelow-end pled with the concentration of the herbicide in all levels or doses in their experimental work are compa- ingested foods and beverages, rable to at least high-end levels of herbicide exposure • The amount of formulated herbicide falling on a experienced by humans in the real world. person’s skin and/or There are important and often significant differ- • The herbicide’s concentration (if any) in the air ences between the level of a herbicide detected in inhaled over the course of a day. some compartment of the environment (water, food, 4 R. MESNAGE AND C. BENBROOK To normalize dose levels across people of different twelve participants consumed a meal with a known sizes, or from experimental animal studies to humans, amount of glyphosate, only 1% of the glyphosate dose the above four routes of herbicide exposure are usually was excreted in urine, suggesting that ‘typical’ absorp- reported as mg of herbicide per kilogram of body- tion estimations applicable to dietary exposures (20% weight per day, or the familiar mg/kg/day. It is innacu- to 30%) may be markedly overestimated (Zoller et al. rate to refer to an experiment’s ‘dose rate’ and ‘levels’ 2020). Further studies need to be performed to under- interchangeably. Likewise, important differences occur stand if glyphosate intestinal absorption becomes pro- between the total delivered dose in an experiment or a gressively saturated at incrementally higher doses. In real-world setting, and the absorbed dose.For exam- addition, all these studies used gavage and thus have ple, metabolism studies show that, in general, between limited value in understanding absorption for dietary 60% and 90% of ingested glyphosate passes through exposure scenarios. The difference between delivered most mammals unabsorbed via faecal matter. and absorbed doses of glyphosate in animal feeding On the other hand, any herbicide measured in a studies, coupled with technical glyphosate’s low acute person’s blood or urine had to enter the body and toxicity, accounts for the very high ‘Maximum Toler- bloodstream somehow and most will ultimately leave ated Dose’ (MTD) in most of glyphosate’s two-year the body primarily via urine. Exceptions can occur as chronic feeding studies in rats and mice (i.e. high-dose a result of metabolism and/or binding in certain tis- levels in animal feed in the 10,000 ppm to 50,000 ppm sues (e.g. bone marrow in the case of glyphosate, or range). paraquat that reachesthe brainand canstayinthe Likewise, the spatial and/or temporal distribution of brain for a long time). For these reasons, it is impor- measured levels of herbicides in the environment, or tant to accurately report how a delivered dose enters in food and beverages, often vary significantly. Hence, thebody, becausethe routeofexposuredrivesthe rela- the delivered doses to organisms living within a partic- tionship between total delivered and absorbed dose, as ular ecosystem, including people, u fl ctuate and some- well as the tissues or organs most likely to be impacted. times by wide margins. For example, herbicide levels In the case of mammalian exposures to glyphosate in drinking water drawn from surface water sources andGBHs, it is generallyassumedthatabout 20% tend to spike during peak herbicide-spray seasons of the glyphosate ingested via food or beverages is and decline sharply over the winter (Kruć-Fijałkowska absorbed; about 3% of the glyphosate landing on skin et al. 2022). Dietary exposures tend to be more stable is absorbed but significant uncertainty persists in the over time andvaryasafunctionofdietary choicesand accuracy and applicability of these estimates despite percent of organically grown food in a person’s diet thefactthatfor over 20 yearsglyphosateand GBHs (Curl et al. 2015). have been by far the most heavily used pesticide glob- The physical and chemical properties of formu- ally (Benbrook 2016). lated pesticides products can also impact the persis- There are other complications in accurately quan- tence of the herbicide in various environmental media tifying and reporting total exposure levels, delivered (Kucharski and Sadowski 2011). These properties does and absorbed doses, especially in the case of impact whether the herbicide bioaccumulates along dietary exposures. For example, the portion of the certain food chains, the rate of dermal penetration, and glyphosate in the food and beverages a person ingests whether the herbicide is slowly or rapidly metabolized in a day that is absorbed as it passes through the GI (Brand and Mueller 2002). All of the above factors and tract is a function of the total amount of glyphosate properties must be taken into account in determining ingested, the health of the person’s GI tract, and his the differences between delivered and absorbed doses. or her microbiome (Mesnage, Teixeira et al. 2021; Mesnage, Calatayud et al. 2022). A total of 7 phar- What makes a herbicide dose or level in the macokinetic studies were performed and submitted to environment ‘acceptable’ or purportedly (and EFSA in order to obtain the extension of glyphosate hopefully) ‘safe’? registrations (EFSA 2015). Thereisalargevariation between the different studies, but very few studies were A variety of health-based regulatory benchmark val- done with low doses corresponding to what the human ueshavebeenderived basedmostlyonthe results population is exposed to. In a recent study where of registrant-submitted animal studies. The maximum ALL LIFE 5 level of total herbicide exposure that regulators regard additional safety factor to accommodate, for example, as hopefully‘safe’iscalledanAcceptable DailyIntake heightened risk of harm during pregnancy, or failure (ADI) in Europe and a Reference Dose or Popula- of a toxicology study to produce a NOAEL (i.e. the tion Adjusted Dose in the U.S. This value is gener- LOAEL is the lowest dose tested). ally derived from laboratory animal toxicity studies Most ADIs and RfDs governing general population conducted according to OECD or U.S. Environmental dietary and water-based exposures are set based on Protection Agency (EPA) guidelines. chronic risks (cADI, cRfD) following exposure to pure The number of incrementally higher doses tested is active ingredient. To the extent surfactants and other an important parameter for toxicology studies used in coformulants are no longer present on or in food and human-health risk assessment. Experimental designs beverages when eaten, this focus on just active ingredi- should ideally support assessment of whether there ents is justified. Since cADI/cRfD exposure thresholds is a dose–response relationship in observed adverse are derived from animal experiments typically con- health impacts. In the case of adverse eect ff s thought ducted with no more than three dose levels, there is to adhere to a linear dose–response relationship, a uncertainty in the determination of the dose–response minimum of three doses is generally recommended relationship (Davis et al. 2011). More recently, Bench- in OECD and EPA guidelines. However, the dose mark Doses (BMD) have become the preferred guid- response of chemicals causing toxic effects through ance values to derive human-health toxicity