By Janet Raloff

Web edition: March 12, 2009

Bottled mineral water may deliver more than a no-calorie thirst quencher. If dispensed in bottles made from a type of plastic known as PET, for polyethylene terephthalate, this water may also pack a substantial quantity of estrogen-mimicking pollution, according to researchers at Johann Wolfgang Goethe University in Frankfurt.

Previous studies had demonstrated polycarbonate bottles — a hard, clear-plastic type — can leach bisphenol A, a chemical that turns on estrogen receptors in the body. But the bottles in the new study were made from a different plastic, the type used in most single-use water bottles.

Martin Wagner and Jörg Oehlmann washed out used water bottles made from either PET or glass and then cultured young mud snails in them. The species that they used (Potamopyrgus antipodarum) is particularly sensitive to estrogen, revving up its production of embryos when exposed to elevated quantities of the hormone. And that’s what was seen in those reared in the PET containers. Compared to snails reared in glass bottles, PET-housed ones produced more young — up to twice as many embryos.

Wagner notes that his team is currently probing the particular agents responsible for the estrogenic response associated with PET bottles, but the “high activity” he witnessed “would seem in indicate we have a mixture” of hormone-mimicking pollutants. The number of bonus embryos that developed in PET bottles was comparable to what developed when glass-housed snails were treated with 25 nanograms of ethinylestradiol — the potent synthetic estrogen in birth control pills — per liter of water.

Several years ago, while the researchers were investigating the estrogenic activity of various materials, Wagner decided on a whim to screen mineral water from the bottle he was drinking out of. And surprise! It tested very positive for estrogens.

That kick-started a series of sampling tests in their lab that eventually included 20 commercially packaged brands of water. All were popular, store-bought products. Nine came in PET bottles, nine more brands in glass, and two in aseptic Tetra Pak containers.

Water from each container was subjected to a test-tube assay. Anything that turns on estrogen receptors in this assay is labeled an estrogen — even though it may not be a natural hormone. The strength of any agents that tested positive were then compared to that of the most potent mammalian estrogen: 17-beta estradiol.

And in the new study, most glass-bottled waters show little or no estrogenicity. One PET-bottled brand of water also registered no estrogenicity,. The other PET-bottled waters all showed some hormone action. Five exhibited dramatic estrogenicity. So did water distributed in Tetra Pak containers, which may have a plastic interior lining.

The findings appear in a paper published online today in Environmental Science and Pollution Research.

Wagner notes that his team is currently probing the particular agents responsible for the estrogenic response associated with PET bottles, but the “high activity” he witnessed “would seem in indicate we have a mixture” of hormone-mimicking pollutants. The number of bonus embryos that developed in PET bottles was comparable to what developed when glass-housed snails were treated with 25 nanograms of ethinylestradiol — the potent synthetic estrogen in birth control pills — per liter of water.

One concern that Wagner mentions in his paper: Perhaps there are phthalates or other plasticizers in PET’s recipes that have mimic estrogens (female sex hormones) — or that block androgens (male sex hormones like testosterone). Indeed, some anti-androgens can appear estrogenic in how they play out in the body.

Endocrinologist Ana Soto of the Tufts University School of Medicine, in Boston, can think of another possible explanation. Plastics are very susceptible to degradation by sunlight. Manufacturers cope with the problem by incorporating antioxidants in a plastic’s recipe. And many of these antioxidants are phenolic compounds. In her lab and others, a number of phenols exhibit varying degrees of estrogenicity.

Of course, the big curiosity is why some glass-bottled waters might be estrogenic. More on that in the next blog.