A new study suggests long-term use of dark hair dye may slightly raise a woman's risk of developing non-Hodgkin lymphoma (NHL). However, experts caution that the findings don't prove that coloring your hair causes cancer, and suggest that more studies are needed to determine if this small increase in risk is real.

"There have been enough epidemiologic studies of hair dye use to discount the possibility that there is any sort of large, wholesale risk of cancer related to hair dyes," said Eugenia Calle, PhD, director of analytic epidemiology for the American Cancer Society.

While some studies have suggested that hair dyes increase the risk of NHL, she noted, others have found the opposite. This study falls in the first camp, but "the results are by no means conclusive," Calle said.

Researchers from Yale University found that women who used hair dyes before 1980 were about 30% more likely to develop NHL than women who never colored their hair, or who used products developed after 1980. Women who used blond dyes did not have an increased risk, but those who used permanent dark colors (red, brown, or black) for more than 25 years had about twice the risk.

Nevertheless, a woman's overall risk of developing NHL is very low, so even if that risk is doubled, it's still unlikely she'll get the disease.

The results, which were published in the American Journal of Epidemiology (Vol. 159, No. 2: 148-154), were based on a study of 601 women with NHL and 717 women who did not have the disease. They were asked about the type of coloring product they used, how long and how often they used it, and how old they were when they stopped using it.

Lead researcher Tongzhang Zheng, BMed, ScD, ScM, associate professor of epidemiology and environmental health at Yale School of Medicine, said his study was the first to consider when a woman dyed her hair as an important factor. Many dyes made before 1980 contained chemicals that caused cancer in mice and rats, and the concentrations of these chemicals were higher in darker dyes.

"Hair coloring products have undergone tremendous change over the last 20 years," he pointed out. "Since 1980, many carcinogens have been removed from some formulas."

That may help explain the different findings for women who used light dyes versus dark, and those who colored their hair before 1980 and after. Or it could be that women who used newer formulations simply haven't been using the products long enough for any cancerous effect to be noticed. The women who used products before 1980 had to use them for at least 25 years before their cancer risk increased.

Zhang and Calle agreed that more studies are needed to determine whether newer dyes are associated with NHL.

Non-Hodgkin lymphoma is a cancer of the lymphatic system, which includes the lymph nodes, spleen, tonsils, and other organs. About 28,850 men and 25,520 women are expected to be diagnosed with the disease in the US in 2004.

Although the number of cases has been increasing, Calle said hair dye is probably not the reason.

"The disease is actually more common and has increased more rapidly in men, so hair dye is not likely to be a large contributor to the increasing rates," she said.