values. altered hormone metabolism is not always linear, TheBMDisaprobabilisticapproachthatconsidersthe andhence requiresahighernumberofdoselev- shapeofthe dose–response curvetocalculate aBMDL els. This was the case of the CLARITY-BPA Core (‘benchmark dose lower confidence limit’). The BMD Study which administered 5 doses of bisphenol A to is essentially the hopefully ‘safe’ toxicity threshold. For Sprague–Dawley rats in order to clarify the existence example, a BMD5 would be the dose leading to an esti- of health effects at low doses (Vandenberg et al. 2019). mated 5% chance of an adverse impact, typically with Dose levels arenot theonlyimportant parameterin 95% certainty. an animal feeding study. The route of administration BMDLs are often more reliable than cADIs/cRfDs to test animals should be selected to reefl ct, as fully that are set on the basis of NOAELs/LOAELs. This is as possible, how humans will likely be exposed to the because the calculation of BMDLs is less dependent same chemical (Tudi et al. 2022). This is why stud- on dose-regime selection and sample size (Davis et al. ies deploying intraperitoneal or gavage dosing some- 2011). Although the BMDL is becoming the preferred times produce results that are questioned (Vandenberg approach within both U.S. and EU regulatory authori- et al. 2014). As a general rule, animal studies designed ties, exposures to most herbicides are still managed to to mimic real-world exposures should be conducted remain below applicable cADIs/cRfDs. using the same or a similar real-world routes of expo- While cADIs/cRfDs set the maximum amount of a sure, as well as formulated product instead of pure herbicide that a person can be exposed to in a day while active ingredients, especially in the case of dermal and meeting health-protection goals (i.e. in the US, ‘a rea- inhalation exposures. For example, the rat study of sonable certainty of no harm’), regulators also establish glyphosate and GBH dermal absorption and impacts Maximum Contamination Levels (MCLs) for herbi- on cancer by George et al. utilized an innovative model cides and other chemicals in drinking water. MCLs and dosing regime that maximized the study’s rele- are set in a similar way as cADIs/cRfDs, and strive to vance in assessment of human applicator exposure and assure that herbicide exposures via drinking water are risk levels (George et al. 2010). notlikelytopushanindividualoverhisorherpersonal Regulators search for the lowest-observed-adverse- cADI/cRfD. effect level (LOAEL) across all available toxicity stud- ies. The next-lower dose that does not cause the The role of tolerance levels and maximum residue adverse eeff ct associated with the LOAEL is called the levels (MRLs) no-observed-adverse-effect level (NOAEL). By divid- ing the NOAEL by a standard safety factor of 100, Expected dietary exposures to herbicides and other regulators estimate a herbicide’s ADI/RfD. In some pesticides are governed by the type and number of cases under U.S. law and EU policy, regulators add an crops a herbicide is registered for use on, when the 6 R. MESNAGE AND C. BENBROOK herbicide is applied in the crop growth cycle, rates and required by regulators prior to setting a pesticide’s number of applications and allowable tolerance lev- ADI/RfD/BMD/MCL. Such benchmark exposure lev- els (also referred to as Maximum Residue Levels, or els are hopefully ‘safe’ for humans, but are ‘safe’ only MRLs). Tolerances/MRLs govern the maximum, legal to the extent that the existing studies on a pesticide’s concentration of a pesticide that can be present in or active ingredient are capable of detecting all possible on a given food item. Under US law and policy, toler- adverse impacts among people exposed to formulated ances are set to cover ∼ 125% of the maximum residue pesticide products. Such an assumption, or assertion, level expected in the harvested crop when it leaves the requires a significant leap of faith. farm. Field studies are carried out in multiple locations Figure 2 displays how many of the above factors deploying maximum label rates in order to quantify impact the relationship between measured levels of the likely maximum level of residue that is likely to glyphosate in the environment, in contrast to regula- be present in a crop sprayed with a legal, on-label tory benchmarks. application. While some benchmarks are generally accepted in Accordingly in the US, tolerances are set based on terms of glyphosate concentrations in different media, unavoidable residue levels in food at harvest. Prior to including people’s bodies and spray solution, the rela- approval of a tolerance, the US EPA must confirm that tionship between typical concentrations and delivered an analytical method is available to detect the pesticide doses is complex and dependent on many factors at the proposed tolerance level, and that the tolerance unique to how a pesticide is applied or ingested and is ‘supported’ by the EPA’s existing evaluation of total the health status of the exposed individual. dietary exposure levels relative to the pesticide’s cRfD or BMDL. However, many tolerances have remained Biomonitoring and human exposure assessment on thebooks formanyyears,ifnot decades, and were initially based on high-rate pesticide-use patterns Measuring the presence of environmental pollutants sanctioned by then EPA-approved labels. But US label in human biological uids fl is a significant techni- directions and typical use patterns change over time. calchallenge.Thisonlybecamepossibleinthe last There are typically large differences between maxi- 20–30 years via major progress in the performance mum, alloweduse of apesticide onagiven crop and of analytical techniques such as mass spectrometry how the pesticide is actually used. Moreover, the way a and nuclear magnetic resonance. Recent advances in pesticide is used in the US or EU often differs markedly Gas Chromatography coupled to Tandem Mass Spec- from the way it is used elsewhere. As a result, hundreds trometry (GC/MS/MS) allow the measurement of of EPA-established tolerances exceed actual residue blood concentrations for 60 persistent organic pol- levels in food at harvest by an order of magnitude or lutants, including polychlorinated biphenyls (PCBs), more. But in many cases, the EPA would not regard polybrominated diphenyl ethers (PBDEs), polycyclic residues in food at or near the tolerance level as ‘safe,’ aromatic hydrocarbons (PAHs), dioxins/furans, as well but since such elevated levels are rarely if ever detected, as some pesticides (Macherone et al. 2015). the agency places a low priority on revisiting the toler- Although pesticide exposures have been linked to ances or lowering or revoking them. For this reason, health effects after residential and occupational expo- it is not appropriate to assert that all published toler- sures (Cognez et al. 2019;Sagiv et al. 2019), few ancesare setat‘safe’levelsunder US law(Benbrook et studieshavefocused on population-wideexposures al. 2021). In addition, such tolerances allow high-risk andriskoutcomes. Anotable exceptionisthe herbi- residues to be present in food imported to the US and cide biomonitoring carried out as