Typically, scientists who focus on reproductive and developmental safety look at two different sources of information: animal studies and epidemiologic investigations. In regard to animal data, glyphosate is relatively unique in having multiple independent companies perform reproductive and developmental toxicology studies in rodents and rabbits. These studies show no reproducible reproductive or developmental effects. Most recently, in 2012, a group of toxicologists conducted a detailed review of all of the animal and epidemiologic data and summarized: “An evaluation of this database found no consistent effects of glyphosate exposure on reproductive health or the developing offspring. Furthermore, no plausible mechanisms of action for such effects were elucidated.”
This analysis can be viewed online (Williams et al., 2012: http://www.tandfonline.com/doi/abs/10.1080/10937404.2012.632361).
So, what other studies are there, and do they in fact provide any convincing evidence of reproductive or developmental effect?
The most commonly cited study would be Paganelli et al. (Carrasco). These authors investigated the effects of a glyphosate-surfactant herbicide using two models: effects on frog embryos and effects following injection into the eggs of chickens. These models are not routine, and the predictive value for effects in mammals (including humans) is not clear. However, on the basis of findings in this study, the authors postulated an effect mediated by changes in retinoic acid (vitamin A) metabolism and speculated that these findings would apply to humans and, indeed, across the animal kingdom. It was a nice theory, but the problem is that there are lots of mammalian studies conducted by different groups, and the effects that Paganelli et al. predicted simply don’t happen in mammals.
The epidemiology literature (see Williams et al., 2012: http://www.tandfonline.com/doi/abs/10.1080/10937404.2012.632361) to date contains six studies looking at a variety of outcomes, including miscarriage, preterm delivery, spontaneous abortion, fetal death, neural tube defects and birth defects in general. Four studies showed no effect. One study (Bell, 2001) was a study of exposure to more pesticides than just glyphosate, and the same author could not replicate the study’s results in a larger study in the same state.
The other study alleging an effect (Garry et al.) demonstrated an overall birth defect rate far above that of earlier studies by the same author. The study asked participants to recall their exposure to chemistries without verifying their recollections, which is not a very reliable process. It resulted in an elevated risk of birth defects across all categories of chemistry studied. Of five studies looking at reproductive health (see Williams, 2012), four demonstrated no statistically significant adverse effects (one study showed statistically significant improved male fertility), and one study involved overall herbicide exposure, inclusive of glyphosate and other chemicals, precluding any ability to draw conclusions related to glyphosate itself. In short, there is no convincing or reproducible epidemiologic evidence of developmental effects related to glyphosate.
The final piece worth noting would be allegations out of Argentina that communities in the vicinity of spray applications of pesticides, including glyphosate and other materials, have experienced an increased rate ofbirth defects. This information has not been systematically collected, and the underlying population from which these individual cases have been collected is not defined. Hence, it is difficult to assess, because there is no measure of the true rates of birth defects. Alleged rates of birth defects actually fall below rates of birth defects seen in the general US population and populations in developed nations globally. This strongly suggests that any changes in rate have more to do with changes in data collection than with changes in actual rates. Finally, there is no way to disentangle exposure to glyphosate from exposure to other agents or, for that matter, from nutritional or other factors in the available data. The bottom line is that birth defect rates alleged in this population are simply not reliable, and conclusions cannot be made regarding relationship to glyphosate.