The <a href="http://www.discover.com/dec_02/featradiation.html">theory of hormesis</a> has been kicking around since the 16th century in cycles of repute and disrepute, and now seems to be swinging back toward repute. Hormesis, from the Greek word for 'excite,' is the theory that small doses of toxic substances stimulate organisms toward health, where larger doses of the same substances cause death and disease. This sounds a bit like homeopathy, and hormesis theory has suffered by association, but hormesis is to homeopathy as astronomy is to astrology. Respected Amherst toxicologist Edward Calabrese has spent the last few decades helping to make this distinction. Calabrese won his reputation proving that even a single dose of a carcinogen can cause disease, making him a bit of a scourge to corporations and a friend to environmentalists. Since then he's attracted a different set of fans and foes, having collected 4,500 cases that demonstrate hormesis, and founding a 12,000 member scientific society to study these effects. Far from being a cold-fusion like theory from isolated researchers, 99.9% of these cases were collected by other researchers, and they are reproducible. Examples of the effect include studies showing that mice thrive on low doses of dioxin, low levels of carbon monoxide prevent heart disease, low doses of radiation boost four-year metastatic cancer survival rates by 50%, and low levels of mercury lessen children's chances of developing neurological tics, delayed speech and other pathologies. The scope of these effects has convinced Calabrese and other researchers that hormesis is a fundamental and unifying aspect of biology. "The mechanisms of the hormesis effect are unclear, but there are many theories. For instance, low doses of radiation may stimulate DNA repair, and stimulate the immune system by creating a constant search and destroy mission for cancerous cells. While all radiation causes damage, low levels may stimulate more repair than damage. In general, hormesis seems to occur when organisms overcompensate for an irritant, reaching a new and healthier equilibrium." While hormesis typically occurs at concentrations around 5x lower than the toxic threshold, the EPA often sets acceptable exposures to 20x lower than the threshold. The distinction between a linear curve and a U curve is more than academic. Arsenic studies suggest that raising arsenic levels from 5 to 50 parts per million could prevent 1,000 cancer deaths per day nationwide. The larger consequence of hormesis may be a significant paradigm shift. We know that too much of a good thing is bad; now it turns out that too little of a bad thing is bad. I guess anti-germ soaps are one of the better and most recent examples.