Review of the earth open source (eos) report " roundup and birth defects: is the public being kept in the dark?"



Download 0.79 Mb.
Page16/19
Date conversion21.02.2016
Size0.79 Mb.
1   ...   11   12   13   14   15   16   17   18   19

A4.5 Developmental and reproductive effects of a glyphosate-based herbicide in rats


Dallegrave et al (2003): Groups of 13 – 16 pregnant Wistar rats (90 days old, 200 – 280 g bw, bred at UFRGS, Porto Alegre, Brazil) received Roundup formulation (Lot BS 1096/98, Monsanto Brazil, containing 360 g/L glyphosate and 18% w/v POEA; no other components specified) by oral gavage at 500, 750 or 1000 mg glyphosate/kg bw/d17 (and ca 250, 375 or 500 mg POEA/kg bw/d) (dose volume of 10 mL/kg in distilled water) from GD 6 – 15. Control rats received vehicle alone. Caesarean sections were performed on GD 21, and foetal bodyweight and the numbers of corpora lutea, implantation sites, live and dead foetuses and resorptions were recorded. Foetuses were examined for external malformations and skeletal alterations. However, there was no investigation of their internal organs.

Maternotoxicity: At 1000 mg/kg, there was 50% maternal mortality between GD 7 and 14, but the study authors did not describe any clinical signs or identify the cause of death. No mortality occurred at 0 – 750 mg/kg. There was no treatment-related effect on maternal water intake. The 750 mg/kg group displayed a consistent deficit of ca 2.0% in food intake over GD 3 – 21; this is not considered to be treatment-related because it was already present before dosing had commenced. Dams in the 1000 mg/kg group showed a deficit of up to ca 4.0% in food intake during the dosing period, maximising on GD 9 but reversing after cessation of treatment. This was accompanied by slight mean bodyweight loss between GD 6 and 9. Subsequent weight gain was similar to the other groups, except for a transient increase over GD 15 – 16. However, there were no statistically significant inter-group differences in food consumption or relative or total gestational bodyweight gain (which was 107, 85, 107 and 102 g at 0, 500, 750 and 1000 mg/kg). Also failing to attain significance was a dose-related trend towards increased relative liver weights (4.57, 4.73, 4.89 and 5.11% in the respective groups). Absolute organ weight data were not presented.

Litter parameters: At Caesarean section, there were 15, 15, 16 and 7 dams and 154, 148, 162 and 75 foetuses available for examination at 0, 500, 750 and 1000 mg/kg. There were no effects on implantation index, resorption rate, mean number of foetuses per dam or mean foetal bodyweight. Gravid uterus weight was not measured. The only remarkable litter parameter was an increase in male:female sex ratio to 1.5:1 at 1000 mg/kg, compared with 1.06:1, 1.01:1 and 0.94:1 in the control, 500 and 750 mg/kg groups. Nevertheless, the finding was not statistically significant (p=0.724, X test) and there is no evidence that it arose from selective mortality of female foetuses in utero. Therefore, despite markedly reducing maternal survival at the high dose, the test formulation does not appear to have compromised foetal survival or growth.

Foetal development: There was no treatment-related effect on the incidence of external foetal malformations. However, as shown in the following table, an unequivocal treatment- and dose-related increase in skeletal alterations (all combined) occurred from 500 mg/kg upwards. These mainly involved ossification deficits suggestive of developmental delay but also included abnormalities such as absent ribs and caudal vertebrae, and wavy ribs. The most common individual alterations (incomplete skull ossification and enlarged fontanel) showed a dose-response relationship, but the incidences of some others were significantly (p<0.05) elevated at 750 and/or 500 mg/kg but not the high dose. It is not possible to exclude a relationship to treatment in these cases, because (a) no historical control or litter incidence data were presented, (b) the range of doses tested was very narrow, and (c) there were only half as many foetuses at 1000 mg/kg as in the remaining groups (which would reduce the chance of observing abnormalities).

