By Gabriel Paijo 
Breast cancer screening has long been a cornerstone of public health, credited with saving countless lives through early detection. However, for decades, a shadow has loomed over the practice in the form of "overdiagnosis"—the identification of tumors that are so slow-growing or indolent that they would never have caused symptoms or threatened a woman’s life. While the medical community has acknowledged overdiagnosis as a potential drawback, the debate over its frequency has been fierce. Some previous estimates, derived from various randomized controlled trials, suggested that as many as 30% to 50% of screen-detected cancers could be overdiagnosed. Now, a comprehensive new study published by an international team of researchers suggests those high estimates were likely the result of premature data analysis and a failure to account for complex timing effects. By re-evaluating the evidence from all major randomized controlled trials and comparing it with long-term data from Denmark, the researchers have concluded that the true rate of overdiagnosis is likely below 5%.
The findings, led by experts from the University of Southern Denmark, Lillebælt Hospital, the University of Copenhagen, and Queen Mary University of London, offer a significant shift in the narrative surrounding mammography. For years, the high-end estimates of overdiagnosis influenced international screening guidelines and colored the information provided to women, sometimes leading to skepticism about the value of regular mammograms. This new analysis aims to correct the record, providing a more stable framework for understanding the balance between the benefits of early detection and the risks of unnecessary treatment.
The Evolution of the Overdiagnosis Debate
The concept of overdiagnosis is rooted in the biological diversity of breast cancer. Not all tumors behave the same way; while some are aggressive and rapidly lethal, others may remain localized or grow so slowly that a woman would die of other causes before the cancer ever became apparent. Screening, by design, picks up both types. When a woman is diagnosed with a cancer that would never have harmed her, she may undergo surgery, radiation, or chemotherapy—treatments that carry physical and psychological side effects—without receiving any survival benefit.
This phenomenon became a central point of contention in the early 2000s. Critics of widespread screening pointed to data from randomized trials, such as the Canadian National Breast Screening Study, which suggested a high degree of overdiagnosis. These figures were used to argue that the harms of screening might outweigh the benefits for certain age groups, leading to shifts in recommendations from organizations like the U.S. Preventive Services Task Force (USPSTF) and the Nordic Cochrane Centre.
"The aim of our study was to bring together the evidence from all randomized controlled trials to get a clearer picture of the extent of overdiagnosis in breast cancer screening," says Sisse Helle Njor, a professor at the University of Southern Denmark and Lillebælt Hospital. According to Njor, the previous interpretation of trial data as evidence of a "substantial problem" was not as straightforward as it appeared.
Methodology: Reassessing the Gold Standard
To reach their conclusions, the research team performed a deep dive into the eight primary randomized controlled trials (RCTs) that form the bedrock of mammography science: the New York Health Insurance Plan trial, the Malmö trial, the Two-County trial, the Edinburgh trial, the Canadian National Breast Screening Study, and the Stockholm, Gothenburg, and UK Age trials.
RCTs are considered the gold standard of medical evidence because they compare a group invited for screening with a control group that is not. However, the researchers identified three critical factors that can artificially inflate overdiagnosis estimates if not properly handled:
- Lead Time and Compensatory Drop: When screening is first introduced, there is a natural "spike" in diagnoses because cancers are being found years earlier than they would have been otherwise. This is known as "lead time." For the statistics to balance out, researchers must wait for a "compensatory drop" in the control group later on. If a study is analyzed too early, the initial spike is mistaken for permanent overdiagnosis.
- Screening of the Control Group: In many trials, women in the control group eventually began receiving mammograms—either because the trial ended or because screening became standard practice in their region. This "contamination" of the control group makes it difficult to compare the two populations fairly.
- Follow-up Duration: Without a sufficiently long follow-up period, the "catch-up" period where the control group’s diagnosis rate should theoretically meet the screening group’s rate is never captured.
The team compared the patterns found in these trials with real-world data from Denmark. Denmark provided a unique "natural experiment" because organized screening was rolled out in different regions at different times. Some areas began screening 17 years before others, allowing researchers to track the long-term incidence of breast cancer in screened versus unscreened populations with high precision.
The Findings: A Dramatic Reduction in Estimates
The researchers discovered that when the data from the original RCTs were adjusted for the timing of screening and the length of follow-up, the results aligned closely with the Danish experience. In Denmark, where the data has had decades to mature, overdiagnosis is estimated to be less than 5%.
"When interpreted in their full temporal context, randomized trial data are consistent with overdiagnosis of less than five percent, rather than with estimates nearing 50%," explains Matejka Rebolj, Senior Epidemiologist at Queen Mary University of London. Rebolj notes that the earlier, higher estimates were often based on evidence collected before the trial data had "fully matured."
