A reversal? No. So far, no harm has been shown in people, but some very smart people are arguing that we're not looking at the question correctly.

The animal literature is scattered with a decent number of studies and reports on various species (normally not rodents) that seem to have shown changes (I don't have it, but have seen a lengthy bibliography somewhere). Whther these animal studies then translate into harm for people--well, we just don't know is my point and where I'm being a bit of a contrarian. The critics point out that we're not looking for the effects of very, very low doses since we traditionally don't examine ultra low doses in human toxicology if higher doses show no effects.

Well, that's one of the problems trying to come to an answer that I didn't cover in my last post--the traditional toxicology paradigm works OK when thinking about the risk from lead or mercury, but maybe not for compounds that mimic hormones. Traditionally, if it doesn't cause problems at some large-ish dose, then the assumption is that it doesn't cause problems at lower doses. The critics point out that substances like BisA demonstrate a "non-monotonic dose-response curve" (almost a cliche phrase in this field now). Instead of a gradually falling risk with falling dose which eventually causes no problem below a threshold dose, you can have a U-shaped or inverted U-shape response instead, so assuming that there is no effect would be correct if the compound does behave this way. So, inject a guy with lot's of estrogen and you see changes. But go down to a fraction of that and you could see no/minute effects, but go down again to a tiny, tiny fraction of that dose and suddenly see a significant effect again. We don't normally do research at these very, very low doses. Currently, toxicology just extrapolates down to super low levels based on the effects at much higher doses.

That said, I think it's going to be really difficult to get much good human data at such mind-boggingly small concentrations. It's just tough to work with such tiny concentrations, particularly if you're trying to do some large-scale epidemiological study, which is probably the number of people you need to study to get enough statistical power to make some definitive judgement. It's like the Chernobyl example. Even a huge study population only turned up a small increase. So, I'm not sure we're ever get really good human data to answer the question. On the flip side of such tiny concentrations, the effects in people could quite subtle, making it extremely difficult to measure, but since these compounds are so ubiquitous, that's a large impact on public health and still would be important to know about, if in fact there's something there.