Breast cancer worry, uncertainty, and perceived risk following breast density notification in a longitudinal mammography screening cohort

High mammographic breast density (BD)—relatively large opaque white areas of mostly fibroglandular tissue on a mammogram—not only raises breast cancer risk but also makes it more difficult for radiologists to detect breast cancer on a mammogram [1, 2]. These factors prompted the passage of state-level legislation designed to inform women if dense breasts were identified on their screening mammogram and a federal amendment to the Mammography Quality Standards Act to mandate the disclosure of BD information to women and their providers [3, 4]. The goal of most dense breast notification (DBN) legislation is to increase awareness of women’s own BD and BD’s impact on breast cancer detection and risk and to inform women of their potential need to obtain supplemental imaging screening via ultrasound and/or magnetic resonance imaging (MRI) techniques, for example [5‐7].

In addition to structural factors such as healthcare access and coverage of costs for supplemental testing, any effect of DBN on breast cancer screening behavior is contingent upon increasing women’s awareness of breast density and its implications as well as women’s cognitive and emotional appraisal of this information [6]. A recent systematic review emphasized that overall, BD awareness tends to be low, and some, but not all, studies found that women residing in states with DBN laws report greater knowledge of general or personal breast density, slightly increased knowledge of density’s masking effect, and no changes in knowledge of the increased risk of high BD on breast cancer [6, 7]. Fewer studies have examined psychological appraisal and responses to DBN. The lack of consensus on screening recommendations for women with dense breasts and providers’ limited training in breast density counseling [8, 9] have raised concerns that DBN may trigger negative emotional and cognitive responses, such as increasing inaccurate risk perceptions, worry, and uncertainty about breast cancer screening choices that may negatively impact women and in turn inadvertently reduce screening participation [7, 10, 11]. Qualitative studies have found some level of increased anxiety or worry as well as confusion over the next steps, while cross-sectional surveys and one study of pre-/post-legislation data revealed no difference in breast cancer worry or fear by DBN [6, 7]. Research is also lacking as to whether any initial psychological reactions in response to notification persist long enough to influence screening behavior at the time of the next screening mammogram [12]. Finally, although prior research has found that Hispanic/Latina (hereafter Hispanic) women, women with lower education levels and with limited English proficiency have lower understanding of breast density, limited empirical research has examined whether women’s psychological responses to DBN may also vary by these factors [12‐15]. Investigating these questions in racially, ethnically, and educationally diverse populations can provide important data for identifying population groups that may benefit from additional educational and clinical support.

In a longitudinal study within a majority Hispanic immigrant screening cohort [13], we specifically evaluated whether psychological outcomes such as breast cancer worry, perceived absolute and comparative risk, and uncertainty about breast cancer risk and screening choices varied in short-term (approximately 1–3 months) and long-term (approximately 9–18 months) surveys collected after receiving a screening mammogram according to women’s prior BD awareness and dense breast status as a proxy for DBN. We assumed that women with mammographically dense breasts had received DBN after their baseline screening mammogram and explored modification by individual-level sociodemographic characteristics.

The New York Mammographic Density (NY MaDe) Study is a study of breast cancer risk factors and screening in a mammography cohort of women in Northern Manhattan, New York City, and has been previously described [13, 16, 17]. Between 2016 and 2018, we recruited 812 women, ages 40–60 years, during their screening mammography appointment at a clinic in Northern Manhattan. Trained bilingual staff conducted in-person baseline surveys with participants in their preferred language (33% in English, 67% in Spanish) to collect information on sociodemographic characteristics, breast cancer risk factors, breast cancer screening history, and medical and lifestyle history. The questionnaire included questions on women’s awareness and knowledge of breast density and breast cancer-related psychological factors such as worry, uncertainty, and absolute and comparative risk.

