Millions of research dollars have been spent worldwide to determine whether cellphones cause brain tumors. Now, what health experts call a large-scale, well-conducted study has yielded the most conclusive evidence of such a link to date.

Researchers at the Karolinska Institute of Environmental Medicine, in Stockholm, Sweden, have found an association between long-term cellphone use and a rare, benign tumor, causing concern among radiation specialists and epidemiologists, though they emphasize that the results haven't been replicated yet. Scientists now eagerly await results from other studies under way around the world.

Published in the November issue of the journal Epidemiology, the Swedish study, led by Stefan Lönn of the Karolinska Institute, looked at 148 people who had acoustic neuroma and compared them with 604 healthy people. It found that people who used cellphones for more than 10 years doubled their risk of developing the tumor, a benign condition affecting one in 100 000 people. Acoustic neuroma grows on the nerve connecting the brain and the inner ear, causing hearing loss. The risk was four times as high on the side of the head where the phone was usually held.

No tumors were associated with less than 10 years of cellphone use, an outcome that has been documented by past studies. So, as more and more people approach this 10-year mark, should they be worried [see chart, "Looming Threat?"] And should they also be concerned about other, malignant tumors? Experts aren't sure.

"The degree of care that went into this study makes it worthy of attention," says James C. Lin, who studies the biological effects of electromagnetic radiation at the University of Illinois at Chicago. But, he warns, "biological responses are so complex that you can't base judgment on one observation."

Broadly speaking, most animal and epidemiological studies have found no connection between tumors and cellphone use, notes Kenneth R. Foster, who studies the health risks of nonionizing electromagnetic radiation at the University of Pennsylvania, in Philadelphia. Because acoustic neuroma is unusual, Foster is not much fazed by Lönn's findings, personally. "The risk of getting such a rare, benign tumor isn't going to keep me up at night," he says. Nor does he believe that the results will have much immediate impact on U.S. government health authorities, or on court cases brought against cellphone makers by victims of brain cancer, which is a different disease.

But the situation warrants watching, say experts. The Lönn paper is second in a series of studies looking at the relationship between cellphone use and the risk of various brain, head, and neck tumors. Thirteen countries, the United States not among them, are a part of this collaborative effort coordinated by the International Agency for Research on Cancer, in Lyon, France. The first research paper, from Denmark, showed no link between acoustic neuroma and cellphone use, but it contained fewer subjects with long-term exposure. The IARC expects to analyze the data from the separate countries' reports on this issue and to have collective results early next year. These reports should cover 1000 cases of acoustic neuroma, as well as many cases of other types of tumors, both benign and malignant.

Linda Erdreich, a senior managing scientist specializing in health risk assessment and epidemiology at the New York City office of Exponent, a science and engineering consulting firm, believes that the Swedish scientists did "a good job" but has some lingering doubts about their findings. One is the potential for what's called detection bias. People might be more likely to be diagnosed with tumors on the side of the head where they usually hold the phone because they may notice hearing loss more easily on that side and seek treatment. Also, Lönn and his colleagues found no correlation between tumor occurrence and number of hours of phone use. If cellphones were the culprit, longer phone use would cause more tumors.

If cellphone users want to be cautious, Foster advises, they can reduce the hours they spend calling on the go or use headsets. Forget about RF-shielding and radiation-reduction devices, says Foster. According to him, they don't work.