Mental Illness and BRCA1/2 Genetic Testing Intention Among Multiethnic Women Undergoing Screening Mammography

Genetic counseling and testing for BRCA1/2 gene mutations and other cancer susceptibility genes related to hereditary breast and ovarian cancer (HBOC) (ATM, BARD1, BRIP1, CDH1, CHEK2, EPCAM, MLH1, MSH2, MSH6, NBN, NF1, PALB2, PMS2, PTEN, RAD51C, RAD51D, STK11, and TP53) are the standard of care for women who meet personal or family history criteria (National Comprehensive Cancer Network [NCCN], 2019). The application of multigene panel testing for hereditary forms of cancer has rapidly changed the clinical approach to genetic testing for at-risk patients and their families (NCCN, 2019). Women who have a deleterious BRCA1/2 gene mutation have a 69%–72% absolute risk of developing breast cancer by age 80 years, compared to a 12% lifetime risk in the general population (National Cancer Institute, 2018). Therefore, the U.S. Preventive Services Task Force ([USPSTF], 2019) recommends that primary care providers assess women with a personal or family history who have an ancestry associated with BRCA1/2 gene mutations with an appropriate brief familial risk assessment tool.

Risk management options include intensive breast cancer screening (Saslow et al., 2007), risk-reducing surgeries (Isaksson et al., 2019), and chemoprevention (Cibula, Zikan, Dusek, & Majek, 2011), which have been shown to improve early detection and reduce cancer incidence and mortality. As reported by Hughes (2017), most at-risk women have yet to be tested. Childers, Childers, Maggard-Gibbons, and Macinko (2017) found that, among 3.8 million survivors of breast and ovarian cancer in the United States, only 14% had been tested. Despite the clinical availability of BRCA1/2 genetic testing for more than 20 years and its associated benefit, there is significant underuse of genetic testing. Less than 20% of eligible women screened in primary care are referred for genetic counseling; of those, only 8% undergo genetic testing (Kurian et al., 2017). Other studies have found that this is particularly true for those of lower socioeconomic status and racial/ethnic minority groups, such as Black and Hispanic women (Jones, McCarthy, Kim, & Armstrong, 2017; Tang et al., 2017; Underhill, Jones, & Habin, 2016). Although Hispanic women are at lower risk for breast cancer than non-Hispanic White women, they have the second highest prevalence of BRCA1/2 gene mutations after Ashkenazi Jews (Weitzel et al., 2013). In addition, from 2006 to 2015, there was an increase in the breast cancer incidence rate (0.4%) annually among Hispanics, while this rate remained stable in non-Hispanic Whites (American Cancer Society [ACS], 2018). Breast cancer remains the leading cause of death among Hispanic women, with an estimated 3,200 deaths in 2018 (ACS, 2018).

Few studies have focused on mental illness and genetic testing, which demonstrates a need to address this gap. Major depressive disorder (MDD) is the most commonly diagnosed form of depression, affecting 16.2 million adults annually; the prevalence is higher among women (National Institute of Mental Health [NIMH], 2019). Anxiety disorders are the most common mental illnesses, affecting 40 million adults aged 18 years or older (Anxiety and Depression Association of America, 2017). In addition, data suggest that MDD is highest among adults from racial or ethnic minorities (NIMH, 2019). Individuals with mental illness are also at greater risk for poor health and inadequate healthcare access (National Council for Behavioral Health, 2018). Depression and anxiety symptoms also can influence patients’ health behaviors and lead to treatment nonadherence (Nipp et al., 2017). Lack of support for women with mental illness may deter them from accessing preventive healthcare services (World Health Organization, 2017), such as breast cancer screening and genetic counseling and testing. In addition, it is unclear how a mental illness diagnosis and symptoms affect intention to complete BRCA1/2 genetic testing.