We extracted clinical mammography reports since January 2013 to obtain the dense breast status of their screening mammograms post-implementation of New York’s DBN (in 2013). In accordance with New York State legislation, the recruiting mammography clinic provided women with dense breasts the state-mandated additional DBN text along with their clinical mammography results, provided in both English and Spanish [13]. We contacted all participants via mail about 1–3 months (short-term follow-up, mean 1.6 months) from the baseline interview to allow them enough time to receive their baseline mammography reports, and again at 9–18 months (long-term follow-up, mean 14 months) from baseline. Bilingual staff administered these interviews by phone or in-person at the mammography clinic, or surveys were self-administered by mail or electronically through a secure link. The baseline and follow-up surveys included questions about breast cancer-related psychological factors. The short-term follow-up was completed by 612 women (75%), and the long-term follow-up was completed by 630 women (77.6%); 538 women (66.3%) completed both follow-ups, while 166 women (20%) completed the baseline and only one of the follow-up surveys. We excluded 4 women with a history of breast cancer and 1 woman for whom mammography reports could not be obtained. All women who completed any follow-up survey were retained for analysis of that survey (n = 607 for short term and 626 for long term).

The Columbia University Medical Center institutional review board approved this study, and all women provided written informed consent prior to data collection.

We investigated whether informing women of their dense breasts may increase women’s anxiety, perceptions of risk, worry, and uncertainty about breast cancer risk and screening, as suggested by qualitative studies [10, 11, 27]. In this sociodemographically diverse screening cohort, we found some evidence that awareness of breast density is associated with increased perceived absolute risk about 1–3 months following a screening mammogram, but that this increase is not observed over the longer term. This short-term association was independent of whether a woman had dense breasts on her own mammogram and hence had received notification. DBN was associated with decreased perceived absolute risk of breast cancer at short-term among women with some college education or greater, but increased perceived absolute risk among women with a high school or less education. We also observed effect modification of the associations of short-term uncertainty with breast cancer risk and screening choices by preferred interview language. Women who were aware of breast density and whose dominant language was Spanish reported greater uncertainty about both breast cancer risk and screening choices than those who were unaware, but no greater uncertainty was reported among those whose dominant language was English. There were also no statistically significant associations between awareness and dense breast status and any emotional or cognitive psychological responses at long-term follow-up (at approximately 1 year after baseline).

The particular outcomes for which we observed associations (absolute risk and uncertainty about breast cancer risk and screening choices) align with the content of the mandated DBN text in the New York State legislation [3, 5]. This text advises women that BD is a risk factor for breast cancer and that it poses limitations to detecting tumors on screening mammograms; these messages may, respectively, influence perceptions of personal risk of breast cancer and uncertainty about screening options shortly after receiving a mammography report containing DBN. Therefore, our results suggest that DBN may be effective in changing women’s cognitive responses that are specific to the information that is communicated to them through DBN. Uncertainty is inherent in most breast cancer screening and risk reducing interventions (e.g., genetic testing). The impact of this uncertainty on women’s psychological states and behavior may depend on personal characteristics such as race [12], but also tolerance to uncertainty around healthcare. While our main analysis suggested opposing directions of effect for DBN and BD awareness, this pattern of associations remained in our stratified analysis only among highly educated women. These effects are difficult to explain as we were not able to tease out more discrete combinations of DBN and BD awareness in our stratified analysis due to small sample sizes. It is possible that higher educated women who have received DBN may be more likely to be risk assessed and knowledgeable about their objective risk, and thus more likely to be reassured by a negative mammogram finding or DBN text; however, our findings and this explanation merit further investigation. Whether increasing perceived risk and uncertainty about breast cancer screening through DBN have adverse consequences for women’s psychological or behavioral outcomes remains unclear; improving the communication and clinical follow-up of the DBN information or clear screening guidelines for women with dense breasts may avoid negative consequences. Given our observed differences by women’s education and language, these efforts would also need to be tailored appropriately to reach all population subgroups and avert potential disparities in breast cancer screening and other outcomes. Additionally, though we did not have the power to explore this fully, women who were aware may have been more likely to remain stable in their reporting of their perceived absolute risk from baseline to short term versus decreasing during that time. This raises the possibility that the greater absolute risk we observed with awareness at short term may result from unaware women decreasing their reported perceived absolute risk after a negative mammogram. This should be explored in future studies.

