A new study…not very neutral
Researcher response
Expression of concerns
The BMJ recently published a study https://www.bmj.com/content/390/bmj-2025-085029
This study would show that women who did not participate in the first screening would represent a large population at greater risk, in the long term, of dying from breast cancer.
The problem is that this study is not a comparison between women who would have been randomly selected to be invited or not to be screened, as is done in randomized controlled trials for example.
Women who did not attend this first mammogram are a group with particular characteristics:
– they also more often missed subsequent appointments
– they more often had cancers detected outside of screening
– they more often have advanced stage cancers (III or IV)
– they have a higher mortality rate from breast cancer.
The authors of the study concluded:
“Our study shows that first screening non-participants represent a large population at an elevated risk of dying from breast cancer decades in advance. This increased mortality is modifiable and primarily attributed to late detection. Targeted interventions are warranted to boost adherence to mammography screening and decrease the mortality risk for those who did not participate in the first screening. “
However, they forget to mention that in all the studies, women who do not get screened are on average of a lower socio-professional level than women who get screened, and they have more frequent and more significant risk factors, such as living conditions, poorer hygiene and dietary habits, etc. If we looked for other pathologies, we would also find that they have a higher risk of stroke or myocardial infarction! This study of descriptive epidemiology can in no way be considered as showing a causal relationship. Poor compliance with medical monitoring of these women, whatever it may be, can be correlated with poorer health in general linked to psychosocial factors. We would also find, in this population, poor medication intake, diabetes and other risk factors for diseases… It is simply a correlation between particular characteristics of a population and its risk of mortality from cancer; once again, correlation should not be confused with causality. By not mentioning this in their conclusion, the authors are behaving like intellectual crooks.
Obviously, the mainstream press, particularly the so-called French women’s press, with its poor scientific background, reports the study without much critical analysis.
But every October we are treated to the famous pink propaganda that is rife every year, with slogans, biased studies, and articles by opinion leaders burdened with vested interests.
An editorial in the BMJ, a researcher’s response
On this same BMJ journal publishes a editorial to highlight this study, which is nevertheless questionable and has fallacious results, which is causing a answer by Nordic researcher Peter Gotsche, Danish physician and researcher, director of the Nordic Cochrane Centre (group of independent experts) and co-founder of the Cochrane Collaboration .
26 September 2025
Peter C. Gøtzsche-Professor emeritus Institute for Scientific Freedom, Copenhagen @PGtzsche1
An editorial in BMJ claimed that “Mammograms can detect breast cancer early, often before a lump can be felt, which improves the chances of successful treatment and survival” (1).
This is false.
First, mammography screening does not detect cancers early but very late. The average tumour size in the randomised trials was 16 mm in the screened groups and 21 mm in the control groups (2). It takes only one more cell division for a 16-mm tumour to become one of 21 mm. If we assume that the observed doubling times are valid from initiation till the tumour becomes detectable, the average woman has harboured the cancer for 21 years before it acquires a size of 10 mm (2).
Second, in screening propaganda, “successful treatment” usually means less invasive treatment (3), which is also false. Because of substantial overdiagnosis, and because the earliest cell changes, carcinoma in situ, are often diffusely spread in one or both breasts, screening increases mastectomies (4,5).
Third, screening does not improve survival. The editorialist claims that screening reduces breast cancer mortality by 15% and then makes the error of equating this with a reduction in mortality. Breast cancer mortality is a flawed outcome that favours screening, mainly because of differential misclassification of cause of death, but also because treatment of overdiagnosed women increase mortality (3,4), and screening does not reduce total cancer mortality (including breast cancer), or total mortality (4). The newest data showed that for the trials with adequate randomisation, the risk ratio was 1.00 (95% confidence interval 0.96 to 1.04), for total cancer mortality, and 1.01 (0.99 to 1.04), for all-cause mortality (6).
The editorialist talks about “potential overdiagnosis.” It is not potential; it is an unavoidable consequence of screening (2-5).
Moreover, the editorialist claims that an observational study (7) provides “concrete evidence that initial screening reduces mortality.” This is false. The study only claimed that screening reduces breast cancer mortality. It is a huge error that the authors of this study, which was performed in Sweden, did not tell their readers about cancer mortality and total mortality, which would have been very easy to document.
Screening doesn’t reduce mortality and observational studies can never demonstrate reliably that screening reduces breast cancer mortality. They are all biased by the healthy screening effect, which no amount of statistical adjustment can make up for. We should ignore observational studies claiming that mammography screening works. And we should abandon mammography screening, as it is harmful (3).
1 Zhen-qiang M. Participation in early mammography screening: Enduring benefits at a population level. BMJ 2025; 390:r1893.
2 Gøtzsche PC, Jørgensen KJ, Zahl PH and Mæhlen J. Why mammography screening has not lived up to expectations from the randomised trials. Cancer Causes Control 2012;23:15-21.
3 Gøtzsche PC. Mammography screening is harmful and should be abandoned. J R Soc Med 2015;108:341-5.
4 Gøtzsche PC and Jørgensen KJ. Screening for breast cancer with mammography. Cochrane Database Syst Rev 2013;6:CD001877.
5 Jørgensen KJ, Keen JD and Gøtzsche PC. Is mammographic screening justifiable considering its substantial overdiagnosis rate and minor effect on mortality? Radiology 2011;260:621-7.
6 Gøtzsche PC. Screening for breast cancer with mammography. Copenhagen: Institute for Scientific Freedom 2023;May 3.
7 Ma Z, He W, Zhang Y, et al. First mammography screening participation and breast cancer incidence and mortality in the subsequent 25 years: population based cohort study. BMJ 2025;Sep 24;390:e085029.
Expressions of concerns
Expressions of concern placed on research paper and linked editorial on attendance at first breast screening appointment and breast cancer death
Expressions of concern have been placed on a research paper showing an association between attendance at first breast screening appointment and breast cancer death and its linked editorial, both of which were press released and published by The BMJ on 24 September 2025.
The decision follows concerns raised after publication that messaging in key areas of these articles may not be sufficiently supported by the data presented or by the wider evidence base.
BMJ is in discussion with the authors about what post-publication change to their work is required to ensure that it accurately reflects the results and other relevant evidence, and is transparent about uncertainties.
Links to articles and Expressions of Concern: https://www.bmj.com/content/391/bmj.r2394
1-Ma Z, He W, Zhang Y, et al. First mammography screening participation and breast cancer incidence and mortality in the subsequent 25 years: population based cohort study. BMJ 2025; 390 DOI: 10.1136/bmj-2025-085029
Expression of Concern
2-Ma Z. Participation in early mammography screening. BMJ 2025; 390 DOI: 10.1136/bmj.r1893
Expression of Concern
🛈 Nous sommes un collectif de professionnels de la santé, rassemblés en association. Nous agissons et fonctionnons sans publicité, sans conflit d’intérêt, sans subvention. Merci de soutenir notre action sur HelloAsso.
🛈 We are an French non-profit organization of health care professionals. We act our activity without advertising, conflict of interest, subsidies. Thank you to support our activity on HelloAsso.