Table 4.6: Percentage incidence of selected skeletal abnormalities in rat foetuses

Region or structure

Abnormality

Glyphosate Dose (mg/kg bw/d)

0

500

750

1000

Whole skeleton

All combined

15

33**

42**

57**

Skull, general

Incomplete ossification

Enlarged fontanel



10

1.9


29*

26*


39*

37*


56*

53*


Interparietal

Bipartite

0.6

19*

4.9*

0.0

Supraoccipital

Bipartite

Incomplete ossification



9.7

3.2


20*

0.0


1.2

1.2


0.0

13*


Maxilla

Short

0.6

0.7

0.0

1.3

Squama

Incomplete ossification

0.0

0.0

3.1*

2.7*

Caudal vertebrae

Absent

1.9

0.0

7.4*

15*

Ribs

Absent

Incomplete ossification

Wavy


1.3

1.9


0.6

2.7

2.0


2.0

3.1

5.6


4.9*

4.0

4.0


0.0

Sternebra

Incomplete ossification

Bipartite



1.9

3.9


14.9*

14.2*


0.0

0.6


2.7

9.3


Limbs

Incomplete ossification

0.0

0.0

17.9*

1.3

Scapula

Incomplete ossification

0.6

3.4

1.2

4.0

Metacarpal bones

Incomplete ossification

1.3

1.4

0.6

2.7

Femur

Incomplete ossification

3.2

3.4

13*

0.0

Tibia / fibula

Incomplete ossification

2.6

2.7

12*

8.0

Metatarsal bones

Unossified

4.5

1.4

14*

11

Hind phalanges

Unossified

7.1

21*

22*

2.7

Ischium

Incomplete ossification

4.5

2.7

9.3

0.0

Pubis

Incomplete ossification

3.9

2.7

11*

0.0

*p<0.05 **p<0.001 vs control (X22 test)

Conclusions

The NOEL for maternotoxicity was 750 mg glyphosate/kg bw/d, based on mortality and depression in food intake at the highest dose of 1000 mg/kg bw/d. There was no NOEL for effects on foetal development, due to increased incidences of skeletal abnormalities at and above the lowest dose of 500 mg glyphosate/kg bw/d.

Comment: Williams et al (2012) have criticised reporting deficiencies and anomalies in this paper, and also noted that foetuses were fixed in formalin and trypsin-digested prior to staining and skeletal examination instead of the standard method of alcohol fixation followed by maceration with potassium hydroxide. According to Williams, proteolysis could have digested peptide bonds in the bone matrix, creating areas that appeared to be incompletely ossified. Also deserving comment are the doses of POEA (ca 250, 375 and 500 mg/kg bw/d), which far exceed the maternal NOEL and LOEL of 15 and 100 mg/kg bw/d in rats (Holson, 1990). The mid and high doses are also greater than the foetal NOEL of 300 mg/kg bw/d18.

Dallegrave et al (2007): Groups of 15 Wistar rats (90 days old, 250 – 350 g bw, bred at UFRGS, Porto Alegre, Brazil) received Roundup formulation (Monsanto Brazil, containing 360 g/L glyphosate and 18% w/v POEA; no other components specified) by oral gavage at 50, 150 or 450 mg glyphosate/kg bw/d (dose volume of 10 mL/kg in distilled water) throughout pregnancy and lactation. Control rats received vehicle alone. At delivery, litter size, the number of living and dead pups, birth weight and sex ratio were recorded. Offspring development was monitored by weekly evaluation of bodyweight and daily assessment of developmental landmarks including ear and eye opening, fur emergence, incisor eruption, testis descent, preputial separation and vaginal opening.

From each litter, one rat/sex was killed at puberty (PND 65 for males; first oestrus after PND 65 for females) and a further animal/sex was killed at adulthood (PND 140). Systemic toxicity was determined on the basis of the relative weights of the heart, lungs, liver, spleen, kidneys, adrenals and brain. Reproductive toxicity in males was evaluated as relative weight of the testis, epididymis, seminal vesicle with coagulating gland and prostate, together with spermatid and sperm numbers in the cauda epididymis, sperm morphology, testicular histology and blood testosterone concentration. In females, assessment of reproductive toxicity was limited to the relative weights of the uterus, oviducts and ovaries without histological examination.



