In 1995, the world was taken by surprise with the announcement of a study linking the excessive wearing of tight bras with an increased incidence of breast cancer, published in the book, Dressed to Kill: The Link Between Breast Cancer and Bras. We are the authors of this book.
As a result of this study, women around the world have been wondering if their bras are harming their breasts. Many stopped wearing bras as a result, and many bra companies have since developed less constrictive designs to avoid the problems caused by tight bras. Studies began to confirm the bra-cancer link.
Meanwhile, the bra-cancer link, and the theory of lymphatic impairment and tissue toxification from tight bras, was adamantly resisted by the leaders in the cancer field. The current paradigms of cancer development and treatment do not include lymphatic impairment by tight clothing as a contributing factor. Admitting this factor would upend current cancer theory and practice.
Nevertheless, information about the bra-cancer link continued to spread. In response to the progressive public acceptance of this problem with bras, a study was funded in 2014 to essentially discredit the theory.
The Fred Hutchinson Cancer Research Center did the study, which was published with great international fanfare and extensive news coverage, announcing that the bra-cancer link is a myth and is refuted once and for all.
The study is called,Bra Wearing Not Associated with Breast Cancer Risk: A Population-Based Case–Control Study. It was hailed, even by the author of the study, as the last word on this issue.
For example, Myth Busted: No link between bras and breast cancer, in USA Today, the author of the study was quoted as saying,
“There is no evidence that wearing a bra increases a woman’s risk of breast cancer,” says Lu Chen, a researcher at Fred Hutchinson Cancer Research Center in Seattle. Chen led the study as part of a broader look at risk factors. She was not surprised, she says, that “there’s just nothing there.”
As we will show, this statement is false, misleading, and biased, and ten years later it is still being used to discredit the bra-cancer link and discourage interest in further research. It is important for women’s health that this paper needs to be retracted for its errors and biases.
Here is an analysis of this paper.
Analysis
The following quotes taken directly from this paper are in italics. This is from the abstract.
Bra Wearing Not Associated with Breast Cancer Risk: A Population-Based Case–Control Study
Despite the widespread use of bras among U.S. women and concerns in the lay media that bra wearing may increase breast cancer risk, there is a scarcity of credible scientific studies addressing this issue. The goal of the study was to evaluate the relationship between various bra-wearing habits and breast cancer risk among postmenopausal women. We conducted a population-based case–control study of breast cancer in the Seattle–Puget Sound metropolitan area that compared 454 invasive ductal carcinoma (IDC) cases and 590 invasive lobular carcinoma (ILC) cases diagnosed between 2000 and 2004 with 469 control women between 55 to 74 years of age. (Emphasis added.)
Here is error number one.
Misleading Title
Note that the title, and media about this study, do not mention a key and essential feature of this study, which is that it only looks at postmemopausal women, ages 55-75. Results from postmenopausal women have unknown application to premenopausal women. In fact, as will be shown below, this study deliberately excluded premenopausal women, and in the text of the study it is admitted that the findings only apply to postmenopausal women.
This essential fact should be in the title. Bra Wearing Not Associated With Breast Cancer in Postmenopausal Women. The title implies the results are applicable to all women. Clearly, the title was designed to overstate the issue, a first hint at the bias of this study. Not everyone looks at the study, and sees that the title tells a different story.
“Control Women”
According to the study, the “control women” are age matched with the groups of women with cancer. But what kind of control group is this, really?
This study is to look at bra use and cancer incidence in women over 55. That means a control group must exist, not only for age, but also for bra usage and for cancer incidence. Therefore, there must be bra-free women for comparison to women who wear bras, and there must be women who do not have breast cancer to compare with those who have had breast cancer (10-13 years ago).
However, this study does not have a bra-free control group. These women are all bra users, as the authors state under Materials and Methods. “There was one participant who reported that she never wore a bra and she was excluded from the analysis. There were seven women who did not currently wear a bra and they were included in our lifetime bra wearing analyses but excluded from the analyses of current bra-wearing habits.”
In the Discussion section it explains, “Because bra wearing was ubiquitous among our participants, we were unable to compare risks among women who never wore a bra to those who regularly wore a bra, and instead, our primary comparison was based on average number of hours per day women wore a bra.”
In addition, the “control women” have not been checked for having breast cancer. They were asked a health history, but were not examined. Hence, an unknown number of these women may have undiagnosed cancer.
