DANA POINT, Calif. (AP) - Women who get regular mammograms may reduce their risk of dying from breast cancer by more than 60 percent - about twice the amount seen in large formal studies, according to a new analysis by the American Cancer Society.

While there is little argument that mammograms are worthwhile, especially for women after menopause, the study suggests they may do substantially more good than most experts had assumed up to now.

``Women should be told that if they get regular screening, they will reduce their risk of dying from breast cancer by about two-thirds,'' said epidemiologist Robert A. Smith, the society's director of cancer screening.

He presented the data at a cancer society meeting in Dana Point. They will be published in the May 1 issue of the journal Cancer.

Seven landmark studies in North America and Europe have shown that mammograms save lives. Together, they suggest regular screening lowers the risk of breast cancer death by about 30 percent.

However, these studies, conducted mostly in the 1970s and 1980s before mammograms became routine, may have underestimated the true benefit.

The reason: Some women in the studies who were offered screening refuse to get mammograms, while others in the unscreened comparison groups sought out mammograms on their own. This diluted the difference between those who got mammograms and those who did not.

The latest analysis sought to see what happened to those who actually get mammograms. It was conducted in two counties in Sweden where the use of these screening tests is very high.

It found that those who comply with screening recommendations reduce their risk of dying from breast cancer by 63 percent when compared to the early 1970s, when mammograms were not routinely done.

Some have questioned whether improving breast cancer survival results from catching the disease earlier through screening or from better treatments.

However, Smith said the data suggest that screening accounts for nearly all the benefit. This is because no significant increase in survival is seen over time in either breast cancer patients who are too young for mammograms or among older women who refuse to get them.

``This is a controversial point, but they make it fairly cogently,'' said Dr. Marilyn Leitch of the University of Texas Southwestern Medical Center in Dallas. ``The contribution of treatment advances is there. But I think a large part of the improvement is related to the screening benefit.''

The study was based on 6,807 women aged 20 through 69 who were diagnosed with breast cancer over 29 years in the Swedish counties of Dalarna and Ostegotland. Of these, 1,863 died of the disease.

The researchers compared mortality rates in three time periods: 1968 to 1977, before mammograms were introduced; 1978 to 1987, when mammograms were offered to half of the women in the two counties in a large study; and 1988 to 1996, when mammograms were available every two years to all women over age 40.

During the most recent period, 85 percent of the women got regular mammograms. By comparison, Smith estimates that between 50 percent and 60 percent of U.S. women get mammograms that frequently.

Looking only at women who get screened opens the possibility of what researchers call selection bias. The worry is that those who get mammograms are already healthier than those who do not. It's their good habits, not their mammograms, that make them live longer.

However, Smith said this is unlikely to have made a big impact on the reduction seen in the study. When the researchers look at all women in the two counties, not just those who got mammograms, the risk of dying from breast cancer has fallen 50 percent since the 1970s.

Copyright 2001 The Associated Press. All rights reserved.