WHO & FAO: Original Study that establishes the glyphosate (Roundup) ADI

Saint Chapelle Paris

Atkinson et al., 1993b.

This is a very important study by Atkinson et. al. (extract below) [1] that determines the safe level in toxicity studies that is then divided by 100 (for a risk margin) to become the ADI, or Acceptable Daily Intake for glyphosate for the World Health Organisation and the FAO following the 2004 glyphosate toxicological evaluations .

Most likely your country is a member of Codex, but you need to research your government database to see if your country subscribes to this ADI recommendation.

I have many questions with this critical study that was funded by Cheminova.

I would encourage a real and proper scientist (unpaid by an agrichemical company) to look at this study and tell me if I am seeing things.  Because the study authors gave the established safe level, the NOAEL at the lowest level of glyphosate administered at this study. That level was 100 mg per kilogram of bodyweight per day.

However when I read this study I can see a list of stuff happening at the same ‘safe’ NOAEL 100 mg level.

NOAEL : No observed adverse effect level.

These effects happened at the level they gave the NOAEL to.  Tell me, are these adverse effects or not?:

  1. Erythrocyte volume fraction and haemoglobin were occasionally increased in males.
  2. An increase in plasma ALP activity occasionally reached statistical significance.
  3. Urine pH tended to be lower in males
  4. absolute weight of the liver was reduced
  5. mean weight of the kidney was reduced in groups of males
  6. cellular alteration of the parotid and mandibular (submaxillary) salivary glands (This alteration was described as the occurrence of hypertrophic and weakly (mandibular gland) or more deeply (parotid gland) basophilic acinar cells)
  7. Neoplasia


The study authors then went on to say ‘In conclusion, administration of glyphosate to rats for 104 weeks produced no evidence of a carcinogenic response.’

Please tell me what is the name for all the responses listed above?  You know, I am not a doctor, but that list doesn’t look safe to me.

We now know that comments such ‘as clear dose–response relationship was lacking’ are unfounded when it comes to the endocrine system.  Toxic effects can be found at low doses but not at higher doses, and different toxic effects can be found at different doses [2,3,4,5,6].  In the complete study below, the authors use this comment again and again to excuse symptoms at lower levels.

I believe that this critical study that apparently gives us the safe level at which we can draw our daily consumption of glyphosate/ Roundup from, simply dismisses effects at that same level, and says the response doesn’t count because the dosage didn’t make sense.  It wasn’t proportional to the dose.  This is flawed.

Following acceptance of the WHO of this study, this ‘safe’ level was divided by 100 to get 1mg per kilogram of bodyweight per day.

Please note that earlier, in 1986, the previous WHO study  (Lankas 1981) used to establish the ADI for glyphosate had stated a lower NOAEL at 31 mg.  At this level  15% of rats were found to have tumours [7].  This was dismissed by the study authors as normal.  These rats are the classic rats used in laboratory tests because they tend to show similar symptoms and results as humans living in industrial countries [8].

It is not normal for 15% of male humans to develop tumours of the testicle at any age (just to let you know).

Thus the lower 1986 level is still questionable.


Why and how was the Atkinson study accepted in the first place?

Who is WHO actually protecting?




Page 127-129  of Pesticide residues in food – 2004    Joint FAO/WHO Meeting on Pesticide Residues  EVALUATIONS  2004  Part II—Toxicological

In a study performed in compliance with the principles of GLP and according to the test guidelines of the US EPA, groups of 85 male and 85 female Sprague-Dawley rats were fed diets containing glyphosate (purity, 98.7–98.9%) at a concentration that was adjusted weekly for the first 12 weeks and every 2 months thereafter to give doses of 0, 100, 300 and 1000mg/kg bw per day for 104 weeks. The doses were selected on the basis of results from a 13-week dietary study of toxicity. Routine analysis of the diet was performed at regular intervals throughout the study, giving acceptable results. Fifty rats of each sex per dose were allocated to the 104-week study of oncogenicity, and 35 rats of each sex per dose were allocated for long-term testing for toxicity. Fifteen rats of each sex per dose from every group testing for toxicity in the long term were killed after 52 weeks, all remaining rats were dosed until scheduled termination after 104 weeks.

Test animals were examined daily for mortalities and clinical signs. Once per week, animals received a detailed clinical examination, with particular regard to palpable masses. Body weight and food consumption were recorded weekly until week 13 of dosing and thereafter every 4 weeks.eks. Ophthalmoscopy examinations were performed on 20 males and 20 females from the control groups and the group receiving the highest dose before initiation of dosing and again during weeks 24, 50 and 102. Blood samples were collected from the orbital sinus of 10 males and 10 females from each group after approximately 14, 25, 51, 78 and 102 weeks. Blood samples were analysed for eight haematology parameters and nineteen clinical chemistry parameters.

At study termination, all surviving animals (and the premature decedents, if possible) were sacrificed and necropsied. Fifteen organs were removed and weighed and some 35 tissues were evaluated histologically from all surviving animals in the control group and at the highest dose, premature decedents being also examined in this way. The kidneys, liver, lungs, sublingual, submaxillary and parotid salivary glands and any abnormal tissue from the groups receiving the intermediate dose were also examined. Survival was not affected by treatment and there were no clinical signs of toxicity that were thought to be related to administration of glyphosate.  Ophthalmoscopy did not reveal any indications of adverse effects. Body-weight gain was reduced in males and females at the highest dose.

At the lower doses, no consistent and clearly dose-related body-weight change was to be seen.