part of the National can place US growers at a competitive disadvantage Health and Nutrition Examination Survey (NHANES) when farmers abroad can continue to use high-risk performed by the US Centers for Disease Control and pesticides in ways not allowed on US product labels. Prevention (CDC). NHANES collects urine and blood It is also important to acknowledge uncertainties samples from several thousand people on a recurring embedded in ADIs, RfDs, BMD, MCLs and tolerances cycle, along with extensive demographic, dietary and governing residues in food. For food-use pesticides, health data (https://www.cdc.gov/exposurereport/). approximately a dozen sub-chronic and chronic tox- Over the last two decades, NHANES has peri- icity studies conducted mostly in rats and mice are odically tested urine for 17 sulfonylurea herbicides, ALL LIFE 7 Figure 2. Defining environmentally relevant doses or concentrations for glyphosate. High urinary concentrations correspond to the levels encountered in applicator studies, but also occasionally in some biomonitoring studies of the general population as in the recent National Health and Nutrition Examination Survey (CDC 2022) For the general public, estimated high-end environmental expo- sures originate mostly from dietary exposure. Exposures of bystanders or residents are poorly characterized. Environmentally relevant concentrations refers to the glyphosate concentrations which are typically found in the food chain or in environmental samples. atrazine,2,4-D,aswellaspyrethroid, organophos- By contrast, the monitoring of hair (Appenzeller phorus insecticide and carbamate metabolites. For et al. 2017;Grundleretal. 2021), adiposetissue(Jack- 2,4-D, urinary levels are available for samples col- son et al. 2017) and meconium (Ostrea et al. 2008) lected in 1999–2000 and 2001–2002 by the NHANES. are increasingly used to provide insights on cumula- Approximately a quarter of the population studied tive exposure. A notable example is a recent French had 2,4-D urinary levels below the limit of detection study measuring pesticides in mothers’ hair samples (0.2 μg/L), with the most exposed individuals having and children’s measurements at birth (Beranger et al. levels around 1 μg/L. The 95th percentile for 2,4-D uri- 2020). Among the 64 compounds measured, numer- nary concentrations was 1.24–1.55 μg/L depending on ous associations between maternal hair concentrations theage groupconsidered. Thereweresignificantasso- and birth measurements were statistically significant. ciations between crop application of 2,4-D and the per An extensivelistofbiologicalmatricesusableatdieff r- cent of NHANES participants with detectable levels of ent life stages to track exposures is available (Barr et al. 2,4-D in urine (Freisthler et al. 2022). 2005). In Europe, the European Human Biomonitoring Initiative (project HBM4EU) is the largest study of Health risk evaluations from human biomonitoring human exposure to chemicals in Europe. It strives to incorporate biomonitoring data in the chemical risk- Human biomonitoring data can be used directly to assessment process (Louro et al. 2019). More recently, estimate health outcomes in human populations. A pyrethroids, chlorpyrifos, dimethoate, glyphosate and classic example is the increase in lifetime mean blood p fi ronil were also included in the priority substance lead levels from 1 to 10 µg/dL, an increase associ- list. Interestingly, two pesticide coformulants (piper- ated with a reduction of 7.4 IQ points (Lanphear et al. onyl butoxide and ethoxylated tallowamine) were also 2005). However, scenarios where both the health out- included in the HBM4EU priority list. Some inde- come and exposure levels can be studied in a human pendent studies were also performed to evaluate the population are uncommon. Health-based guidance exposure to these active ingredients (Mesnage, Bowyer values allowing the estimation of health risk are gener- et al. 2022), or develop biomonitoring methods for ally based on toxicity tests performed in rodents before their coformulants (Mesnage, Mazzacuva et al. 2021). a product is released on the market. 8 R. MESNAGE AND C. BENBROOK Figure 3. Overview of the chronic health risk assessment of dietary exposures to glyphosate – switching from biomonitoring to risk assessment. Considering that the high-end of glyphosate urinary levels resulting from environmental exposures is approximately 7µg/L (Gillezeau et al. 2019), it can be estimated that environmental exposures to glyphosate result in an internal dose of 0.2 µg/kg/bw and an external dose of 1 µg/kg/bw. Glyphosate concentrations cannot be considered as environmentally relevant if they are far higher than these doses, except in the case of relatively high applicator or occupational exposures. When the cADI/cRfD is used to perform health 7μg/Lglyphosateismeasuredinurine andthe per- risk evaluations, biomonitoring equivalent (BE) values sonexcretes2Lofurine perday (EFSA 2012), daily are sometimes calculated in order to compare ‘envi- excretion of glyphosate would be 14 μg per day via ronmental’ doses in humans to references doses in urine. Metabolism studies suggest that at least 80% rodents. In this context, ‘environmental’ refers to ‘real- of ingested glyphosate passes through the GI tract world’ exposure levels based on actual herbicide levels unabsorbed, exiting the body in faeces (Figure 3). In in food, water, the air and spray solution. A BE level addition, another 2% to 5% of ingested glyphosate is the concentration of a given chemical in urine or is likely retained beyond 24 h in the body, mostly in plasma whichisexpectedwhenanaverage individ- bone marrow, liver and kidney tissues. Glyphosate ual is exposed to the cADI/cRfD, see (Aylward and does not bioaccumulate in mammals and metabolism Hays 2008) for a 2,4-D example. Estimating the inter- is minimal given how quickly glyphosate exits the nal dose in humans requires pharmacokinetic knowl- body, although some studies suggest that glyphosate edge to estimate the delivered dose required to pro- canbemetabolized by thegut microbiome (Mes- duce agiven levelinurine or blood.Whenthe inter- nage and Antoniou 2020;Mesnage,Calatayud et al. nal dose reaching the bloodstream (or a target tissue) 2022). of a pesticide is estimated from urinary concentra- A70kgpersonwould have an internal doseof tions. It is necessary to incorporate average human 0.2 μg/kg body weight (internal dose = daily excre- physiological data in the model such as body weight, tion [14 μg] / body weight [70 kg]) (Figure 3). The food or water consumption or urine excretion (EFSA corresponding external dose, based on the calcula- 2012). tions described above, is 1 μg/kg body weight. This Health risk can be estimated from the results of ani- is because it is estimated that 80% of a glyphosate mal studies even when information is not sufficient dose that enters the body via food/beverages passes to adequately derive BE values. In a recent review of through unabsorbed and exists the body in faeces. The glyphosate biomonitoring studies, it was found that 14 μg of glyphosate in urine is thus likely to origi- the highest glyphosate urinary levels detected