Maternotoxicity and litter parameters: There were no maternal deaths or effects on relative bodyweight gain of dams during pregnancy or lactation. There were also no effects on litter parameters at birth, the survival and growth of pups during lactation or attainment of general developmental landmarks.

Female sexual characteristics: Vaginal patency was delayed by two to three days in the treated groups, which was statistically significant (p<0.05, ANOVA-Bonferroni test) vs controls. Latencies of 34.9, 37.6, 36.9 and 36.7 days were recorded at 0, 50, 150 and 450 mg/kg respectively. Nevertheless, the study authors did not consider the finding to be biologically significant because the latency period was “well within” historical control values (these were not cited, however). There was no effect on the weights of the reproductive organs.

Male sexual characteristics: Although there was no effect on attainment of testicular descent, preputial separation was advanced by one day in the 450 mg/kg group (see following table). Despite achieving statistical significance, this was not considered treatment-related because the latency was within the historical control range (not cited). Testis and accessory sex organ weights were not affected by treatment.

However, the numbers and morphology of sperms in the treated groups showed noteworthy displacements from control values, which the study authors considered were biologically significant. As shown in the table below, these comprised:



  1. Statistically significant deficits of ca 25% in sperm numbers and daily sperm production at adulthood in the 50 and 450 mg/kg groups, although not at 150 mg/kg.

  2. A statistically significant doubling in the proportion of abnormal sperm at puberty in the 50 mg/kg, with a non-significant increase at 450 mg/kg but little or no effect at the mid dose. At adulthood, all treated groups displayed a ca 1.5-fold elevation in abnormal sperm incidence relative to controls, which did not achieve significance (p=0.066, ANOVA). Furthermore, in the treated groups the proportion of sperm-producing tubules was depressed by ca 6 – 11% at puberty and 18 – 29% at adulthood.

  3. Dose-related depression in serum testosterone levels, seen at all doses at puberty (significant at 450 mg/kg) but wholly or partially reversing by adulthood.

  4. Histological abnormalities within the testis. At puberty, there were growth disorders and degeneration characterised by spermatid vacuolisation and a decrease in elongated spermatids at and above 150 mg/kg. At adulthood there was dose-related, intense tubular degeneration characterised by the absence of tubular lumen (see table).

Based on the above findings, the study authors considered that there was no NOEL for effects on the male reproductive system, and suggested that the test formulation was a probable endocrine disruptor. However, they acknowledged that the study had not elucidated a mechanism of action or identified which component of Roundup was causing the observed effects.

Table 4.7: Reproductive parameters (mean values) in male offspring

Parameter

Maternal glyphosate dose (mg/kg bw/d)

0

50

150

450

Age at preputial separation (d)

31.7

31.7

31.5

30.7*

Bodyweight at preputial separation (g)

73.0

68.1

72.2

70.7

Daily sperm production (x 106) (n=15) PND 140

20.5

15.3*

19.7

14.7*

Sperm number (x 106) (n=15) PND 140

345

251*

369

257*

Abnormal sperm (%) (n=15) PND 65

PND 140

8.6

5.4


16.7*

8.3


9.2

8.4


11.6

7.7


Tubules with spermatogenesis (%) PND 65

(n=5) PND 140

84

92


77

74


79

75


75

65


Blood testosterone concentration (ng/mL) PND 65

(n=15) PND 140

5.2

3.9


4.0

3.4


3.2

6.3


1.5*

3.3


Testis: spermatid vacuolisation & decrease in elongated spermatids (incidence at PND 65)

NS

NS

4/5

4/5

Testis: tubular degeneration (incidence at PND 140)

