Filed at 8:27 a.m. ET

LONDON (AP) -- Amid a debate over whether mammograms save lives, major new research has reaffirmed that routine screening reduces breast cancer deaths, though not as much as previously believed, experts say.

Many experts say the latest findings, published this week in The Lancet medical journal, should settle the issue.

But Danish researchers who have questioned studies showing the value of mammographies say the new report does not remove their doubts.

In recent years doctors around the world have encouraged women to be screened for breast cancer. They have based their recommendations on seven landmark studies over the last 30 years that conclude mammograms can cut deaths from breast cancer significantly.

Last fall, the Danish scientists analyzed those studies and concluded that five of them were so flawed that it was not possible to tell if routine mammograms save lives.

Some of the most provocative of the seven landmark studies were done in Sweden, where researchers who have continued to track the effect of the procedure have now come together to update their findings.

This group, led by Lennarth Nystrom, an epidemiologist at Umea University in Sweden, found that mammograms reduce the risk of dying from breast cancer by 21 percent in the long term. A few years ago the consensus was 30 percent.

The benefit was greatest for women in their 60s, who had a 33 percent less chance of dying from breast cancer if they were offered regular mammograms compared with women who were not.

Women aged 55 to 64 reduced their risk by 27 percent. Women older than 64 and those younger than 55 also benefited, but the results were not so clear for those groups.

Dr. Peter Greenwald, cancer prevention chief at the U.S. National Cancer Institute, said the institute has independently reviewed all the landmark evidence and reached similar conclusions.

``The original studies did have flaws, but the flaws are not big enough to refute the evidence that mammography saves lives,'' Greenwald said.

His views echo those of other expert groups that recently have taken a fresh look at the decades-old studies.

However, Donald Berry, biostatistics chairman at M.D. Anderson Cancer Center in Houston, said the new research does not settle the question.

``Although the Swedes have admirably addressed some of the problems pointed out by the Danes, some issues of quality of the trials remain,'' he said.

Berry also said he believes the data in the new research reveal a 15 percent reduction in the risk of breast cancer deaths, not the 21 percent proposed by the Swedish scientists.

``Whether there is a benefit to screening is not clear. If there is a benefit then the benefit is small.''

Nystrom, who conducted one of the original studies, said the update was not prompted by the Danish criticism but does contain information on methods in an effort to address some of the concerns.

In a Lancet commentary, experts from the British Columbia Cancer Agency said the study ``reassures us that the Swedish data are believable and that they can be used to develop guidelines and assist individual women to make informed health care choices.''

However, Dr. Peter Gotzsche, director of the Nordic Cochrane Center in Copenhagen, who conducted the review studies questioning the lifesaving benefits of mammograms, said the new research has not convinced him.

``Some of our very important concerns (about the methods) have still not been addressed,'' he said.

``This is not an analysis of all the trials. One very good trial has been excluded and some trials we find problematic have been included,'' Gotzsche said. ``It is more necessary than ever that the investigators give all their raw data to an international body that can analyze all the evidence together.''

Breast cancer deaths in the United States and Europe have fallen by nearly 30 percent since 1990. Experts are not sure how much of this is due to catching the disease early with mammograms and how much to better treatment.

Mammograms can detect small tumors up to two years earlier than breast exams, providing more options for treatment.

However, they are not perfect. They miss some cancer. They also too often flag benign lumps, causing unnecessary anxiety, additional testing and biopsies.

Over 10 years of screening, about 23 percent of women will have an abnormal mammogram result and about 80 percent of those will be false alarms.

^------

On the Net:

The Lancet, http://www.thelancet.com