The research, published in Thursday's New England Journal of Medicine, is the latest to show that the benefits of mammography are limited.
"It's not the great lifesaver that people think it is. It's not a magic bullet," said Georgetown University researcher Dr. Jeanne Mandelblatt who was not involved in the study.
Mandelblatt headed six teams that helped shape the new mammogram guidelines issued last year by an influential government task force. The U.S. Preventive Services Task Force concluded that women at average risk for breast cancer don't need mammograms in their 40s and should get one just every two years starting at 50.
The World Health Organization estimates that mammograms reduce the breast cancer death rate by 25 percent in women over 50. Other groups put the figure at 15 to 23 percent.
Among women in the screening group, the breast cancer death rate declined by 7.2 deaths per 100,000 people compared with women in the decade before the screening program started. The death rate in the non-screening group fell by 4.8 deaths per 100,000 people compared with its historical counterpart.
That means that mammography reduced mortality by only 2.4 deaths per 100,000 people – a third of the total risk of death.
A second part of the study bore this out: Women over 70, who weren't eligible for screening, had an 8 percent lower risk of dying from breast cancer compared to the previous decade, pointing to the benefit of better care.
The study was funded by the Cancer Registry of Norway and the Research Council of Norway. It was led by Dr. Mette Kalager of Oslo University Hospital with collaboration from Harvard University and the Dana-Farber Cancer Institute.