NS

3/5

4/5

4/5

*p<0.05 vs control, ANOVA – Bonferroni test

NS = Not stated



Comment

Interpretation of the results is hindered by the lack of historical control data, which may have defined effect levels and clarified whether there were genuine treatment-related effects on variables that did not show dose-response relationships. These include daily sperm production, sperm numbers in the cauda epididymis and the proportion of abnormal sperms, which showed the least displacement at 150 mg/kg. The reviewing toxicologist considers that the reporting of histological findings in the testis was insufficiently detailed, as it lacked descriptive detail, severity gradings and control data. The study would also have been strengthened by histological examination of the female reproductive organs.

In an independent assessment of this study, Williams et al (2012) have remarked that:


  • In the 450 mg/kg bw/d group, the age at preputial separation was within the physiological range for rats;

  • Hastening of puberty would be not be expected, given that the 450 mg/kg group had the lowest mean circulating testosterone level on PND 65;

  • The increased percentage of abnormal sperm at 50 mg/kg bw/d may be a random finding, given the lack of effects at higher doses;

  • Dallegrave et al’s reporting of the testicular histology was deficient and the abnormalities described may be a tissue processing artefact, rather than an effect of treatment;

  • Testicular abnormalities have not been reported in offspring in reproduction studies with glyphosate, all of which involved much greater glyphosate exposures.

Conclusions

In the absence of any apparent maternotoxicity, the NOEL in dams was 450 mg glyphosate/kg bw/d. The study did not demonstrate treatment-related effects in female offspring at up to and including the highest dose of 450 mg glyphosate/kg bw/d. The study is considered to be insufficiently reliable enough to demonstrate whether there were treatment-related effects in male offspring.



Romano et al (2010): The test compound in this study was Roundup Transorb (Monsanto Co, St Louis, MO, USA / Monsanto of Brazil Ltda, Sao Paulo, Brazil; containing glyphosate isopropylamine salt 648 g/L equivalent to 480 g/L glyphosate, with 594 g/L of unidentified “inert ingredients”). The formulation was diluted in water to yield a dosage volume of 0.25 mL/100 g bw, and administered PO by gavage to newly weaned male Wistar rats (16 – 18/group) from PND 23 – 53 at 5.0, 50 or 250 mg/kg bw/d. A control group received vehicle alone. The study authors described their test compound as “glyphosate-Roundup Transorb”, so it is ambiguous whether they were referring to the active or product. However, given that their choice of doses was based on a NOEL of 50 mg/kg bw/d for glyphosate in another study, it will be assumed that the doses are equivalent to 5.0, 50 or 250 mg active/kg bw/d.

Pups were weighed daily throughout the treatment period and examined to determine the age of puberty (balano-preputial separation) from PND 33 onwards. At termination on PND 53, serum was collected via cardiac puncture for measurement of testosterone, oestradiol and corticosterone concentrations. The testes and adrenal glands were weighed and processed for histological examination. Quantitative morphometry of the seminiferous tubules was then performed to examine for disturbance of spermatogenesis. However, spermatozoa were not examined or quantified.

There were no treatment-related effects on bodyweight throughout the dosing period, including puberty (p>0.05). However, attainment of puberty was delayed by ca 1.0 and 1.5 days at 50 and 250 mg/kg respectively (p<0.01 and <0.001 vs control). As shown in the following table, relative testicular weight increased dose-relatedly by up to ca 9%, attaining statistical significance at 250 mg/kg. At this same dose, there was also a significant, 29% increase in relative adrenal weight. Absolute organ and terminal body weights were not provided.

Serum testosterone concentrations were depressed by 30%, 45% and 50% at 5, 50 and 250 mg/kg bw/d. Histologically, this finding was correlated with decreased numbers of germ cells, seen as a dose-related reduction in the height of the seminiferous tubule germinal epithelium and increased diameter of the lumen. Displacements from control were statistically significant at all doses (see table below). By contrast, serum corticosterone and oestradiol concentrations, adrenal morphology and the overall diameter of the seminiferous tubules were not affected.