In the authors’ previous work, they conducted a retrospective cross-sectional study to evaluate the impact of mental illness among 308 multiethnic women with newly diagnosed breast cancer who were eligible for genetic testing and seen in an academic urban medical center (Ackerman, Shapiro, Coe, Trivedi, & Crew, 2017). The authors found that 57% of women who met NCCN guidelines for BRCA1/2 testing underwent genetic counseling. In addition, mental illness did not affect the completion of genetic counseling. The current study builds on this research by exploring the impact of mental illness on intention to have genetic testing among predominantly Hispanic women. This study is guided by the theory of planned behavior (TPB) (Ajzen, 2011; Roncancio et al., 2015), one of the most widely used socio-cognitive theories, which encompasses behavioral intention (i.e., the stronger the intention to perform the behavior, the more likely the behavior will be performed). Because the purpose of the study is to determine factors that increase behavioral intention, no other constructs of TPB were explored. Genetic testing intention is a necessary first step to genetic testing use. Therefore, the aim of this study is to examine associations between validated measures of mental illness symptoms, patient-reported mental health history, and BRCA1/2 genetic testing intention in multiethnic women at high risk for breast cancer.

Methods

Sample

From November 2014 to June 2016, the authors approached women during their screening mammography visit at the Avon Foundation Breast Imaging Center at Columbia University Irving Medical Center (CUIMC) in New York. This center provides screening mammography to about 15,000 women per year in the Washington Heights, New York, catchment area, which serves a diverse patient population with a predominantly Medicaid/Medicare payer mix. Participants consented to participate in a survey study, called Know Your Risk: Assessment at Screening (KYRAS) for breast cancer, at the time of screening mammography (McGuinness et al., 2018). The KYRAS survey included the Six-Point Scale (SPS), a family history screener that determines eligibility for BRCA1/2 genetic testing based on USPSTF guidelines (Joseph et al., 2012; Stewart et al., 2016). Scores greater than 6 on the SPS warrant referral for genetic testing (Stewart et al., 2016). Among those who agreed to be contacted again for future studies, these women were later contacted via telephone for participation in the mental health substudy if they met the following inclusion criteria: aged 18 years or older; met family history criteria for BRCA1/2 genetic testing, based on the SPS family history screener; spoke English or Spanish; and provided verbal or written informed consent. Participants completed an interviewer-administered survey via telephone in English or Spanish; this survey assessed patient-reported mental health history, current depression and anxiety symptoms, and genetic testing intention. This study was approved by the CUIMC Institutional Review Board.

Measures

The authors collected demographic characteristics, such as age, highest education level, and race/ethnicity, and breast cancer risk factors, including parity, age at first live birth, menopausal status, number of first- and second-degree relatives with breast cancer, and any blood relative who ever tested positive for a BRCA1/2 gene mutation, based on the SPS family history screener (Stewart et al., 2016). Perceived breast cancer risk was measured with one item that asked, “Compared to other women the same age, do you think your chance of getting breast cancer is: higher, same, or lower?” (Lipkus et al., 2000). The survey also included validated measures of anxiety and depression and patient-reported mental health history. The Center for Epidemiologic Studies Depression Scale (CES-D) is a widely used and validated self-report scale designed to measure symptoms associated with depression experienced in the past week using 20 items (Radloff, 1977). Psychometric equivalence of the CES-D has been previously studied within the heterogeneous population of Hispanic women (Hahn, Kim, & Chiriboga, 2011; McCabe, Vermeesch, Hall, Peragallo, & Mitrani, 2011). The Generalized Anxiety Disorder–7 (GAD-7) is a widely used and validated seven-item screen for generalized anxiety disorder (Terrill, Hartoonian, Beier, Salem, & Alschuler, 2015); Mills et al. (2014) reported that the GAD-7 in Hispanic men and women was reliable and structurally valid with strong internal consistency and reliability (a = 0.93). Patient-reported mental illness history was assessed with the following items that were previously used in a national survey on drug use and health (Substance Abuse and Mental Health Services Administration, 2018):