In summary, this study does not really study the impact of bra usage, since there is no bra-free control group. It does not study how many “control women” have undiagnosed cancer. And many of the participants in this postmenopausal cancer study were premenopausal when diagnosed between 2000 and 2004, and some were not.
As the paper continues, we see bias admitted in the study purpose.
Bias
To our knowledge, the only epidemiologic evidence on bra wearing and breast cancer risk comes from a case–control study published in 1991, which reported a nonstatistically significant two-fold higher risk among premenopausal women who wore a bra versus those who did not, but no elevation in risk was observed for postmenopausal women. Given that questions in the lay media have been raised about breast cancer risk and bra wearing, we evaluated relationships between various aspects of bra wearing and breast cancer risk among postmenopausal women enrolled in a population-based case–control study. (Emphasis added.)
The 1991 study, called Breast Size, Handedness, and Breast Cancer Incidence, in the European Journal of Cancer, used bra size to estimate breast size, and discovered that bra-free women had half the incidence of breast cancer compared to bra users. Put differently, bra users had a 100% higher incidence of breast cancer compared to bra-free women. The numbers in the study were too small to make this statistically significant, as the Hutchinson authors point out, but it is significant, enough that the 1991 study pointed out this finding in their abstract. “Premenopausal women who do not wear bras had half the risk of breast cancer compared with bra users (P about 0.09).”
Note that this finding was found for premenopausal women in that study, and not postmenopausal women. Older women did not show a bra-cancer link in the 1991 study. The authors of the Hutchinson study knew that, and designed their study to confirm the negative finding of a bra-cancer link in postmenopausal women, not look at the positive finding of a bra-link in premenopausal women.
If you want to study a bra-cancer link, use the group for which this link was positive.
This reveals a bias and agenda, which is further revealed by a statement from Fred Hutchison about this study. “We weren’t really surprised,” Chen said. “We knew that the biological plausibility of a link between bras and breast cancer was really weak.”
Bad Data
This study assessed bra usage through asking questions to women 55-75 years old about details of bra usage since they were teens and until the time they were diagnosed with cancer, or an equivalent age for the “control women”.
“Women were asked a series of structured questions to assess lifetime patterns of bra wearing, including bra cup sizes and band sizes, age at which they started regularly wearing a bra, whether they wore a bra with an underwire, number of hours per day, and number of days per week they wore a bra at different times in their life, and if these patterns ever changed.”
That’s a lot of recall for 50-60 years of bra usage. Bra styles changed since the early 1940s, when some of the participants were born. Body size also changed. And many women wear the wrong size bra, as the lingerie industry attests. Bra sizing is variable between bra makers and different styles. It is weak evidence when you rely on the memory of women in their 60’s and 70’s to recall the details of their bra usage since puberty.
The authors admit this in their Discussion section. “It is important to acknowledge some of the limitations of this study. Data on bra-wearing habits were all self-reported, which are subject to recall bias and/or nondifferential misclassification.
However, the authors hedge. “This said, there is no more reliable measure of this exposure other than self-report.” Of course, no evidence is offered to show the reliability of these reports, since you can’t confirm past bra usage for the last 50-60 years.
We also observed bra-wearing habits to be relatively stable over a woman’s lifetime (e.g., 47.6% of women reported that their bra-wearing habits never changed over their lifetime), which may make the recall task less complex and, thus, improve accuracy in self-reporting these data.”
This statement makes no sense. If the accuracy of this self-reported data needs improvement, then how can you rely on the accuracy of statements about stable lifetime bra usage habits, which are self-reported? Without objective verification of their past bra usage, you cannot confirm the accuracy of any of their statements.
Results
To determine whether bra use time was related to cancer incidence, the women were divided into groups depending on time worn. The time categories were very close, making distinctions extremely difficult, especially when this is considered the sole daily bra usage time for women wearing bras for over 50 years.
For the lifetime average daily bra usage, the time categories used were 10 hours or less, 10.1 -11.5 hours, 11.6-13.9 hours, and 14 hours or more. They did not find any statistically significant difference between the cancer groups and the “control women” using this time categorization.
Again, note that there is no group of zero hours of bra usage, which would be the control group.
Their results mean that in lifetime bra users between 55-75, you cannot tell any statistically significant differences between women who wear bras for 10 hours compared to 11.5 hours, or even compared to 14 hours when it comes to developing breast cancer. Therefore, they conclude, the time each day a bra is worn does not impact breast cancer incidence.