were nate from an exposure to 70 μg glyphosate [70 μgx 7µg/L (Gillezeau et al. 2019). If a concentration of 0.2]. However, recent studies suggest that for dietary ALL LIFE 9 exposures, as little as 1% of the delivered dose of Pesticide exposure levels are modelled using the glyphosate was excreted in urine (Zoller et al. 2020). In median residue value derived from residue trials, or this case the estimated intake from human urine data maximum residues limits. These levels of residues is 1400 μg glyphosate [70 × 0.01]. The acceptable daily reflect the concentration in pesticide residues which intake of glyphosate set by EFSA is 500 μg/kg body are expected in a crop cultivated according to good weight per day, suggesting a 500-fold margin of safety agricultural practices. They can thus be combined between current levels of exposure and the applica- with national commodity intake levels to estimate the ble EFSA cADI. However, this large margin-of-safety dailyintakeofagivenactivesubstance,asdonewith depends on glyphosate’s relatively high cADI/cRfD of the pesticide residue data collected by the UK-FSA 500 µg/kg/day in Europe (currently 1.0 mg/kg/day in and USDA via the Dietary Risk Index (DRI) system the US). (Benbrook and Davis 2020). These values can also be If a new toxicology study emerges that leads EFSA refined across food forms by considering data on the to reduce glyphosate’s cADI to 5 µg/kg/day, the 500- typical impact of food processing on residue levels fold safety margin would shrink to just 5-fold. Recent in fresh food forms, coupled with available monitor- animal studies assessing the role of glyphosate-based ing data on various food forms (e.g. fresh, frozen, herbicide exposures in triggering non-alcoholic fatty canned, dried, juice, sauce). For glyphosate, Interna- liver disease suggest that a cADI reduction well below tional Estimates of Daily Intake (IEDI) (in % ADI) for 5µg/kg/day might actually be necessary to fully pro- theWHOGEMS/Food17ClusterDietsrangebetween tect against damage to the liver (Mesnage et al. 2017). 1.7 and 4.9 μg/kg bw/day in a recent analysis (Stephen- Other adverse health effects yet to be associated with son and Harris 2016). In this case, the highest con- GBH use and exposures may be recognized. For exam- tributor to the total glyphosate intake was barley in ple, birthcohortstudies focusedonglyphosateand its the Irish diet. It is estimated that the Irish population primary breakdown product aminomethylphosphonic consumes on average 65 grams of barley in breakfast acid (AMPA) have reported associations with health cerealswhich maybecontaminatedby5.85mg/kg of outcomes such as breast cancer, aging, or preterm glyphosate (Stephenson and Harris 2016). Barley con- birth, at real-world levels of exposure occurring in sev- sumption in Ireland results in a glyphosate daily intake eral areas (Franke et al. 2021;Silveretal. 2021; Lucia of 0.38 μg/kg bw/day. Rachel et al. 2022). It is widely recognized that for many chemicals, adverse impacts on reproduction and Challenges in the design and conduct of children’s development occur at the lowest dose levels pesticide studies designed to inform risk across all other endpoints (Council 1993). assessment and promote public health Guidelines exist to ensure that toxicity studies in ani- From dietary exposures to internal doses mals and cell assay systems are technically sound and When biomonitoring data is not available, it is possi- reproducible. But there are no formal or verified meth- ble to estimate pesticide dietary exposures using stan- ods to determine whether such studies are likely to dard daily intake levels for a variety of foodstusff for detect the adverse effect on humans likely to occur whichcontaminating levels areknown.Thisisper- at the lowest intake level, especially among vulner- formed in Europe using the EFSA pesticide Residues able and/or heavily exposed populations. Regulatory Intake Model PRIMO model (Brancato et al. 2018). studies using doses several orders of magnitude higher PRIMO estimates pesticide external doses for a variety than dietary doses, and unrealistic routes of exposures of dietary patterns. such as gavage continue to support nearly all pesticide Dietary risk assessment methods for pesticides are dietary risk assessments. comparable in the US. Chronic exposures rely on food- Concentrations used in tissue culture experiments consumption data from the NHANES ‘What We Eat cannot be directly compared to doses arising from in America’ survey (Steinfeldt et al. 2019). In PRIMO, concentrations in the environment. A variety of factors food intake levels are derived from WHO cluster diets, drive the degree to which concentrations in formu- as well as diets considered representative of different lated products, spray solutions, food, water, soil or air EU countries, for adults, toddlers or infants. translate into delivered doses to an individual. Another 10 R. MESNAGE AND C. BENBROOK set of complex factors then drive blood and tis- have to drink approximately one litre of his 10 g/L sue concentrations, longevity of exposures and health spray to absorb 2 g of glyphosate (assuming 20% outcomes. intestinal absorption), which would be diluted to a Evaluating the health eeff cts of pesticides in differ- concentration of 0.036 g/L glyphosate at the tissue level ent scenarios requires a good understanding of the assuming that glyphosate distributes evenly in differ- concept of dose, as well as the differences between a ent tissues. dose level and a concentration. In brief, a concentra- Academic scientists are also free to create their own tion expresses the mass of molecules in a volume (e.g. setofrules to definethe relevanceoftheir results. grams per litre), or the number of molecules in a vol- Authors sometimes claim that a study is relevant for ume of liquid (e.g. moles per litre). A dose is typically public-health assessment in the hope such relevance normalized to take account of the weight of a given will enhance the odds of favourable passage through individual, the level in food or beverages consumed peer review. For instance, a recent glyphosate toxicity and the duration of the exposure (e.g. grams per kilo- study which tested 0.05% concentration of a Roundup gram body weight perday). As aconsequence,agiven formulation on breast cancer cell lines concluded that dose can be achieved by eating or drinking different Roundup altered cell cycle and DNA repair ‘at much concentrations of chemicals depending on the body lower doses than the ones used in agriculture’ (Stur weight of the exposed individuals. In some cases, like et al. 2019). This studycannotbeusedtoinformreg- for multi-component mixtures, it is also common to ulators on the eeff cts of glyphosate on the mammary express a given level as a dilution (volume per vol- gland because the authors do not discuss expected ume, v/v). This v/v unit can also be expressed as a glyphosate concentrations in the mammary gland after percentage (%) or as a part-per-million (1 ppm being expected, real-world exposures. A recent study inves- a dilution of one in a million) for smaller quantities. tigated for the first time the transgenerational