•  Have you ever had a serious mental illness or emotional problem? (yes or no)

•  Have you ever seen a psychiatrist, psychologist, social worker, or other health professional for a psychological or emotional problem? (yes or no)

•  Have you ever stayed overnight or longer in a hospital or treatment facility because of any mental or emotional problem? (yes or no)

•  Has a doctor ever given you any medicine for a psychological or emotional problem? (yes or no)

The primary outcome was BRCA1/2 genetic testing intention (Kessler et al., 2005) and was based on TPB. Using a single-item measure, the authors asked participants, “At the present time, which of the following statements describes your thoughts about having genetic testing for susceptibility to breast cancer?” Responses ranged from 1–6 and included the following:

•  “I have not thought about it.” (1)

•  “I definitely will not get tested.” (2)

•  “I probably will not get tested.” (3)

•  “I probably will get tested.” (4)

•  “I definitely will get tested.” (5)

•  “I was already tested.” (6)

The authors dichotomized genetic testing intention into two groups: those who intended to have genetic testing (“probably/definitely will get tested/already tested,” yes) and those who did not intend to have testing (“definitely/probably will not get tested/have not thought about it,” no). In addition, five participants reported that they already had genetic testing, which was not validated by medical record review; subsequently, the authors conducted a sensitivity analysis that excluded these five participants. Participants who reported that they were already tested were included in the genetic testing intention (yes) group because the authors were unable to perform a medical record review to determine concordance between self-report and completion of genetic testing.

Data Analysis

Descriptive statistics included relative frequencies for categorical/short scale ordinal variables, and means and standard deviations for normally distributed variables. Differences in categorical or ordinal variables between women who reported genetic testing intention and those who did not were assessed using chi-square, Fisher’s exact, or Mantel–Haenszel chi-square tests. Wilcoxon rank-sum tests assessed differences between intention and mental illness scales. Bivariate analysis of genetic testing intention included patient characteristics, such as age, education, race, ethnicity, breast cancer risk factors, validated mental illness symptom measures, and patient-reported mental health characteristics. For variables from bivariate analyses that yielded p values less than 0.2 or that were clinically important predictors of breast cancer risk, the authors devised an initial multiple logistic regression model with GAD-7 and CES-D continuous scores and patient-reported mental illness variables, adjusting for covariates, with intention for genetic testing as the dichotomous dependent variable. The authors then devised a second model with GAD-7 and CES-D continuous scores and patient-reported mental illness variables and adjusted for age and education level. A monitored stepwise procedure for which the variable in the subset yielding the greatest p value exceeding 0.05 was eliminated first. This approach was iterated until the final model included only variables with p values less than 0.05 with the retained mental health variables. Analyses were conducted using SAS, version 9.4.

Findings

Participant Characteristics

Among 18,502 women who had screening mammography at CUIMC from November 2014 to June 2016, 3,558 (19%) were approached for participation in the KYRAS survey study and 3,055 (86% of total approached) enrolled (McGuinness et al., 2019). Demographic characteristics of the enrolled women were similar to the entire screened population, based on electronic health record data (Jiang et al., 2019). Of these women, the authors found that 369 (12%) were eligible for BRCA1/2 genetic testing according to the SPS family history screener. Of these, 269 either declined, were unable to be reached by telephone, or were missing data. Among women eligible for BRCA1/2 genetic testing, 100 women were contacted a median of 171 days (range = 50–288) after enrollment in the KYRAS parent study and agreed to participate in the mental health substudy. A comparison of baseline characteristics of KYRAS participants eligible for genetic testing who enrolled in the substudy compared to those who did not is shown in Table 1. Results revealed that education level and breast cancer risk perception differed between participants who enrolled in the mental illness substudy and those who did not. Women who participated in the mental illness substudy had higher breast cancer risk perception compared to those who did not (34% versus 20%).

