I’ve written before about the effect of complex chemical pollutants on the environment, in particular the “feminization” of animals such as fish and amphibians. There is little doubt that significant changes are being seen in many animal populations, and there is no credible explanation other than chemical pollution. However, the exact mechanism is unclear, as is the level of danger to more complex animals such as humans.

Up until now the favorite explanation for this effect has been estrogen. We know that there is a lot more of it going into the water than there used to be, and it is a popular villain chemical in a number of ways. For male journalists, estrogen is a physical embodiment of girl cooties. It allows them to write about how our environment is being polluted by “girl stuff” that is “turning our sons into sissys”. Puritans of various stripes have put the blame on vain women. If people didn’t use hormone replacement therapy, the argument goes, we wouldn’t have this problem. And the Catholic church has leaped at the chance to identify contraceptive pills as a major pollutant.

What does the science say? As usual, it is complicated. But a new study by some UK scientists suggests another potential cause. Estrogen isn’t the only hormone pollutant around, and the new study looked at what are called “anti-androgens” – that is chemicals that block the effects of testosterone. These chemicals are used in the treatment of prostate cancer, and also in fertilizer. They have also been shown in lab work to be important in producing the sorts of feminizing effects seen in nature.

The new study is statistical. That is, it doesn’t demonstrate actual causality, it simply shows that there is a strong correlation between high levels of anti-androgen pollution and high numbers of feminized fish. Nor does it exonerate estrogen – there could well be some complex effect requiring more than one chemical going on here. But the correlation between estrogen pollution alone and feminized fish is a lot lower.

As ever, there is more work to be done. But if there is a lesson here it is that we shouldn’t jump to conclusions simply because they happen to fit our particular prejudices.

I can’t find the original paper online, but you can read the abstract here.