inheri- Timing is also important. It is common to normalize tance of glyphosate induced-obesity, prostate, kidney exposure to daily intakes. Most references doses such and ovarian disease in rats (Kubsad et al. 2019). The as theacceptable dailyintakeare expressedasadaily authors claimed that they used ‘an environmentally rel- dose (e.g. mg/kg bw/day). evant exposure’ by administering daily intraperitoneal Although this is common knowledge for toxicol- injections of glyphosate (25 mg/kg bw/day). This dose ogists, concentrations and doses are sometimes not is at least 10,000 times too high to be qualified as accurately communicated when authors with expertise ‘environmentally relevant’ following exposures via the in specific test methods, but not general toxicologi- human diet. In addition, intraperitoneal injections are cal principles and methods, test pesticides, describe not regarded as comparable to environmental expo- their protocols and report experimental findings. For sure scenarios. instance, In a glyphosate toxicity study (de Liz Oliveira Other animal studies have used intraperitoneal Cavalli et al. 2013), authors indicated that glyphosate administrations to assess the effects of transgener- was tested at a low dose of 36 ppm in vitro,which ational exposures which are more likely to occur could suggest that they tested a dose of 36 mg/kg through oral exposures (Kubsad et al. 2019). Although body weight or a concentration of 36 ppm in some the use of this type of experimental design can be medium to which the test organism was exposed. A suitable to study biochemical mechanisms, they are concentration of 0.036 g/L was actually tested in an in less valuable when conducting health risk assessments. vitro system, and so does not reflect a daily dose of Overall, few studies are performed where the exper- 36 mg/kg/day. Concentrations used in tissue cultures imental design is tailored to specific exposure scenar- experiments cannot be directly compared to doses in ios. In the next sections, we recommend a set of criteria animal studies and concentrations in the environment. which can be used in scientific studies to enhance their The authors indicated that ‘It is important to empha- relevance. size that Roundup is used in agricultural work at dilu- The concentration of a pesticide ingredient in an tions ranging from 10,000 to 20,000 ppm (10 to 20 g/L), experimental system that can be regarded as envi- concentrations much higher than those described in ronmentally relevant will differ between application our results.’ If the glyphosate was ingested, an and exposure scenarios. Exposure levels are higher agricultural worker with a body weight of 70 kg would among human populations living in intensive cropping ALL LIFE 11 areas or urban habitats where pesticides are frequently be used to claimthateeff ctsat0.1μg/L or aboveare a used. It is also organism dependent. Environmentally source of health risks. relevant concentrations will be dieff rent for worms and Environmental pollutants such as pesticides are soil fungi in agricultural fields regularly sprayed with found at very low concentrations in human biological pesticides compared to untreated cropland, different u fl ids. While food components, drugs, or endogenous among organisms at various branches in the tree of compounds are found in blood at concentrations in life and different depending on the adverse impact of the range of the μM level, pollutant concentrations are concern. generally found at the nM level (Rappaport et al. 2014). Caution must be exercised in attempts to draw haz- We estimated in the previous sections that a realistic ard or risk conclusions from in vitro cell assays. In dose in studies aiming to mimic environmental expo- order to characterize their dosage levels as ‘environ- sure scenarios for glyphosate would be below 10 μg/kg mentally relevant,’ some authors of in vitro studies bw/day. Such exposures would be predominantly via erroneously compare their dose levels (e.g. 1 g/L) to food and drinking water. Further research is needed to the recommended 10–20 g/L spray solution concen- estimate comparable ‘environmentally relevant’ dose trations on product labels. In a study aiming to char- levels in thecaseofdermaland inhalation exposures acterize theeeff ctsofglyphosateonliver cells, agroup among those handling and spraying GBHs, and espe- from Italy exposed hepatoma tissue culture (HTC) cially among those spraying a GBH for several hours cellsto1–10mMRoundup.Theauthorsqualifiedtheir a day for many days per year over many years using dosage as ‘very low concentrations of Roundup.’ How- small-scale application equipment. ever, calculations using standard conversion factors A surprisingly common mistake is failure to fully reveal that a rat would need to be exposed to a diet and carefully describe the test substance. For example, contaminated by approximately 10 g/kg glyphosate to some authors just report a test involving glyphosate be exposed to an internal concentration of 1–10 mM withoutclarifyingwhatchemicalformwas secured, or Roundup (Malatesta et al. 2008). whether a pure active ingredient or formulated prod- Some authors also claim that the doses in their uct was used (Sivikova and Dianovsky 2006;Chanetal. studies correspond to ‘environmental relevant’ levels 2007;Hokansonetal. 2007). Ideally, authors should because they tested permitted levels like the EFSA state not just the precise chemical formulation tested, cADI or a tolerance level. The fact that a concentration but also provide full details on where the product was of glyphosate is allowed in a certain exposure scenario purchased, date of purchase and lot numbers when (i.e. in soybeans) does not mean that the concentra- available. tion is an accurate estimate of delivered or absorbed dose from any or all routes of exposure. In a study Conclusion of glyphosate reprotoxicity (Milesi et al. 2018), the authors indicated that ‘The dose of 2 mg/kg bw/day We identify and describe gaps in the scientific litera- is representative of theglyphosateresiduesfound in ture regarding the concept of environmental relevance soybean grains.’ Although a glyphosate concentration of exposures and doses in laboratory settings. Our aim of 2mg/kg maybefound in soybeans,thisdoesnot is threefold: to explain the many factors that must mean that their consumption will automatically result be taken into account in determining environmen- in a dose of 2 mg/kg bw/day. A typical human being tally relevant levels; second, to improve the ability to (70kg) would need to eat 70kg of soybeans contami- properly translate results from laboratory studies into natedby2mg/kgevery daytoachieve adoselevel of human-health risk assessment; and third, to enhance 2 mg/kg bw/day. In other studies, maximum residue opportunities to compare results across studies using limits are confounded with actual environmental lev- different experimental designs, organisms and routes els. This was the case in a long-term toxicity study of exposure. investigating toxic effects of a GBH at the levels per- Clear definitions and full specification of details on mitted by regulatory authorities in drinking water in test substances, formulations, organisms tested, how the EU (0.1 μg/L) (Seralini et al. 2014). Levels of pesti- doses are delivered and pharmacokinetics are vital to cides allowed in tap water in the EU are not established fullyappreciatethe findingsinagivenstudy.Wealso based on the results of toxicity studies and should not provide an overview of the roll of concentrations and 12 R. MESNAGE AND C. BENBROOK doses in chemical risk assessment, highlighting the from infections by mosquito borne diseases (Southwell emerging use of human biomonitoring data in the et al. 2018). regulatory assessments. It is crucial to indicate in sci- There is equally compelling need to more accurately entific studies what environments the concentrations compare regulatory-set, herbicide-exposure thresh- are coming from, and which organisms and exposure olds to levels of herbicides measured in the environ- routes are encompassed in a given study. We also rec- ment,food,sediment, theair andhuman bioufl ids ommend that Editors and reviewers should pay par- and tissues. Such levels must be tracked to deter- ticular attention to misuse of exposure thresholds in mine whether contemporary and legal herbicide uses defining ‘acceptable’ concentrations in various media are leading to exposures reliably below, roughly equal in light of various routes of exposure. to, or sometimes above ‘acceptable’ regulatory bench- This paper can serve as a point of departure for marks (e.g. cRfDs/ADIs). The path from measured lev- disciplinary experts compiling and vetting recommen- els of herbicides in various media, to estimates of expo- dations to help scientists design and carry out projects, sure that can be compared to ‘acceptable’ doses derived report results and conduct peer-reviews. The goal is to from toxicology studies is fraught with technical com- better understand and more clearly define and approx- plexity. The path is too often clouded by unstated imate ‘environmentally relevant’ levels and doses of assumptions and missing information. Imprecise char- pesticides across the diversity of use patterns, exposure acterization of test substances and dosing methods, scenarios, test systems and human vulnerabilities. routes and levels often complicate interpretation of Important next steps in regions with rising herbi- results. Confusion over the relationship between con- cide use should include focused, clinical studies of centrations, regulatory thresholds and ‘safe’ expo- pregnant women and newborns designed to determine sure levels is all too common and must be overcome whether elevated levels of prenatal herbicide exposures to accelerate progress in identifying those pesticides are associated with heightened risk of adverse birth and exposure scenarios worthy of regulatory inter- outcomes spanning low birth weight, pre-term deliv- ventions and those that do not based on current ery, birth defects and developmental abnormalities. knowledge. The research objectives and protocols of the Heartland Study serve as an example of a clinical study designed Author’s contributions to explore potential, herbicide-induced adverse birth RM and CMB designed and wrote this review. outcomes (Benbrook et al. 2021). Improving science reporting needs to be associated Disclosure statement with an improvement of science translation for pub- CMB has and continues to serve as an expert witness in pes- lic audiences (Uhrig et al. 2020). Erroneous statements ticide litigation, including cases involving Roundup and non- about environmental relevance have been publicized Hodgkinlymphomaand paraquat andParkinsonsdisease.RM in a variety of media, sometimes inadvertently but has served as a consultant on glyphosate risk assessment issues sometimes with intent (Hansen 2016). This has been as part of litigation in theUSoverglyphosatehealtheeff cts. RM well documented for the case of tobacco (Baba et al. is a section editor of the section ‘Environmental Toxicology and 2005). However, this is not unilateral. Pesticide manu- Health’in‘AllLife’. facturers develop strategies to preserve their commer- cial interests while advocate groups deploy strategies Data availability statement to ban these same products. The way scientific find- This manuscript is a review and do not contain experimental ings are presented can also profoundly shape public data. All the results reviewed are from published studies. reactions. Since most adults do not have experience in participating in scientific research activities, and have Ethical statement for most of all limited education on the specialized top- The research presented in this manuscript did not involve any ics (Burke et al. 2022), it is important for scientists to animal or human participants. adopt strategies which improve public acceptance of research information. For instance, mental models of ORCID how a person conceptualizes a phenomenon have been proposed to improve public understanding of the risks Robin Mesnage http://orcid.org/0000-0003-1732-4741 ALL LIFE 13 & Engineering Indicators. NSB-2022-1. National Science References Foundation. AppenzellerBMR,Hardy EM,Grova N, ChataC,FaÿsF, CDC. 2022. National Health and Nutrition Examination Sur- Briand O, Schroeder H, Duca RC. 2017.Hairanalysisfor the vey - Glyphosate (GLYP) - Urine (SSGLYP_H). https://www. biomonitoring of pesticide exposure: comparison with blood ncdcgov/Nchs/Nhanes/2013-2014/SSGLYP_Hhtm. and urine in a rat model. Arch Toxicol. 91(8):2813–2825. Chan YC,Chang SC,Hsuan SL,Chien MS,Lee WC,KangJJ, DOI:10.1007/s00204-016-1910-9 Wang SC, Liao JW. 2007. Cardiovascular eeff cts of herbicides Aylward LL, Hays SM. 2008.Biomonitoring equivalents(BE) and formulated adjuvants on isolated rat aorta and heart. dossier for 2,4-dichlorophenoxyacetic acid (2,4-D) (CAS No. Toxicol In Vitro. 21(4):595–603. DOI:10.1016/j.tiv.2006. 94-75-7). Regul Toxicol Pharmacol. 51(3 Suppl):S37–S48. 12.007 DOI:10.1016/j.yrtph.2008.05.006 Cognez N, Warembourg C, Zaros C, Metten MA, Bouvier G, Baba A, Cook DM, McGarity TO, Bero LA. 2005.Legislating Garlantézec R, Charles MA, Béranger R, Chevrier C. 2019. “sound science": the role of the tobacco industry. Am J Pub- Residential sources of pesticide exposure during pregnancy lic Health. 95(Suppl 1):S20–S27. DOI:10.2105/ajph.2004. and the risks of hypospadias and cryptorchidism: the French ELFE birth cohort. Occup Environ Med. 76(9):672–679. Barr DB, Wang RY, Needham LL. 2005. Biologic monitor- DOI:10.1136/oemed-2019-105801 ing of exposure to environmental chemicals throughout Council NR. 1993.Pesticidesinthe dietsofinfants andchil- the life stages: requirements and issues for consideration dren. for the national children’s study. Environ Health Perspect. Curl CL, Beresford SA, Fenske RA, Fitzpatrick AL, Lu C, Net- 113(8):1083–1091. DOI:10.1289/ehp.7617 tleton JA, Kaufman JD. 2015. Estimating pesticide exposure Benbrook C, Perry MJ, Belpoggi F, Landrigan PJ, Perro M, from dietary intake and organic food choices: the Multi- Mandrioli D, Antoniou MN, Winchester P, Mesnage R. Ethnic Study of Atherosclerosis (MESA). Environ Health 2021. Commentary: novel strategies and new tools to cur- Perspect. 