Baseline characteristics for the 100 evaluable women are shown in Table 2. Participants had a mean age of 60.65 years (SD = 11.78). The majority were Hispanic (73%), and 59% had no more than a high school education. More than 85% were parous, and more than 80% were postmenopausal. About half of the participants had one or more first-degree relatives who had breast cancer (51%), and few women had a personal history of breast cancer (n = 6) or ovarian cancer (n = 6). Twelve percent of women reported that they did have a blood relative who had tested positive for a BRCA1/2 gene mutation.

Bivariate Analyses

Among all evaluable participants, the majority reported that they intended to complete genetic testing (76%). More specifically, 5% were already tested, 37% reported that they would definitely get genetic testing, 34% reported they would probably get tested, 9% reported they would probably not get tested, 8% reported they will definitely not get tested, and 7% said they had not thought about it. In an unadjusted analysis, the authors found that women who reported genetic testing intention (yes) were younger (mean age of 58.74 years versus 66.71 years, p = 0.007), more likely to have a high school education or less (66% versus 38%, p = 0.014), and more likely to be Hispanic (79% versus 55%, p = 0.031). In addition, on the validated screening measures, women who intended to complete genetic testing had higher mean depression scores on the CES-D (15.39 versus 8.13, p = 0.009) and higher anxiety scores on the GAD-7 (5.63 versus 1.46, p = 0.004) compared to those who did not. From the patient-reported mental illness variables (see Table 3), 32% of the participants reported that they had a serious mental illness or emotional problem; 53% had ever seen a psychiatrist, psychologist, social worker, or other health professional for a psychological or emotional problem; 12% had ever stayed overnight or longer in a hospital or treatment facility because of any mental or emotional problem; and 41% had ever been prescribed medicine for a psychological or emotional problem. In bivariate analyses, the authors found that women who intended to complete genetic testing were more likely to report having a serious mental illness or emotional problem (40% versus 8%, p = 0.004); ever seeing a psychiatrist, psychologist, social worker, or healthcare professional for a psychological or emotional problem (62% versus 25%, p = 0.001); and ever being prescribed medicine for a psychological or emotional problem (47% versus 21%, p = 0.021) compared to those who did not.

Multivariate Analyses

In the multivariate logistic regression model (see Table 4), the authors present associations between validated mental illness measures, patient-reported mental health history, and genetic testing intention after controlling for age, education, race/ethnicity, and breast cancer risk factors. The authors found no statistically significant associations between validated mental illness measures of anxiety and depression, patient-reported mental illness, and genetic testing intention when adjusted for known confounders. Younger age was the only covariate that was significantly associated with genetic testing intention in the initial model. For the second model that included fewer covariates, younger age and less education remained statistically significant. Results for the sensitivity analysis were similar: Younger age and less education remained statistically significant.

Discussion

In the current study, the authors demonstrated that, among a predominantly Hispanic population undergoing screening mammography who met eligibility criteria for BRCA1/2 genetic testing, intentions to complete BRCA1/2 genetic testing for breast cancer risk were high. However, completion of genetic testing was low, with only 5% reporting previously having genetic testing performed. The authors also found a high prevalence of patient-reported mental illness among this multiethnic cohort. In addition, women who intended to complete genetic testing had higher mean depression scores on the CES-D and higher anxiety scores on the GAD-7. Despite this high prevalence of mental illness, a majority (76%) of the participants reported that they intended to complete BRCA1/2 genetic testing.

