By Nimal Chandrasena* and Sunil Wimalawansa** –

An article highlighting a recent paper titled, “Glyphosate & hard water: Is it a deadly CKDu combination for farmers?” by Ulrich et al.[published in Environmental Science & Technology Letters in Sept 2023] [1] was published in Sunday Times on 23 November.

It is surprising to see many shortcomings and misinformation in the original Ulrich et al. article and the uncritical tenor of the newspaper article. To prevent people from being misled, it is necessary to refute false claims in both.

The study above provided no scientific data to substantiate that glyphosate is involved in, or can be implicated as a causative agent in, Chronic Kidney Disease of unknown aetiology (CKDu). The authors of this article neither provided supportive scientific evidence nor did they discuss the major pitfalls of the paper under limitations.

Examples of shortcomings include (1) non-systematic sample selection and faulty handling procedures, (2) falsely assuming that all villagers in NCP‒CKDu affected regions had similar poor water quality, (3) mistaken comparison of water from the southern region, (4) likely errors in detecting glyphosate at the ultra-low levels, which has no relevance to human health, reported without proper analytical quality assurance, (5), failure to provide evidence that glyphosate detected was not an artifact, and (6) failure to have quality assurance and quality control procedures in the field and during measurements essential to establish authenticity and to convince the scientific community of the validity of findings.

The data presented are grossly inadequate and unconvincing. The authors failed to take adequate scientific steps to prove they were detecting glyphosate. A rigorous scientific study with a better study design and methodology would have recognized that the samples obtained were highly heterogeneous (in terms of matrices that would interfere with analyses) as the water samples were from a diverse geographical area.

Glyphosate use was banned in Sri Lanka in 2014. Water samples were collected in October/November 2021 from wells in the so-called ‘CKD affected’ areas and controls from the ‘unaffected area’ from the southern region. It is a well-established scientific fact that sometime after application, glyphosate does not exist in its original form in water or soil.

The herbicide glyphosate is rapidly chelated—firmly bound to cations (such as Ca²⁺ and Mg²⁺ ions) in hard water present in the NCP and becomes insoluble. Similarly, glyphosate quickly binds to similar ions in the soil, which makes it unavailable. Besides, glyphosate breaks down through microbial actions, making the molecule unavailable. Therefore, glyphosate is unlikely to remain intact without change and in measurable amount.

Standard glyphosate analysis always detects some amounts of the main breakdown product (i.e., AMPA). However, in the method used in the study, the so-called ‘ultra-sensitive’ analysis failed to detect and/or report AMPA and even to discuss AMPA non-detection. This makes the analytical results dubious.

In a field sampling exercise of this nature and sensitive laboratory analysis in the USA, the extent of matrix interferences varies considerably from source to source, depending on the nature and diversity of the sampled matrix.

Despite the magnitude and repercussions of the claim, the authors failed to state what steps were taken to ensure that the analytical results were reliable. Unfortunately, the article fails in this regard. It is a fatal scientific error not to provide QA evidence and steps, such as an inter-lab comparison, which would have ensured that the analyst was not detecting false positives and artifacts.

Ours and other field research in Sri Lanka showed that the prevalence of CKD is a heterogeneous and localized water quality issue. Along with others, we have witnessed multiple times that in some villages where over 20% of adults are affected with CKDu, in adjacent villages or the other side of the same village, the prevalence of CKDu is less than 2%. The authors have failed to consider these basic facts in their study— another major flaw in the sampling design.

It is a false assumption to state that a large geographical area, like a province in Sri Lanka, is CKD-affected. The disease is prevalent not in regions but in scattered villages within some affected provinces. Consequently, using the terminology of ‘CKD-affected’ region versus ‘non-affected’ (e.g., Southern region) in water sample collection is a mistake many local scientists are making. No effort has been made to do unbiased ‘random’ sampling, potentially because of a ‘pre-formed’ bias against glyphosate.

The authors seem over-enthusiastic about proving what they believe rather than making conclusions based on data or testing a hypothesis; they failed to follow standard frameworks of good science. In addition to failing to discuss the study’s limitations, the article has many improperly cited references purporting to support the findings.

Just because about half the well-water samples tested in ‘CKDu areas’ had very low detectable but clinically insignificant amounts of glyphosate in water does not support glyphosate as a causative agent (or even associated) of CKDu. The author failed to cite several pertinent scientific publications by other groups, making the paper less credible [2, 3]. These recent articles have convincingly proven that glyphosate and AMPA levels in urine/blood, detected in people even potentially exposed to glyphosate, are almost always well below the EPA-stipulated safe levels implicated in its chronic toxicity.

The authors also failed to answer critical questions—Did the detected glyphosate exist in a free, unchanged form, without chelation? Should the samples be subject to other analytical techniques to detect glyphosate chelation? If chelation had occurred over time, a critical question is – was glyphosate bio-available to affect people via drinking water?

The authors also failed to report whether the study found clinically relevant concentrations (in humans) before making a direct claim implicating glyphosate. The negative effects of glyphosate authors reported in laboratory Zebra fish (using/simulating glyphosate levels in well-water) is over ten-thousand-fold higher exposure-equivalent in humans.

The median concentration (unstated) of glyphosate reported in this study is approximately one µg/L (or 1 ppb). This level is significantly lower than the US EPA and WHO stipulated safety standards and upper limits for glyphosate in drinking water of 500-600 micrograms/L (or parts per billion, ppb) [4]. Therefore, using an astronomical dose exposure of glyphosate is supra-pharmacological, unscientific, unrealistic, and misleading.