This reassessment included both invasive breast cancer and ductal carcinoma in situ (DCIS), a non-invasive condition that is often a focal point of the overdiagnosis debate because its progression to invasive cancer is not guaranteed. Even including DCIS, the adjusted figures remained remarkably low compared to previous literature.
Chronology of Breast Cancer Screening Research
To understand why this new study is so pivotal, one must look at the timeline of mammography research:
- 1963: The Health Insurance Plan (HIP) of New York begins the first large-scale RCT of mammography, eventually showing a reduction in mortality.
- 1970s-1980s: Several trials in Sweden (Malmö, Two-County, Stockholm, Gothenburg) further establish the efficacy of screening.
- 1980s-1990s: National screening programs are implemented across Europe and North America.
- 2000: The Cochrane Collaboration publishes a controversial review questioning the mortality benefits of screening and highlighting the risk of overdiagnosis.
- 2012: An influential study in the New England Journal of Medicine suggests that overdiagnosis in the U.S. had reached 31%, sparking widespread media coverage and patient anxiety.
- 2024: The current re-analysis of all eight RCTs suggests that these mid-period estimates were largely the result of statistical artifacts related to the timing of the studies.
Technical Analysis: Why the Numbers Were Wrong
The "lead time" effect is perhaps the most misunderstood aspect of screening statistics. Imagine a group of 1,000 women. Without screening, 50 of them might be diagnosed with breast cancer over a 10-year period. If you screen them in Year 1, you might find 15 of those cancers immediately. At that moment, it looks like screening has increased the cancer rate by 30%. However, if you follow both groups for 20 years, the unscreened group will eventually "catch up" as their cancers become symptomatic.
"If researchers do not take these factors into account, the initial increase can be mistaken for overdiagnosis," says Elsebeth Lynge, professor emerita at the Department of Public Health, University of Copenhagen. The new study highlights that many of the most-cited papers on overdiagnosis were published while the "catch-up" period was still ongoing, or they failed to account for the fact that the control groups were no longer "pure" because those women had sought screening on their own.
Implications for Public Health and Patient Communication
The downward revision of overdiagnosis rates has profound implications for how doctors talk to patients. For years, clinicians have had to navigate a difficult conversation, telling women that while mammograms save lives, there was a significant chance—perhaps one in three—that a detected cancer might be an "overdiagnosis."
"Most women will not develop breast cancer, but with this study we can now be reassured that the benefits of detecting breast cancer early and preventing premature death will outweigh the small risk of unnecessary treatment," says Professor Njor.
This reassurance comes at a time when screening guidelines are being lowered to age 40 in many jurisdictions due to a rising incidence of breast cancer in younger women. If overdiagnosis were truly as high as 50%, screening younger women (who have more years of life ahead of them for a slow-growing tumor to potentially be found) would be highly problematic. At a rate of less than 5%, the benefit-to-harm ratio shifts heavily in favor of screening.
Responses from the Medical Community
While the study has been welcomed by many radiologists and oncologists, it is expected to face scrutiny from researchers who have built their careers on the overdiagnosis narrative. Traditionally, groups like the Nordic Cochrane Centre have argued that for every life saved by screening, roughly ten women are overdiagnosed and overtreated.
The new analysis challenges this "1-to-10" ratio, suggesting it may be closer to "1-to-less-than-1." This shift could lead to a re-evaluation of patient decision aids—the brochures and websites used to help women decide whether to get screened. Many of these aids currently feature prominent warnings about overdiagnosis based on the older, higher percentages.
Future Directions and Funding
The study was supported by major research institutions, including the Novo Nordisk Foundation and Cancer Research UK. The researchers hope their work will provide a more realistic framework for interpreting medical evidence in the future.
As breast cancer screening moves toward more "personalized" approaches—such as using AI to read mammograms or tailoring screening frequency to a woman’s genetic risk—having an accurate baseline for overdiagnosis is essential. If the baseline risk is indeed below 5%, the focus can shift more decisively toward improving detection technology and ensuring equitable access to screening, rather than debating the fundamental utility of the program itself.
In conclusion, while overdiagnosis remains a biological reality that requires clinical vigilance, this new evidence suggests it is a manageable risk rather than a systemic failure of breast cancer screening. For the millions of women invited for mammography each year, the message is clearer than it has been in decades: the protection offered by early detection is substantial, and the risk of being diagnosed with a "harmless" cancer is significantly lower than previously feared.