TPB was useful in understanding genetic testing intention because the theory posits that one’s intention is an indicator of readiness to perform the behavior—in this case, BRCA1/2 genetic testing. Of note, among women who intended to have BRCA1/2 testing performed, 40% reported that they had a serious mental illness or emotional problem and 62% had ever seen a mental healthcare provider for a psychological problem, highlighting the role of psychological functioning in genetic testing. This high prevalence of mental health problems and low completion of BRCA1/2 genetic testing is likely explained by lower socioeconomic status and high stress experienced by racial and ethnic minority women. The current findings are consistent with a previous study (González-Ramírez et al., 2017) that found that 16% of Mexican women undergoing genetic counseling for HBOC had depressive symptomology and 29% had anxious symptomology. Those results suggest that anxious and depressive symptomatology, worries, grief, and sleep problems affect the well-being of participants undergoing genetic counseling. Similar to the current findings, a study by Holden, Ramirez, and Gallion (2014) of 117 Latina breast cancer survivors showed that 32% had CES-D scores above the threshold, about three times those of the general population. Cancer screening rates were extremely low among this cohort, with only 5 (4%) women who screened for ovarian and colorectal cancers. The authors concluded that depressive symptoms may be a barrier to cancer screening.

Although the current authors found bivariate associations between validated measures of anxiety and depression, patient-reported mental illness, and genetic testing intention, mental illness variables were not significantly associated with genetic testing intention in the multivariate adjusted analysis. Instead, only younger age and lower education level were associated with genetic testing intention. Consistent with previous studies, patients who undergo genetic testing tend to be younger (Ayme et al., 2014). This sample was predominantly Hispanic, and the majority had less than a high school education. The authors found that education level was inversely associated with intention to have genetic testing performed. This finding is consistent with another study (Jones et al., 2016) that found an inverse relationship between education level and BRCA1/2 genetic testing; however, that study did not include women of Hispanic ethnicity. One possible explanation for this finding is that women of Hispanic ethnicity with lower education levels may have a trusting relationship with their healthcare providers and be more likely to follow through with their providers’ recommendations to have genetic testing performed, when healthcare system–related barriers are removed. In this same KYRAS screening cohort, the authors found that Hispanic women underwent more frequent screening mammography compared to non-Hispanic Whites, despite having lower breast cancer risk (McGuinness et al., 2018). Previous research of intention to pursue genetic testing for HBOC risk has found that attitudes and beliefs about genetic testing are a significant predictor of intention (Braithwaite, Sutton, & Steggles, 2002; Kessler et al., 2005). Future research can include constructs from the TPB as a framework to understand barriers to the low completion of genetic testing that the authors observed in the current study of primarily Hispanic women at high risk for breast cancer; this population may have unique needs and require additional support to remove barriers to completion of genetic testing.

About 25% of patients who attend genetic counseling experience clinically significant levels of anxiety, and anxiety levels have been associated with decision making and adherence to screening methods and to risk-reduction measures (González-Ramírez et al., 2017). Psychoeducation, a form of psychosocial intervention, has been well documented in the literature to improve coping among individuals with mental illness (Bevan Jones et al., 2018). Future studies are needed to explore whether the use of psychoeducation during genetic counseling and testing sessions could reduce anxiety levels among women who intend to complete genetic testing. To increase BRCA1/2 genetic testing, more efficient models of genetic counseling, such as telephone-based counseling, telemedicine counseling, and group counseling, could be used to facilitate greater access to genetic testing services.

Limitations

Several limitations of this study warrant discussion. The study was conducted at a single institution with a relatively small sample of primarily Hispanic women undergoing mammography screening, and the findings may not be generalizable to other populations geographically or to those who do not seek mammography. Multigene panel testing is the standard of care; however, the study was limited to BRCA1/2 genetic testing intention. Mental illness history was based on self-report from patients, and the authors were unable to confirm clinical diagnosis of mental illness via medical record review. However, the authors included validated measures of anxiety and depression. The study assessed prior completion of BRCA1/2 genetic testing in the survey. However, only five participants reported having completed genetic testing, and the authors were unable to confirm genetic testing via medical record review; therefore, these women were included in the genetic testing intention (yes) group. In addition, the theoretical framework focused on intention, and a further limitation is that intention may not lead to completion; other constructs of this theory can be explored in future studies. Of note, although intentions to have genetic testing were high, other possible barriers affecting uptake of genetic testing include lack of systematic family history screening with a screening tool, such as the SPS, at the mammography site to identify appropriate candidates and refer them to have genetic counseling or testing and limited access to cancer genetic services.