123(5):475–483. DOI:10.1289/ehp.1408197 tail the health eeff cts of pesticides. Environ Health. 20(1):87. Dallmann R, Okyar A, Lévi F. 2016. Dosing-time makes the poi- DOI:10.1186/s12940-021-00773-4 son: circadian regulation and pharmacotherapy. Trends Mol Benbrook CM. 2016. Trends in glyphosate herbicide use in Med. 22(5):430–445. the United States and globally. Environ Sci Eur 28(1):3. Davis JA, Gift JS, Zhao QJ. 2011.Introductiontobench- DOI:10.1186/s12302-016-0070-0 mark dose methods and U.S. EPA’s Benchmark Dose Benbrook CM. 2019.How didthe US EPAand IARC Software (BMDS) version 2.1.1. Toxicol Appl Pharmacol. reach diametrically opposed conclusions on the genotoxic- 254(2):181–191. DOI:10.1016/j.taap.2010.10.016 ity of glyphosate-based herbicides? Environ Sci Eur. 31(1):2. de Liz Oliveira Cavalli VL, Cattani D, Heinz Rieg CE, DOI:10.1186/s12302-018-0184-7 Pierozan P, Zanatta L, Benedetti Parisotto E, Wilhelm Filho Benbrook CM, Davis DR. 2020. The dietary risk index sys- D, Mena Barreto Silva FR, Pessoa-Pureur R, Zamoner tem: a tool to track pesticide dietary risks. Environ Health. A. 2013. Roundup disrupts male reproductive functions 19(1):1–18. by triggering calcium-mediated cell death in rat testis Béranger R, Hardy EM, Binter AC, Charles MA, Zaros and Sertoli cells. Free Radical Biol Med. 65:335–346. C, Appenzeller BMR, Chevrier C. 2020. Multiple pesti- DOI:10.1016/j.freeradbiomed.2013.06.043 cides in mothers’ hair samples and children’s measure- EEA. 2013. Late lessons from early warnings: science, precau- ments at birth: results from the French national birth tion, innovation. ISBN 978-92-9213-349-8. DOI:10.2800/ cohort (ELFE). Int J Hyg Environ Health. 223(1):22–33. DOI:10.1016/j.ijheh.2019.10.010 EFSA. 2012. Guidance on selected default values to be used by Bopp SK, Kienzler A, Richarz AN, van der Linden SC, the EFSA Scientific Committee, Scientific Panels and Units PainiA,ParissisN,Worth AP. 2019.Regulatoryassess- in the absence of actual measured data. EFSA J. 10(3):2579. ment and risk management of chemical mixtures: chal- DOI:10.2903/j.efsa.2012.2579 lenges and ways forward. Crit Rev Toxicol. 49(2):174–189. EFSA. 2015. Conclusion on the peer review of the pesticide DOI:10.1080/10408444.2019.1579169 risk assessment of the active substance glyphosate. EFSA J. European Food Safety Authority (EFSA); Brancato A, Brocca 13(11):4302. DOI:10.2903/j.efsa.2015.4302 D, Ferreira L, Greco L, Jarrah S, Leuschner R, Medina P, EFSA. 2022. Guidance on the assessment of exposure of opera- Miron I, Nougadere A, et al. 2018. Use of EFSA pesticide tors, workers, residents and bystanders in risk assessment of residue intake model (EFSA PRIMo revision 3). EFSA J. plant protection products. EFSA J. 20(1):e07032. 16(1):e05147. DOI:10.2903/j.efsa.2018.5147 Festing MF, Vesell ES. 1987. Genetic factors in toxicology: Brand RM, Mueller C. 2002.Transdermal penetrationof implications for toxicological screening. CRC Crit Rev Tox- atrazine, alachlor, and triu fl ralin: effect of formulation. Tox- icol. 18(1):1–26. icol Sci. 68(1):18–23. Franke AA, Li X, Shvetsov YB, Lai JF. 2021.Pilot study BurkeA,OkrentA,HaleK,Gough N. 2022.The StateofUS on the urinary excretion of the glyphosate metabolite Science & Engineering 2022. National Science Board Science aminomethylphosphonic acid and breast cancer risk: The 14 R. MESNAGE AND C. BENBROOK multiethnic cohort study. Environ Pollut. 277:116848. Lanphear BP, Hornung R, Khoury J, Yolton K, Baghurst P, DOI:10.1016/j.envpol.2021.116848 Bellinger DC, Canfield RL, Dietrich KN, Bornschein R, Freisthler MS, Robbins CR, Benbrook CM, Young HA, Greene T, et al. 2005. Low-level environmental lead expo- Haas DM, Winchester PD, Perry MJ. 2022.Association sure and children’s intellectual function: an international between increasing agricultural use of 2,4-D and popula- pooled analysis. Environ Health Perspect. 113(7):894–899. tion biomarkers of exposure: findings from the National DOI:10.1289/ehp.7688 Health and Nutrition Examination Survey, 2001–2014. Env- Louro H, Heinälä M, Bessems J, Buekers J, Vermeire T, iron Health. 21(1):23. DOI:10.1186/s12940-021-00815-x WoutersenM,van EngelenJ,BorgesT,Rousselle C, Ougier George J, Prasad S, Mahmood Z, Shukla Y. 2010.Studies on E, et al. 2019. Human biomonitoring in health risk assess- glyphosate-induced carcinogenicity in mouse skin: a pro- ment in Europe: current practices and recommendations teomic approach. J Proteomics. 73(5):951–964. DOI:10.1016/ for the future. Int J Hyg Environ Health. 222(5):727–737. j.jprot.2009.12.008 DOI:10.1016/j.ijheh.2019.05.009 Gillezeau C, van Gerwen M, Shaer ff RM, Rana I, Zhang Lucia RM, Huang WL, Pathak KV, McGilvrey M, David- L, Sheppard L, Taioli E. 2019.The evidence of human Dirgo V, AlvarezA,GoodmanD,MasunakaI,Odegaard exposure to glyphosate: a review. Environ Health. 18(1):2. AO, Ziogas A, et al. 2022.Association of glyphosate expo- DOI:10.1186/s12940-018-0435-5 sure with blood DNAmethylation inacross-sectional Goumenou M, Djordjevic AB, Vassilopoulou L, Tsatsakis AM. study of postmenopausal women. Environ Health Perspect. 2021. Endocrine disruption and human health risk assess- 130(4):047001. DOI:10.1289/EHP10174 ment in the light of real-life risk simulation. In: Tsatsakis Macherone A, Daniels S, Maggitti A, Churley M, McMullin M, AM, editor. Toxicological risk assessment and multi-system Smith M. 2015. Measuring a slice of the exposome: targeted health impacts from exposure. Elsevier; p. 147–162. GC-MS/MS analysis of persistent organic pollutants POPs) Grundler F, Séralini G-E, Mesnage R, Peynet V, Wilhelmi de in small volumes of human plasma. In: 63rd AMS confer- Toledo F. 2021. Excretion of heavy metals and glyphosate in ence on mass spectrometry and allied topics, St Louis, MO, urine and hair before and after long-term fasting in humans. Abstract TP, Vol. 309. Front Nutr. 8. DOI:10.3389/fnut.2021.708069 Malatesta M, Perdoni F, Santin G, Battistelli S, Muller S, Big- Hansen A. 2016. The changing uses of accuracy in sci- giogera M. 2008. Hepatoma tissue culture (HTC) cells as a ence communication. Public Underst Sci. 25(7):760–774. model for investigating the effects of low concentrations of DOI:10.1177/0963662516636303 herbicide on cell structure and function. Toxicol In Vitro. Hayes AW, Dixon D. 2017. Cornerstones of toxicology. Toxicol 22(8):1853–1860. DOI:10.1016/j.tiv.2008.09.006 Pathol. 45(1):57–63. DOI:10.1177/0192623316675768 Mebane CA, Sumpter JP, Fairbrother A, Augspurger TP, Can- Hokanson R, Fudge R, Chowdhary R, Busbee D. 2007.Alter- field TJ, Goodfellow WL, Guiney PD, LeHuray A, Maltby ation of estrogen-regulated gene expression in human cells L, Mayfield DB, et al. 2019. Scientific integrity issues induced by the agricultural and horticultural herbicide in environmental toxicology and chemistry: improving glyphosate. Hum Exp Toxicol. 26(9):747–752. DOI:10.1177/ research reproducibility, credibility, and transparency. Integr 0960327107083453 Environ Assess Manag. 15(3):320–344. DOI:10.1002/ieam. Jackson E, Shoemaker R, Larian N, Cassis L. 2017.Adi- 4119 pose tissue as a site of toxin accumulation. Compr Physiol. Mesnage R, Antoniou MN. 2020. Computational modelling 7(4):1085–1135. DOI:10.1002/cphy.c160038 provides insight into the effects of glyphosate on the shiki- Kruć-Fijałkowska R, Dragon K, Drożdżyński D, Górski J. 2022. mate pathway in the human gut microbiome. Curr Res Seasonal variation of pesticides in surface water and drink- Toxicol. 