Food consumption and water intake were not affected. Haematological changes were not considered to be treatment-related, although erythrocyte volume fraction and haemoglobin were occasionally increased in males and females at the highest dose.

However, a similar increase was also observed at other doses, in particular in males receiving a dose of 100 mg/kg bw per day, and a clear dose–response relationship was lacking.  In addition, the differences observed were rather small and no consistent trend became obvious throughout the study.

In contrast, clinical chemistry investigations and urine analysis elucidated some changes that could be attributed to administration of the compound.  An increase in plasma ALP activity became most apparent in males and females at the highest dose, but was also noted and occasionally reached statistical significance at the intermediate doses of 300 and 100 mg/kg bw per day (Table 21).

All other changes in clinical chemistry were not considered to be unequivocally treatment related.

Urine pH was consistently decreased in males at the highest dose and tended to be lower from a dose of 100mg/kg bw per day onwards. However, a similar effect was not observed in females.

At interim sacrifice after 52 weeks, absolute weight of the liver was reduced at doses of 1000, 300 and 100 mg/kg bw per day. For males, however, this finding was not confirmed by the sensitive means of covariance analysis, i.e. with correction for final body weight. At terminal sacrifice, no statistically significant decrease in liver weight was noted.

In contrast, mean weight of the kidney was reduced in groups of males at 100 and 1000 mg/kg bw after 104 weeks, but a clear dose–response relationship was lacking.

A probably treatment-related impact on weight of the salivary gland was noted in both sexes at interim kill. At study termination, weight of the submaxillary (mandibular)/sublingual glands in both sexes and of the parotid salivary gland (females only) still tended to be higher at the two higher doses.  However, when compared with these values in the controls, the difference was rather small, statistical significance was not achieved and there was no clear dose–response relationship.

Gross necropsy did not reveal indications of treatment-related non-neoplastic changes. The only remarkable histopathological finding attributed to administration of glyphosate was a dose-related increase in the number of animals exhibiting cellular alteration of the parotid and mandibular (submaxillary) salivary glands at the highest dose and at both intermediate doses (intermediate doses are 300 and 100 mg/kg bw/day). The changes were seen after 52 weeks. This alteration was described as the occurrence of hypertrophic and weakly (mandibular gland) or more deeply (parotid gland) basophilic acinar cells without any evidence of degeneration or other toxic damage. The severity of alteration was graded by the study pathologist on a scale ranging from “slight” to “very severe” (slight/very mild, mild, moderate, severe, very severe). The changes graded as “moderate” or “severe” were seen more frequently at 300 and 1000mg/kg bw per day (Table 22). The sublingual salivary gland was not affected. Neoplasia* was present in all groups, but there was no relationship with dose in the incidence of any individual tumour or in the total incidence of animals with tumours.

In conclusion, administration of glyphosate to rats for 104 weeks produced no evidence of a carcinogenic response. The liver and the salivary glands were identified as the main target organs of glyphosate-related toxicity in the long term. At 100 mg/kg bw per day, the changes in salivary glands were only minimal with respect to severity and were not considered to be of toxicological significance. Thus, the NOAEL was 100mg/kg bw per day on the basis of the more pronounced cellular alteration of salivary glands at 300mg/kg bw per day and greater (Atkinson et al., 1993b).

*Neoplasia is the abnormal growth or division of cells.

SAFE SAYS WHO NOTE:  2 other studies were used to back up the above findings “The Joint Meeting established a group ADI for glyphosate and AMPA of 0–1.0mg/kg bw on the basis of the NOAEL of 100mg/kgbw per day for salivary gland alterations in a long-term study of toxicity and carcinogenicity in rats and a safety factor of 100. The ADI is supported by NOAELs of 141 and 197mg/kgbw per day from the 1-year study and the two-generation study of reproductive toxicity in rats, respectively.”



[1]  Atkinson, C., Strutt, A.V., Henderson, W., Finch, J. & Hudson, P. (1993b) Glyphosate: 104 week combined chronic feeding/oncogenicity study in rats with 52 week interim kill (results after 104 weeks.). Unpublished report No. 7867, IRI project No. 438623, dated 7 April 1993, from Inveresk Research International, Tranent, Scotland. Submitted to WHO by Cheminova A/S, Lemvig, Denmark.

[2] Open Source:  Roundup & Birth Defects: Is the public being kept in the dark?

[3]  Myers, P., Hessler, W. 2007. Does the dose make the poison? Extensive results challenge a core assumption in toxicology.  OurStolenFuture.org, May 25.

[4]  Welshons, W. V., Thayer, K. A. et al. 2003. Large effects from small exposures. I. Mechanisms for endocrine-disrupting chemicals  with estrogenic activity. Environ Health Perspect 111(8): 994-1006.

[5]  Wetherill, Y. B., Petre, C. E. et al. 2002. The xenoestrogen  bisphenol A induces inappropriate androgen receptor activation and mitogenesis in prostatic adenocarcinoma cells. Mol Cancer Ther 1(7): 515-524.

[6]  Gierthy, J. F. 2002. Testing for endocrine disruption: how much is enough? Toxicol Sci 68(1): 1-3.


[8]  Soffritti M, Belpoggi F, Degli Esposti D. Cancer prevention: The lesson from the lab. In: Biasco G, Tanneberger S, eds. Cancer Medicine at the Dawn of the 21st Century: The view from Bologna. Bologna: Bononia University Press; 2006:49–64.

, , , No Comment

Comments are closed.