This is like studying cigarette smoking in old smokers and looking for a difference in lung cancer incidence, depending on whether they smoked an average of 20 cigarettes each day for 60 years, or 22 cigarettes a day, or 30 cigarettes a day, and not including any non-smokers for comparison, and, upon not finding a statistically significant difference, saying there is no smoking-cancer connection.
Of course, if they compared lifetime smokers to lifetime non-smokers, you will see a big difference. Lifetime bra users must also be compared to lifetime bra-free women to have any meaningful comparison.
This leads us to a basic problem with this study design and its selection of postmenopausal women. Choosing this group to confirm a negative impact of bras creates another bias that the authors did not admit.
Survivor Bias
What the authors do not acknowledge is the existence of a survivor bias when using older people who have been practicing a lifestyle daily for 50 years or more. Time selects out the more susceptible people, leaving survivors. Studying the survivors will underestimate the hazards of a lifestyle. See survivor bias.
We raised this issue to Fred Hutchinson. When the authors of this study were preparing to announce their findings, we were personally contacted by Fred Hutchinson’s communications department, asking for our response prior to their publishing. They sent us an advanced copy of their study to get our objections so they could address them in their upcoming press release. We asked why they were excluding premenopausal women, when the 1991 study they referenced found a bra-cancer link in younger women. We suggested their design to confirm a negative finding and ignore the positive one reflected a bias and agenda.
Fred Hutchinson released a statement soon thereafter, saying, “The “Dressed to Kill” authors aren’t backing down. In an email, Sydney Ross Singer questioned why the Fred Hutch authors studied post-menopausal women rather than pre-menopausal women and said that their conclusion “suggests a pro-bra bias.”
Not at all, said the Fred Hutch’s Dr. Christopher Li, an epidemiologist and breast cancer expert in the PHS Division. Li, along with fellow PHS epidemiologist Dr. Kathleen Malone and researcher Chen, used data from a larger study of post-menopausal breast cancer patients because most breast cancers are diagnosed in older women – and they also have the longest lifetime exposure to wearing bras.”
Their response shows a lack of appreciation for the survivor bias. It also shows that they were working to refute our research.
Lastly, under bias, it must be mentioned that, while the authors claimed no conflicts of interest in their study, they do not mention that they are women who wear bras. That’s like having smokers study the harms of smoking. It’s a conflict of interest.
Surprising Result
Despite the apparent design to show no evidence of a bra-cancer link, this study did find a link.
(T)here was some suggestion that women who wore an A-cup bra had increased risksof both IDC (OR, 1.9; 95% CI, 1.0–3.6; P = 0.039) and ILC (OR, 1.8; 95% CI, 1.0–3.3; P = 0.057)…”
In other words, the proportion of women in the cancer groups wearing A-cup bras was nearly 90% higher than the proportion of women in the “control” group wearing A-cup bras.
This statement contradicts the conclusion and title of their paper.
The authors reported these results on women who wore A-cup bras, but then hedged that, “…neither these trends nor trends for any of the other continuous variables assessed were statistically significant.
That is not true, as we shall show.
Keep in mind that statistical significance tells how reliable the results are. Very significant results can be statistically insignificant. It’s a reflection of the study design, not the situation being studied.
A “P” value of less than 0.05 means that you can be 95% confident of the results. A P value of 0.01 means there is a 99% confidence. A P value of 0.1 gives a 90% confidence.
In this study, the P value for IDC is 0.039 in A-cup women, which is a P number below 0.05 and is statistically significant. And for ILC, the P number is close, at 0.057.
Taken together, these cancer groups have a statistically significantly higher rate of A-cup bra users compared to the “control women”. The authors of the study said this was insignificant.
Unfortunately, the lack of a bra-free control group, and all the other errors exposed above, make these results meaningless.
There was no data given to show how cup size relates to time worn. Any difference, if it is real, would more likely be due to breast size. But who knows with faulty data, even if statistics say it is significant?
Breast size is considered directly proportional to cancer incidence, with the larger breast having the greater incidence, particularly for postmenopausal women. However, in this study, breasts in A-cups have a higher link to cancer than larger breasts, the opposite of what you would expect from breast size.
Could it be that A-cup bras are worn more tightly, perhaps to increase the appearance of cleavage? Tightness of the bras was never considered in this study, which is another major flaw. Comparing time worn without regard to tightness is useless.
Tightness Ignored
The impact of bras is a function of tightness and time worn. It is not just about whether a bra has an underwire or not. Any bra can be worn too tightly. It is also possible that women of different breast and body shapes and sizes may choose to wear bras to different degrees of tightness. Research is needed to study this.