Based on this data, it is implausible that a person’s daily water intake would reach glyphosate levels that would harm their health. The authors have not attempted to use the globally accepted standard drinking water quality risk assessment approach (which is based on a person’s daily intake of water by a person). Instead, the study used an inappropriate correlation analysis to link glyphosate with CKDu, resulting in false conclusions.

The article suggests that “multiple” factors (i.e., water hardness, moderately alkaline water, elevated F^– and Vanadium, etc.) might contribute to this complex problem. However, the authors jumped the gun, perhaps due to bias, with a distorted title, implicating ‘glyphosate,’ thus feeding the misinformation narrative. We have reported that CKDu/CKD-CTN is caused by consuming naturally contaminated groundwater, consisting of hard water having a high calcium-to-magnesium ratio (alkaline), with fluoride in the presence of several other factors, including chronic dehydration.

The correlations and statistical analysis used in the study also show thirteen other elements (out of 27 measured) at elevated levels in well-water in CKD-affected areas. Nevertheless, the study ignored these. Based on wrongful methodologies, it is likely that such elements might also have been ‘wrongfully’ implicated, along with glyphosate as ‘causing’ CKD. In our view, the authors drew a long bow by using faulty correlations to unjustly extrapolate results to implicate glyphosate as a cause of CKD. Such a conclusion is scientifically not valid.

As reviewers and editors of other scientific journals, we have reviewed innumerable articles. Surprisingly, the referees of this paper failed to detect the major scientific misconceptions mentioned above. We doubt whether glyphosate was actually detected at all in the wells in Sri Lanka. Even if it was, the levels represented were clinically wholly insignificant.

Notwithstanding the technological errors associated with this study, it simply does not correlate the detected low glyphosate levels to clinically relevant and meaningful concentrations that could cause CKDu even with prolonged exposure. This article does not provide evidence of a cause-and-effect that can be associated with glyphosate or any other component measured by the authors.

We believe there is much more convincing evidence of combining multiple factors, such as water hardness‒elevated Ca²⁺, elevated fluoride levels, alkalinity, and other factors linked with CKDu than glyphosate levels or any heavy metal in stagnant well-water. Other factors, such as insufficient daily water intake by the affected people causing chronic dehydration likely to also be implicated in causing CKDu [5-7].

As made in this study, the harm from uncritical and biased conclusions is that glyphosate is unfairly accused of causing CKDu without scientific evidence. This study has so many flaws; thus, it should not be publicized in newspapers or social media to mislead the public. As discussed in global literature, this is not the first time spurious conclusions have been made against glyphosate as a cause of CKD [8].

We believe that vested interests (politicians and other unscrupulous elements) will likely explore such misinformation, causing further injustice to the affected communities and diverting funds away from providing clean water.

Citations

1. Ulrich, J.C., et al., Glyphosate and Ffuoride in high-hardness drinking water Are positively associated with chronic kidney disease of unknown etiology (CKDu) in Sri Lanka. Environ Sci Technol Lett, 2023. 10: p. 916-923.

2. Soukup, S.T., et al., Glyphosate and AMPA levels in human urine samples and their correlation with food consumption: results of the cross-sectional KarMeN study in Germany. Arch Toxicol, 2020. 94(5): p. 1575-1584.

3. Lemke, N., et al., Glyphosate and aminomethylphosphonic acid (AMPA) in urine of children and adolescents in Germany – Human biomonitoring results of the German Environmental Survey 2014-2017 (GerES V). Environ Int, 2021. 156: p. 106769.

4. USA-EPA, USEPA Office of Ground Water and Drinking Water; 2008; Contaminant Candidate List 3 (CCL3);. 2008, USEPA Office of Ground Water and Drinking Water; http://www/epa.gov/OGWDW/ccl/ccl3.html.

5. Dharma-Wardana, M.W.C., Chronic kidney disease of unknown etiology (CKDu) – The search for causes and the impact of Its politicization. En route to One Health-Medical Geology, ed. P.V. Narasimha M, Vithanage M. Vol. Chapter 7. 2023: Wiley.

6. Wimalawansa, S.J., Escalating chronic kidney diseases of multi-factorial origin in Sri Lanka: causes, solutions, and recommendations. Environ Health Prev Med, 2014. 19(6): p. 375-394.

7. Wimalawansa, S.J., Dissanayake, C.B, Factors Affecting the Environmentally Induced, Chronic Kidney Disease of Unknown Aetiology in Dry Zonal Regions in Tropical Countries—Novel Findings. Environments, 2019. 7(1): p. 1-26.

8. Val Giddings, L. Points to consider: Unsubstantiated claims that glyphosate is linked to kidney disease. 2014 [cited 2024, Nov 16; Available from: https://itif.org/publications/2014/05/19/points-consider-unsubstantiated-claims-glyphosate-linked-kidney-disease/.

*Professor Nimal Chandrasena (formerly, Department of Botany, University of Colombo), Editor-in-Chief WEEDS, Journal of the Asian-Pacific Weed Science Society nimal.chandrasena@gmail.com

**Sunil Wimalawansa, Former Professor of Medicine, Endocrinology, and Human Nutrition: Distinguished University Professor of Medicine, Cardiometabolic and Endocrine Institute, North Brunswick, NJ, USA suniljw@hotmail.com

The authors declare no conflicts of interest.  Our work is not funded by any organization but was purely an academic scrutiny of a dubious claim.