Table 4.8: Treatment-related effects in rats

Variable examined

Dose (mg/kg bw/d)

0

5

50

250

Mean testicular weight (mg/100 g bw)

531

539

553

580*

Mean adrenal weight (mg/100 g/bw)

11.3

12.8

12.3

14.6*

Serum testosterone concentration (ng/dL)

155

109**

85***

77***

Seminiferous tubule: Germinal epithelium height (µM)

Lumen diameter (µM)

86

94


72**

117**


69**

114**


65**

130**


*p<0.05 **p<0.001 vs control

Comment

The study was performed before the publication of the EPA OPPTS Test Guideline 890.1500 for investigating pubertal development in male rats19, but the treatment period (PND 23 – 53) was in line with the Guideline-specified protocol. However, the study was not Guideline-compliant in numerous other aspects of its design and reporting. In particular, there were no bodyweight data except for the mean values at preputial separation. It is therefore impossible to verify independently that inter-group variation in bodyweight and/or bodyweight gain did not influence the timing of puberty, or other parameters. It is also unclear whether the experimenters ensured that litter mates were not allocated to the same experimental group, as required by the Guideline.

Furthermore, Williams et al (2012) have questioned the reliability of the preputial separation data and morphometric analysis of testis pathology, claiming that the latter was affected by tissue fixation artefacts and confounded by variation in the maturity of seminiferous tubules.

Conclusions

The study is considered to be insufficiently reliable to demonstrate whether there were treatment-related effects in the experimental model used.



Romano et al (2012): Roundup Transorb (see Romano et al, 2010) was administered to pregnant Wistar rats PO by gavage at a dose equivalent to 50 mg glyphosate/kg bw/d from GD 18 to PND 5. The test compound was diluted in water and given at a dose volume of 2.5 mL/kg bw. A control group (size unspecified) received water alone. On PND 4, litters were culled to eight pups/dam and then maintained until weaning at PND 21. Their bodyweight was recorded on PND 21, 30, 40 and 60. Throughout the post-weaning period, male offspring were evaluated for preputial separation, indicating attainment of puberty.

Preference test: On PND 60, subgroups of five male offspring from treated and control dams alternately underwent a sexual preference test, in which they were placed individually on a circular stage with one male and one female stimulus rat, housed in separate cages on opposite sides of the apparatus. The stage was divided into neutral, male and female areas, with the male and female areas divided into seven zones. Stimulus males were gonad-intact and sexually mature, whereas the stimulus females had been ovariectomised and brought into oestrus with oestradiol (50 µg/kg SC at -54 hours) and progesterone (2.0 mg/kg SC at -6 hours). After a five-minute adaptation interval, there was a 20-minute observation period during which the test males’ stay times in the two zones nearest the stimulus males and females were recorded. Preference scores were calculated by subtracting the total time spent in the male zones from the time spent in the female zones. Following the preference trial, the test males were not subjected to other experiments.

Mating behaviour: Four males from treated and control dams were scored for the numbers of mounts, attempted mounts, intromissions and ejaculations over a 40-minute interval when placed individually with an oestrus-induced female rat. The time to first ejaculation and ejaculatory intervals were also recorded.

Reproductive tract: On PND 60, the testes, epididymides (caput, corpus and cauda) and seminal vesicles were weighed, sperm counts were performed, and the histology and morphometry of the seminiferous epithelium examined by light microscopy.

Other parameters: Serum concentrations of testosterone and oestradiol were measured by RIA, and FSH and LH concentrations were measured using chemiluminescence immunoassay. Pituitary mRNA and protein levels of β-LH, β-FSH and GH were analysed by real-time PCR (for mRNA) and SDS-PAGE followed by nitrocellulose membrane hybridisation / antibody detection (for proteins).

Results

Maternal observations: No information was provided on the survival, appearance, behaviour or bodyweight of dams during or after the dosing period. It is therefore unknown whether any maternotoxicity occurred.