1:25–33. DOI:10.1016/j.crtox.2020.04.001 ing water wells in the annual cycle in western Poland, MesnageR,BowyerRCE,ElBalkhiS,Saint-MarcouxF,Gardere and potential health risk assessment. Sci Rep. 12(1):3317. A, Ducarmon QR, Geelen AR, Zwittink RD, Tsoukalas DOI:10.1038/s41598-022-07385-z D, Sarandi E, et al. 2022. Impacts of dietary exposure Kubsad D, Nilsson EE, King SE, Sadler-Riggleman I, Beck to pesticides on faecal microbiome metabolism in adult D, Skinner MK. 2019. Assessment of glyphosate induced twins. Environ Health. 21(1):46. DOI:10.1186/s12940-022- epigenetic transgenerational inheritance of pathologies 00860-0 and sperm epimutations: generational toxicology. Sci Rep. Mesnage R, Calatayud M, Duysburgh C, Marzorati M, 9(1):6372. DOI:10.1038/s41598-019-42860-0 Antoniou MN. 2022.Alterationsininfantgut micro- Kucharski M, Sadowski J. 2011.Behaviour of metazachlor biome composition and metabolism after exposure to applied with additives in soil: laboratory and field studies. glyphosate and roundup and/or a spore-based formula- J Food Agric Environ. 9(3/4 part 2):723–726. tion using the SHIME technology.Gut Microbiome.1–30. Landrigan PJ, Fuller R, Acosta NJR, Adeyi O, Arnold R, Basu DOI:10.1017/gmb.2022.5 NN, Baldé AB, Bertollini R, Bose-O’Reilly S, Bouoff rd JI, Mesnage R, Mazzacuva F, Caldwell A, Halket J, Antoniou MN. et al. 2018. The Lancet Commission on pollution and health. 2021. Urinary excretion of herbicide co-formulants after Lancet. 391(10119):462–512. DOI:10.1016/S0140-6736(17)32 oral exposure to roundup MON 52276 in rats. Environ Res. 345-0 197:111103. DOI:10.1016/j.envres.2021.111103 ALL LIFE 15 aminomethylphosphonic acid (AMPA), and preterm birth: MesnageR,RenneyG,Seralini GE,WardM,AntoniouMN. anestedcase-controlstudy in thePROTECT Cohort 2017. Multiomics reveal non-alcoholic fatty liver disease (Puerto Rico). Environ Health Perspect. 129(5):57011. in rats following chronic exposure to an ultra-low dose DOI:10.1289/ehp7295 of roundup herbicide. Sci Rep. 7:39328. DOI:10.1038/ Sivikova K, Dianovsky J. 2006. Cytogenetic effect of technical srep39328 glyphosate on cultivated bovine peripheral lymphocytes. Int Mesnage R, Séralini G-E. 2018.Editorial:toxicityofpesti- J Hyg Environ Health. 209(1):15–20. DOI:10.1016/j.ijheh. cides on health and environment. Front Public Health. 6. 2005.07.005 DOI:10.3389/fpubh.2018.00268 Southwell BG, Ray SE, Vazquez NN, Ligorria T, Kelly BJ. Mesnage R, Teixeira M, Mandrioli D, Falcioni L, Ducar- 2018. A mental models approach to assessing public under- mon QR, Zwittink RD, Mazzacuva F, Caldwell A, Halket standing of Zika virus, Guatemala. Emerg Infect Dis. J, Amiel C, et al. 2021. Use of shotgun metagenomics 24(5):938–939. DOI:10.3201/eid2405.171570 and metabolomics to evaluate the impact of glyphosate or Steinfeldt L, Goldman J, Moshfegh A. 2019.Usual intake of roundup MON 52276 on the gut microbiota and serum food patterns components by older adults from WWEIA. metabolome of sprague-dawley rats. Environ Health Per- NHANES. 2013–2016. (P18-119-19). Curr Develop Nutr. spect. 129(1):17005. DOI:10.1289/EHP6990 3(Suppl 1). DOI: 10.1093/cdn/nzz039.P18-119-19. MesnageR,ZallerJG. 2021. Herbicides: chemistry, efficacy. In: Stephenson CL, Harris CA. 2016.Anassessmentofdietary MesnageR,ZallerJG, editors. Toxicology,and environmen- exposure to glyphosate using refined deterministic and tal impacts. Elsevier. probabilistic methods. Food Chem Toxicol. 95:28–41. DOI: Milesi MM, Lorenz V, Pacini G, Repetti MR, Demonte 10.1016/j.fct.2016.06.026. LD, Varayoud J, Luque EH. 2018. Perinatal exposure Stur E, Aristizabal-Pachon AF, Peronni KC, Agostini LP, Waigel to a glyphosate-based herbicide impairs female repro- S, Chariker J, Miller DM, Thomas SD, Rezzoug F, Detogni ductive outcomes and induces second-generation adverse RS, et al. 2019. Glyphosate-based herbicides at low doses effects in Wistar rats. Arch Toxicol. 92(8):2629–2643. affect canonical pathways in estrogen positive and nega- DOI:10.1007/s00204-018-2236-6 tive breast cancer cell lines. PLoS One. 14(7):e0219610. Myers JP, Zoeller RT, vom Saal FS. 2009.Aclashofold DOI:10.1371/journal.pone.0219610 and new scientific concepts in toxicity, with important Tudi M, Li H, Li H, Wang L, LyuJ,YangL,TongS,Yu implications for public health. Environ Health Perspect. QJ,RuanHD, AtabilaA,etal., 2022.Exposureroutesand 117(11):1652–1655. health risks associated with pesticide application. Toxics. Ostrea EM Jr, Bielawski DM, Posecion NC Jr, Corrion M, 10:6. DOI:10.3390/toxics10060335 Villanueva-Uy E, Jin Y, Janisse JJ, Ager JW. 2008. A com- Uhrig JD, Lewis MA, Poehlman JA, Southwell BG. 2020. parison of infant hair, cord blood and meconium anal- Beyond evidence reporting: evidence translation in an era ysis to detect fetal exposure to environmental pesticides. of uncertainty. The Medical Care Blog [online publication Environ Res. 106(2):277–283. DOI:10.1016/j.envres.2007. of Medical Care]. https://www.themedicalcareblog.com/ 08.014 evidence-translation/. Rappaport SM, Barupal DK, Wishart D, Vineis P, Scalbert Vandenberg LN, Hunt PA, Gore AC. 2019. Endocrine dis- A. 2014. The blood exposome and its role in discovering ruptorsand thefutureoftoxicologytesting—lessons from causes of disease. Environ Health Perspect. 122(8):769–774. CLARITY–BPA. Nat Rev Endocrinol. 15(6):366–374. DOI:10.1289/ehp.1308015 Vandenberg LN, Welshons WV, Vom Saal FS, Toutain PL, Sagiv SK, Bruno JL, Baker JM, Palzes V, Kogut K, Rauch S, Myers JP. 2014.Shouldoralgavagebeabandoned in toxic- Gunier R, Mora AM, Reiss AL, Eskenazi B. 2019.Prena- ity testing of endocrine disruptors? Environ Health. 13(1):46. tal exposure to organophosphate pesticides and functional DOI:10.1186/1476-069x-13-46 neuroimaging in adolescents living in proximity to pesticide Wang B, Gray G. 2015. Concordance of noncarcinogenic end- application. Proc Natl Acad Sci U S A. 116(37):18347–18356. points in rodent chemical bioassays. Risk Anal. 35(6):1154– DOI:10.1073/pnas.1903940116 1166. DOI:10.1111/risa.12314 Séralini GE, Clair E, Mesnage R, Gress S, Defarge N, Malat- Weltje L, Sumpter JP. 2017. What makes a concentration envi- esta M, Hennequin D, de Vendômois JS. 2014.Republished ronmentally relevant? critique and a proposal. Environ Sci study: long-term toxicity of a roundup herbicide and a Technol. 51(20):11520–11521. DOI:10.1021/acs.est.7b04673 roundup-tolerant genetically modified maize. Environ Sci Zoller O, Rhyn P, Zarn JA, Dudler V. 2020.Urine glyphosate Eur. 26(1):14. DOI:10.1186/s12302-014-0014-5 level as a quantitative biomarker of oral exposure. Int J Silver MK, Fernandez J, Tang J, McDade A, Sabino J, Rosario Hyg Environ Health. 228:113526. DOI:10.1016/j.ijheh.2020. Z, Vélez Vega C, Alshawabkeh A, Cordero JF. 2021.Prena- tal exposure to glyphosate and its environmentaldegradate,
Journal
Frontiers in Life Science
– Taylor & Francis
Published: Dec 31, 2023
Keywords: Environmentally relevant; risk assessment; pesticides; exposure