The limited body of empirical literature to date has not established a large or consistent effect of notification on emotional or cognitive psychological outcomes although unlike this study, the majority of prior research was small qualitative studies immediately after reading DBN text, or examined cross-sectional associations, and focused on BD knowledge and awareness [6, 7, 13, 18, 28]. Cross-sectional studies may not be able to differentiate effects of DBN or awareness from baseline characteristics such as whether some woman have greater tendencies toward greater worry or risk perceptions. In our study, the use of repeated assessment of psychological outcomes starting with the mammography appointment and over the course of about 1 year allowed for adjustment of baseline psychological factors. We were also able to assess psychological outcomes at each timepoint and observed that the increased perceived risk at short term with BD awareness was not observed around the time of the next mammogram (i.e., 12–18 months following receiving mammography). Although cross-sectional studies show mixed results with DBN and perceived risk [6], several small studies assessing short-term responses to DBN have reported results in alignment with our findings. In one randomized control trial of women with dense breasts who either did or did not receive DBN, women who received DBN were less likely to report “a lot lower” perceived comparative risk (10.5% vs. 15.5% for women who did not receive DBN) both 4 weeks and 6 months after DBN receipt, while worry and absolute risk perception did not significantly differ with DBN [29]. Another study assessing women before and immediately after reading a DBN also found significantly greater perceived lifetime risk of breast cancer after notification [30]. Our finding that education and preferred interview language modified the associations adds new data for these outcomes and is broadly consistent with previous literature that found variation of the impact of DBN legislation on other factors such as awareness, knowledge, and understanding of breast density, including by educational attainment [6], suggesting that the messaging of the DBN text and breast density awareness in general may be complex or, not uniformly understood across subgroups, and that more tailored patient education may be appropriate. Interventions to accompany DBN such as additional written information or community health worker interpersonal interactions as a way to clarify DBN messaging, address women’s questions, and facilitate follow-up with health care providers are being evaluated and may be successful at increasing awareness of BD [15], but more research is necessary to assess the impact on psychological outcomes.

We ascertained DBN status using BI-RADS data for dense and non-dense breast classification levels retrieved from medical records, which provides an accurate measure of women’s eligibility for DBN. However, we did not assess whether women who had mammographically dense breasts at their baseline mammogram received, read, or absorbed the notification information. We also did not have a clear assessment of women who were newly made aware of breast density to see the impact of notification. Although we were able to use medical records to see how many times women had dense breasts on their mammogram in the past, this only accounted for mammograms done at the recruiting facility. Finally, a full exploration of multiplicative interaction with education and our 4-level combination density and awareness variable was impossible due to sparse cell size at this level. Our study was strengthened by the use of an ethnically and racially diverse cohort with a range of education levels, allowing us to explore effect modification by education and language. We were also strengthened by our use of validated constructs for psychological outcomes, and that we assessed these psychological outcomes in the same women at multiple timepoints.

In a predominantly Hispanic screening cohort, awareness of breast density, regardless of one’s own dense breast status, appears to increase one’s perceived risk of breast cancer for a short time after undergoing mammography, but was otherwise not associated with emotional or cognitive psychological factors such as worry or long-term changes. Women with lower educational attainment had increased perceived absolute risk if they were eligible for dense breast notification, and women with a dominant language of Spanish showed increased short-term uncertainty about breast cancer risk and screening options if they had BD awareness. Taken together, these findings suggest that women with lower educational attainment or with language barriers or lower acculturation are at greater risk for short-term uncertainty among breast cancer choices with BD awareness and perhaps could specifically benefit from outreach clarifying the implications of breast density on screening choices.