This morning someone was explaining that all sorts of otherwise harmless substances were in fact harmful in super low doses. He mentioned parts per trillion, parts per billion, and parts per million. According to this fellow, standard tests for harmful effects are always conducted at high dose levels, and such testing will miss the terrible effects that occur when the dose is very very small. He cited one (just one) example, involving a breast cancer drug that I have never heard of.
Of course, this guy ignores an ancient principle, if a little bit is bad for you, a lot is worse. If you want to make sure something is harmless, feed a lot of it to a laboratory rat. If the rat survives, the stuff is harmless. This principle, and others like it in other fields, goes way back, and makes good sense.
What this guy is really saying is horrible things can happen with undetectably small exposures and we ought to go out and ban all sorts of things. He mentioned pthallates, a plasticizer that has attracted a lot of bad publicity but has passed a number of professional tests for toxicity.