Growth of offspring and attainment of puberty: The study authors did not present data on bodyweight or pituitary GH levels, but claimed that neither was affected by treatment. In males from Roundup-treated dams, however, age and bodyweight at preputial separation were decreased by about two days (mean of 45 vs 47 days; p<0.05) and 30 g (mean of 215 vs 245 g; p<0.05).

Preference test: As shown in the table below, male rats from the Roundup-treated dams spent significantly longer in close proximity to female stimulus animals, and had a significantly higher preference score.

Table 4.9: Results of sexual preference test

Parameter

Time (sec)

Control

Roundup

Mean total time in male area

431

312

Mean total time in female area

502

625**

Mean partner preference score

71

313**

**p<0.01 vs control (Student’s t-test) N = 5/group

Mating behaviour: Based on the interquartile ranges, the study authors claimed a significant increase in mounting, intromission and ejaculatory latency for males from Roundup-treated dams. The remaining parameters did not differ significantly between the groups.

Table 4.10: Results of mating behaviour evaluation

Parameter

Time (min)

Control

Roundup

Latency for the first mount^

0.6 – 1.0

5.2 – 7.0*

Latency for the first intromission^

0.6 – 1.0

5.2 – 7.0*

Latency for the first ejaculation^

1.0 – 1.7

5.5 – 7.0*

^Data are interquartile range (25 – 75%) N = 4/group

*p<0.05 vs control (Mann-Whitney U-test)



Reproductive tract: There were no effects on the relative weights of the testes or undrained seminal vesicles on PND 60. However, the relative weight of drained seminal vesicles was 10% higher in the Roundup group, suggesting a lower fluid volume. The corpus and cauda segments of the epididymis were slightly but significantly heavier in the Roundup group than controls. Compared with controls, sperm production was approximately twice as high in rats from Roundup-treated dams (see table below), and sperm reserves in the caput + corpus were increased by 50%. Sperm transit time through the cauda was reduced by ca 1/3. In the absence of any significant difference in the diameter of the seminiferous tubules, the Roundup group displayed a minor but statistically significant increase in epithelial height and decrease in luminal diameter.

Table 4.11: Findings in the reproductive system of male rats

Parameter

Control

Roundup

Total sperm production (X 106/testis)

52

99*

Total sperm production (X 106/g testis)

35

71*

Daily sperm production (X 106/testis)

8.5

16*

Daily sperm production (X 106/g testis)

5.7

12*

Sperm reserve, caput + corpus (X 106)

14

21*

Sperm transit time through cauda (days)

6.3

4.0*

Seminiferous

epithelium



Tubular diameter (µm)

467

451

Epithelial height (µm)

92

98*

Luminal diameter (µm)

257

239*

Seminal

vesicle


Weight, undrained (mg/100 g bw)

160

155

Weight, drained (mg/100 g bw)

100

110*

Epididymis

Weight, corpus (mg/100 g bw)

10

13*

Weight, cauda (mg/100 g bw)

36

43*

*p<0.05 vs control (Student’s t-test) N = 8/group

Other parameters: In males from Roundup-treated dams, serum testosterone and oestradiol concentrations were approximately twice as high as in controls (see following table). Pituitary LH and FSH mRNA levels were very slightly but significantly increased by Roundup treatment. However, although there were concomitant increases of ca 70% in pituitary LH protein and serum LH levels, there was no treatment-related effect on FSH levels in the pituitary or serum.