But this issue is academic, since the data is based on the recall of senior citizens, with unknown memory abilities, about lifetime bra usage over a 50-60 year period. And none of them were compared to women of the same age who were bra-free their entire lives.
This study is useless, misleading, misinformation, and needs retraction. Currently, it is confusing and misleading the public, doctors, and cancer researchers.
Since this study was done there have been numerous studies done internationally which show a bra-cancer link. See resources below for references.
Biomechanics research is showing the mechanism for this link, and how tight bras interfere with the lymphatic system to the detriment of immune function, increasing cancer risk.
Unfortunately, this poorly designed, biased, uncontrolled study has been elevated to prominence by an advertising campaign that ran stories about this study around the world as the final word on this controversial issue.
Indeed, the authors of this paper admit in their funding section that “The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.” (Emphasis added.)
In other words, this study and paper is an advertisement defending bras, and discouraging interest in the bra-cancer link, and has been reported around the world to reassure women about the safety of bras.
As the authors conclude, “This is the first study to characterize various bra-wearing habits in relation to breast cancer risk using a rigorous epidemiologic study design. The findings provided reassurance to women that wearing a bra does not seem to increase the risk of the most common histologic types of postmenopausal breast cancer.” (Emphasis added.)
Note that saying this study is reassuring shows a bias towards bras. Given that this is one study, there is no basis for reassurances to women about the bra-cancer link. The study design makes meaningful results impossible, as the study tries to confirm negative findings. Even if they were valid, the results are also of unknown relevance to premenopausal women, a limitation which is not mentioned in the paper.
The authors should have called for further studies that are better designed. Their eagerness to put this issue to rest is clear.
As new research emerges which supports the bra-cancer link, this one study is still being used to discourage interest, 10 years since its publication. But this study is now in stark contrast to dozens of studies that support a bra-cancer link, all done within the past 10 years.
It is imperative for women’s health and the prevention of breast cancer that this sham study be retracted and taken out of publication. Research into the causes of breast cancer needs to consider the biomechanical impact of tight bras on the breasts and how that can lead to cancer development. But this Hutchinson study is getting in the way of that research.
For example, the American Cancer Society is referring to this study when denying the bra-cancer link. “Online and social media rumors and at least one book have suggested that bras cause breast cancer by obstructing lymph flow. There is no scientific or clinical basis for this claim, and a 2014 study of more than 1,500 women found no association between wearing a bra and breast cancer risk.”
Note that the limitations of this study are not mentioned by the ACS, such as the absence of a bra-free control group, or that the results from postmenopausal women may not apply to younger, premenopausal women. Omission of this information is misleading.
Wikipedia has an entry about our book, Dressed to Kill, which discredits the book and refers to the Hutchinson study as proof that the bra-cancer theory is discredited. “A study conducted by the Fred Hutchinson Cancer Research Center found “no aspect of bra wearing, including bra cup size, recency, average number of hours a day worn, wearing a bra with an underwire, or age first began regularly wearing a bra, was associated with risks” of breast cancer.The study included detailed studies of women’s lifestyle and bra-wearing habits and found no correlation between bra use and cancer.”
Again, no mention is made of the limitations of this study. Clearly, this study is being misused and misquoted by authoritative sources, making the retraction and removal of this study very important to stop misinformation about this issue.
This work was funded by a grant from the National Cancer Institute R01 CA 85913 (to C.I. Li and K.E. Malone).
It is unfortunate that the public cannot be reimbursed for the taxpayer funds that paid for this flawed, misleading study.
Conclusion:
This study is misleading the public about this vital issue. This Fred Hutchinson study is part of a medical industry effort to suppress and censor the bra-cancer link. This study should be retracted, and all those involved in its publication and dissemination should be held responsible for the pain, suffering, and deaths caused by keeping this information from women.
Clearly, the public cannot trust these cancer industry leaders to give truthful information about breast cancer, or to have the integrity to retract this propagandistic study.
Resources:
References to studies which show a bra-cancer link can be found at
https://brasandbreastcancer.org/supportive-references
See also,
The Biomechanics of Bras and Lymphatics and the Link to Breast Cancer
Dressed to Kill: The Link Between Breast Cancer and Bras, Second Ed.
Tight Clothing Affects Tumor Microenvironment
What Breast Cancer Inc. Doesn’t Want You to Know about Bras
Fake Breast Cancer News from the American Cancer Society