Implications for Nursing

The current study indicates that the majority of Hispanic women with self-reported mental illness who are at high risk of developing breast cancer have high intentions of completing BRCA1/2 genetic testing. The findings indicate that healthcare providers, particularly nurses, should be aware of the high prevalence of mental illness among Hispanic women who meet family history criteria for BRCA1/2 genetic testing. Although genetic testing intention was high, very few Hispanic women completed BRCA1/2 testing. Because of the underuse of BRCA1/2 genetic testing among women at high risk who are eligible for testing, healthcare providers, particularly nurses, should be aware of barriers impeding completion of genetic testing, particularly for racial and ethnic minorities who experience cancer health disparities. Nurses should be aware that Hispanic women with a personal history of mental illness and high anxiety or depression levels may require additional psychosocial support to facilitate completion of genetic testing. Healthcare providers, particularly nurses, who are on the frontline of health care, are well positioned to identify women who are eligible for genetic testing through family history screening, to assess mental health status, and to provide psychosocial support. Prior to genetic testing, mental health history should be considered because individuals with a psychiatric history may be at greater risk for anxiety post–genetic testing (Hirschberg, Chan-Smutko, & Pirl, 2015). Useful screening tools, such as the psychosocial aspects of hereditary cancer questionnaire and the psychological health interview, assess psychosocial functioning of individuals who intend to undergo genetic counseling and testing (González-Ramírez et al., 2017). These tools may facilitate greater completion of genetic testing, particularly among multiethnic women, who are more likely to experience cancer health disparities.

Conclusion

High-risk multiethnic women who had a high prevalence of anxiety or depression had high intentions of completing BRCA1/2 genetic testing, but actual reported completion of genetic testing was low. Although mental illness was not significantly associated with BRCA1/2 genetic testing intention after adjusting for covariates, healthcare providers’ attempt to increase completion of BRCA1/2 genetic testing among high-risk multiethnic women should include assessment of mental health status and other competing barriers at the patient, provider, and healthcare system level. Addressing psychosocial needs, such as anxiety and depression, in women at high risk for breast cancer may increase the rate of those who intend to have BRCA1/2 genetic testing.

About the Author(s)

Tarsha Jones, PhD, RN, PHNA-BC, is an assistant professor in the Christine E. Lynn College of Nursing, and Katherine Freeman, DrPH, is a professor and Biostatistics Core Leader, both at Florida Atlantic University in Boca Raton; Marra Ackerman, MD, is a clinical assistant professor and psychiatrist at New York University Langone Medical Center in New York; and Meghna S. Trivedi, MD, MS, is an assistant professor of medicine at the Herbert Irving Comprehensive Cancer Center, Thomas Silverman, MPH, is a staff associate in the Department of Biomedical Informatics, Peter Shapiro, MD, is a psychiatrist in the Department of Psychiatry, Rita Kukafka, DrPH, MA, is a professor of biomedical informatics and sociomedical sciences, and Katherine D. Crew, MD, MS, is an associate professor of medicine and epidemiology at the Columbia University Irving Medical Center, all at Columbia University in New York. No financial relationships to disclose. Jones, Ackerman, Shapiro, Kukafka, and Crew contributed to the conceptualization and design. Crew completed the data collection. Jones, Freeman, and Crew provided statistical support. Jones, Freeman, Trivedi, Silverman, and Crew provided analysis. Jones, Freeman, Ackerman, Trivedi, Silverman, Kukafka, and Crew contributed to the manuscript preparation. Jones can be reached at jonest@health.fau.edu, with copy to ONFEditor@ons.org. (Submitted December 2018. Accepted July 8, 2019.)