Table 4.12: Hormonal levels in the serum and pituitary

Parameter

Control

Roundup

Serum testosterone conc. (ng/dL) (N = 12)

60

140**

Serum oestradiol conc. (pg/mL) (N = 12)

1.4

2.8**

Pituitary LH mRNA content (AU) (N = 8)

1.00

1.02*

Pituitary LH protein content (AU) (N = 8)

1.1

1.9**

Serum LH conc. (pg/mL) (N = 8)

70

120*

Pituitary FSH mRNA content (AU) (N = 8)

1.00

1.02*

*p<0.05 **p<0.01 vs control (Student’s t-test)

Conclusions

The study authors interpreted their findings as indicating that maternal glyphosate exposure during the perinatal period caused hypersecretion of androgens in the male offspring, combined with hastening of puberty, increased gonadal activity and sperm production, greater predilection for the company of female rats and increased libido (the latter notwithstanding the statistically significant increase in the delay before copulation). The authors acknowledged that their findings contradicted the depression in serum testosterone level and sperm production and reduced height of the seminiferous epithelium observed by Romano et al (2010) and Dallegrave et al (2007) (see above). However, they attributed the discrepancies in experimental outcome to differences in timing of exposure, which occurred over GD18 to PND 5 in this study but extended through gestation to the end of lactation (PND 21) in Dallegrave et al (2007) and was from PND 23 to 53 in Romano et al (2010).



Comment

Numerous aspects of the design of this study and its findings deserve comment.



  • Although the study authors attribute their findings to glyphosate, dams were treated with a commercial formulation containing 594 g/L of unidentified “inert ingredients”. Offspring may consequently have been exposed to these formulation adjuvants in utero or via maternal milk and it is possible that they influenced the experimental outcome, either directly or by interaction with the active constituent. The study did not control for the presence of adjuvants.

  • Since no observations on the dams were presented, it is unknown whether maternotoxicity (including effects on maternal nursing behaviour) occurred. The study authors appear not to have considered the possibility that at least some experimental findings in offspring arose from effects on the mothers.

  • The study authors did not state when serum and pituitary hormone parameters were measured.

  • Rats that underwent the sexual preference test were not used for other experiments, but no information was provided on whether those undergoing evaluation of mating behaviour were also subjected to hormone assays and/or reproductive tract histology. Either of these end-points could have been affected by sexual activity.

  • In a mating evaluation, one would expect relatively large variation in the behaviour of individual males, especially given that the outcome would be partially dependent on the behaviour of the partnering females. However, the group sizes were very small (N = 4). No group mean values were provided; data were reported as interquartile ranges (25 – 75%). In a set of four observations, there would be only one data point per quartile. Therefore, because they were based on so few observations, it is open to question whether the apparent increases in mounting, intromission and ejaculation latency time were biologically significant, even though statistical significance was attained.

  • In an extensive critique of this study, DeSesso and Williams (2012) point out that surfactants inhibit the enzyme aromatase, which is responsible for conversion of circulating testosterone to oestradiol. Surfactants, if present in the test formulation, could therefore have disrupted the expression and function of endocrine hormones in the dams and/or offspring.

  • The study authors did not identify from which dams/litters the evaluated males had originated. DeSesso and Williams question whether the study was controlled for litter effects, adding that because litter mates are more similar to each other than offspring from separate litters, the observed inter-group differences may be due to animals being derived from the same limited number of litters rather than a true effect of treatment.

  • DeSesso and Williams note the lack of evidence that precautions were taken to prevent the sexual preference test being confounded by environmental cues including auditory and visual stimuli, odours and pheromones.

  • These authors also observe major differences in the control values for attainment of puberty, serum testosterone and oestradiol concentrations and seminiferous tubule morphometry when comparing Romano’s 2010 and 2012 studies. The magnitude of these differences exceeds the size of the treatment-related changes within each study.

  • Romano et al (2010; see above) report that preputial separation in controls occurred at means of ca 37 days and 146 g bw, compared with 47 days and 245 g bw in their 2012 paper. Mean values from test animals in 2012 (45 days and 215 g bw) lie within this range, and also within the range specified for control Wistar rats in US EPA TG 890.1500 (40 – 46 days and 177 – 241 g bw)rd. By contrast, mean values from controls in both studies lie outside the EPA’s Guideline ranges (DeSesso and Williams, 2012).
1   ...   11   12   13   14   15   16   17   18   19


The database is protected by copyright ©essaydocs.org 2016
send message

    Main page