When US government regulators hired the firm Sciences International to help review a potentially dangerous chemical, they almost certainly didn't intend to ignite a political firestorm.

But the company's links to the chemical industry have enraged environmentalists and forced the National Institute of Environmental Health Sciences (NIEHS) to pull the contract. Last week, in the face of growing pressure from both Congress and the public, the agency announced that it would re-review the 20 chemicals that the company had been involved in studying.

Scientists, environmental groups and even private contractors say that the situation highlights a weakness in federal environmental and health regulation. Private companies hired to do risk assessments are rarely required to disclose conflicts of interest. And often, the companies have relationships with the industries that the regulators are charged with overseeing. “It is definitely a problem,” says Sandra Schubert, director of government affairs for the Environmental Working Group, the watchdog based in Washington DC that criticized the work.

The review involves bisphenol A, a chemical commonly found in plastics, including baby bottles and microwave containers. The chemical disrupts the endocrine system in mice, and some researchers think that it might cause similar effects in humans. Both Japanese and European regulators say that more research is needed to determine whether the chemical is harmful.

A review on the safety of bisphenol A, found in baby bottles, is under scrutiny for conflicts of interest. Credit: CORBIS

In July 2005, the NIEHS hired Sciences International of Alexandria, Virginia, as part of a panel to summarize hundreds of studies on bisphenol A. The firm had done the same sort of work many times before for the agency. But this time, the Environmental Working Group caught wind of the study and learned that Sciences International had two clients — Dow Chemical and BASF — that produce bisphenol A. In late February, the advocacy group wrote to the agency demanding a review of Sciences International's role in the study. After several weeks of mounting pressure, the NIEHS announced that it was scrapping its contract and reviewing the work already done on bisphenol A and the other chemicals.

Several scientists involved in the review told Nature they think Sciences International had behaved ethically and impartially. “I really don't think that they were trying to influence the process,” says Simon Hayward, a cancer biologist at Vanderbilt University in Nashville, Tennessee.

The panel should be shut down.

But others contend that the company was deliberately stacking the review literature in favour of industry. “All those reports are suspect,” says Frederick vom Saal, an endocrinologist at the University of Missouri at Columbia, who was not involved in the review. “The panel should be shut down.”

Independent scientists on the bisphenol-A panel were required to sign conflict-of-interest forms, but Sciences International was not. The company did volunteer to report any potential conflicts of interest, according to Sciences International's president Herman Gibb. But he says that there was no conflict because the staff working on bisphenol A were not working for private clients. “We didn't do anything wrong,” he says. “This is unfair with a capital 'U'.”

Sciences International's website says that about half of its business comes from private companies and half from the government, and it is hardly alone in that regard. A brief investigation by Nature turned up at least half a dozen other firms that work simultaneously for federal regulators and industry — sometimes in the same area.

For instance, in 2005, Syracuse Research Corporation, a non-profit company based in New York, was hired by the US Centers for Disease Control and Prevention to help review perchlorates, which are reactive salts that can damage the thyroid. At the same time, the company's defence division was partnered with Lockheed Martin, a firm that has been sued for leaking perchlorates. Syracuse has an internal panel to review potential conflicts of interest, says spokeswoman Lisa Mondello. In this case, she says, the company collaborated with a division of Lockheed that doesn't handle perchlorates. “It's really not connected,” she says.

“The larger the company, the more it becomes an issue,” says James Lamb, a toxicologist with the Weinberg Group, a consultancy firm based in Washington DC that does regulatory work for private industry. Lamb says that some companies, like Sciences International, try to segregate industrial and government work to limit conflicts, but few if any regulators have rules about disclosing or managing such issues. “Conflicts of interest have never really been completely ignored,” he says. “But I really don't know what the legal process has been.”

In fact, the relationships are a legal grey area that has troubled regulators in the past. In 2004, the US Environmental Protection Agency and the NIEHS came under fire for accepting US$2 million from the American Chemistry Council — an industry lobby group — to conduct a children's health study (see Nature 432, 6; doi:10.1038/432006a 2004). The agencies eventually withdrew from the arrangement, and a subsequent investigation by the US Government Accountability Office found that stronger guidelines were needed. Both agencies say that they have since rewritten their guidelines about directly taking funds from industry.