Science, 288 (16 June 2000):1941-1944

Just How Bad Is Dioxin?

Jocelyn Kaiser

The verdict is in--again: Dioxin is even worse for human health than previously believed. But, as has been true with earlier pronouncements on dioxin's risks, that judgment is controversial and may be appealed.

This latest assessment comes in an eagerly awaited draft report from the U.S. Environmental Protection Agency (EPA), which concludes that many Americans may have enough dioxin in their bodies to trigger such subtle harmful effects as developmental delays and hormonal changes in men. But the draft's most explosive finding is that the risk of getting cancer from dioxin is 10 times higher than previously estimated--a conclusion based largely on new data linking dioxin to cancer in workers.

That conclusion has flabbergasted many outside researchers, who first heard about it when the report was leaked to the press last month (Science, 26 May, p. 1313). A few told Science that they are concerned that EPA scientists may have fumbled again--when this was their chance to finally get it right. Indeed, agency scientists have spent the past 6 years revising the dioxin report, analyzing new data and reassessing earlier data after portions of their last draft were blasted by outside reviewers. "After all this time, if it doesn't fly, it will be an embarrassment to the agency," says environmental scientist Morton Lippmann of New York University, who chaired the earlier review panel and will lead the new one.

Dioxins are chlorinated chemicals produced mainly by incinerators and paper bleaching. They accumulate in the food chain, winding up in body fat when people eat animal products. In the 1980s, EPA concluded there was no safe level of dioxin--even the lowest exposure was hazardous. Then in the late 1980s, molecular biologists suggested that more than one dioxin molecule, perhaps considerably more, have to latch onto the cell receptor for dioxin to trigger toxic effects. Dioxin experts thought EPA may have overestimated the risk, so the agency set out to reassess it again in 1991.

Instead of downgrading the risk, agency scientists came back in 1994 with a draft report that supported EPA's earlier conclusion that there is no exposure threshold below which dioxin is harmless. But EPA's Science Advisory Board (SAB), while praising much of the reassessment, sent two key chapters back for revision, charging that agency scientists mixed science and policy and failed to mention alternate hypotheses and data that contradicted their conclusions (Science, 26 May 1995, p. 1124).

As requested, EPA has now rewritten the report's summary based on new dose-response modeling. It also added a new chapter to clarify how agency scientists reached their conclusions about the cumulative risks from dioxin-like chemicals by assigning each a "toxicity equivalency factor" and adding up their effects. The agency has "significantly updated" the report, says William Farland, chief of risk assessment in EPA's Office of Research and Development. "We have quite a bit of new information"--for example, from a study of Dutch infants exposed to polychlorinated biphenyls and dioxins--that even at background levels, dioxin may cause subtle neurobehavioral and immune effects.

As for cancer effects, the report upgrades dioxin from a "probable" to a "known" human carcinogen. For the most exposed people, such as those eating a diet high in animal fat, EPA puts the risk of developing cancer at between 1 in 1000 and 1 in 100. Farland says this controversial number comes from two changes in EPA's analysis. First, when scientists extrapolated results from rats to humans, they used a new metric that factors in dioxin's far longer half-life in human tissue than in rats. Second, EPA drew on new studies of three worker populations exposed to dioxin in the United States, Germany, and Holland. Those studies include information on the levels of dioxin to which workers were exposed, enabling experts to calculate how cancer risk rises with a given dose. That analysis, which "overlaps" with dose-response estimates from animal studies, results in a dioxin cancer potency that is 30 times higher than the 1985 estimate, Farland says. The agency factors in the threefold drop in dioxin exposure since the mid-1980s to conclude that the cancer risk today is 10 times higher.

Farland acknowledges that this number can be confusing to the public, explaining that this is the highest possible risk for the most exposed individuals, but for most people the risk will likely be lower or even zero. Even so, Farland says the report's new findings that dioxin in soil, water, and sediments may be a major source of exposure could warrant new measures to protect the food supply, for example, by cutting back on feeding lard and fish meal to cattle and pigs.

Whether such steps are reasonable will depend on whether the report passes muster with skeptical outside scientists. Several who spoke with Science asserted that the new worker studies of cancer effects are inconclusive. Even to those who have closely watched EPA's new analysis, the 10-fold increase "is a lot more than anybody expected," says Dennis Paustenbach, a risk assessment consultant with Exponent in Menlo Park, California. "It's going to require a lot of discussion before there's widespread acceptance."

That scrutiny will come in the form of public comments, a review by an outside science panel in late July, and another review by the SAB in September. Farland is urging scientists to take a close look at the report and the new data before passing judgment: "We'll have to see what they think after they've read the document."