ROCKVILLE, MARYLAND--The first major epidemiological study of bisphenol A, a common ingredient in baby bottles and drink containers, suggests that relatively higher doses of the chemical double the risk of diabetes and cardiovascular disease. The finding comes as a scientific review panel meets here to begin evaluating the U.S. Food and Drug Administration's (FDA's) draft assessment of the health risks of the compound.
Many laboratory studies have examined the health impact of bisphenol-A, but the results have been inconsistent. One cause for concern is that the chemical weakly binds to the estrogen receptor. Some researchers have found in rodents evidence of reproductive problems, developmental harm, cancer, and neurological problems, whereas others have not (ScienceNOW, 16 April).
A team led by David Melzer, an epidemiologist and physician at the Peninsula Medical School, in Exeter, U.K. examined data from the U.S. Centers for Disease Control and Prevention's National Health and Nutrition Examination Survey, a representative sample of the general population. More than 90% of people had detectable levels of bisphenol A in their urine.
Melzer and his colleagues then looked at the bisphenol A levels and health status for 1455 adults, sampled in 2003 and 2004. Divided up by the amount of bisphenol A measured in urine, those in the top fourth had double the risk of type 2 diabetes and cardiovascular disease compared with those in the bottom forth. Levels of enzymes commonly used to diagnose liver trouble were also higher, they report in tomorrow's issue of the Journal of the American Medical Association (JAMA).
The diabetes link makes sense, Melzer says, pointing to recent animal and cell studies finding that bisphenol A interferes with insulin production in mice. In contrast, the heightened risk of cardiovascular disease was a surprise, he admits. Roughly gauging exposure from the urine levels, Melzer suspects that the adults in the survey with higher levels of bisphenol A were nevertheless consuming much less than the current daily level regarded as safe by the FDA for adults (50 micrograms per kilogram body weight).
The study was published early to coincide with today's hearing. An ad-hoc advisory panel of six scientists is reviewing the FDA's draft assessment of bisphenol A, released last month. In the assessment, the FDA evaluated the current literature. Focusing heavily on two large, industry-funded studies, the agency decided that the current safe level remains adequate. It rejected many other animal studies, primarily done by academics, as too small, inconsistent, or unreliable. (That contrasts with a final report released 3 September by the National Toxicology Program, run by the National Institute of Environmental Health Sciences, which found "minimal" to "some" concern when it evaluated these studies.)
The JAMA paper is not included in the draft assessment, and some panel members were skeptical of its findings. "I'm concerned that this paper will leave people with the message that bisphenol A causes heart disease and diabetes," says cardiovascular biologist Garret FitzGerald of the University of Pennsylvania School of Medicine in Philadelphia, who worries about the small sample size; only 79 people reported cardiovascular disease.
Environmental groups urged the panel to recommend that FDA consider the JAMA paper, as well as the complete range of animal studies, and revise its calculation of a safe level. That's not a simple task. "How in the world do you put all of this stuff together in a quantitative risk assessment?" panel member Philip Bushnell of the Environmental Protection Agency's neurotoxicology division wondered aloud. The panel will report back to FDA's science board by 31 October.