Fax: (801) 585-6458
The impact of receiving genetic test results on general and cancer-specific psychologic distress among members of an African-American kindred with a BRCA1 mutation
Version of Record online: 12 OCT 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 11, pages 2508–2516, 1 December 2005
How to Cite
Kinney, A. Y., Bloor, L. E., Mandal, D., Simonsen, S. E., Baty, B. J., Holubkov, R., Seggar, K., Neuhausen, S. and Smith, K. (2005), The impact of receiving genetic test results on general and cancer-specific psychologic distress among members of an African-American kindred with a BRCA1 mutation. Cancer, 104: 2508–2516. doi: 10.1002/cncr.21479
- Issue online: 18 NOV 2005
- Version of Record online: 12 OCT 2005
- Manuscript Accepted: 5 JUL 2005
- Manuscript Revised: 9 JUN 2005
- Manuscript Received: 19 APR 2005
- National Human Genome Research Institute
- National Institute of Nursing Research
- National Cancer Institute. Grant Numbers: 1 R01 HG02241, 1 R01 HG02241-02S1
- BRCA1 gene mutation;
- genetic testing;
- African Americans;
- psychologic adjustment;
Numerous studies have examined short-term and long-term psychologic responses to genetic testing for breast/ovarian carcinoma susceptibility in clinic samples and among families who participated in genetic linkage studies. However, to the authors' knowledge, the vast majority of studies focused on non-Latino whites and women. In this prospective study, the authors investigated the psychologic impact of receiving carrier-specific BRCA1 test results as part of a genetic education/counseling intervention in female and male members of an African-American kindred with a BRCA1 mutation.
Eighty-five of 101 participating kindred members (84%) underwent genetic counseling/education and testing according to an established protocol. Participants completed in-person or telephone-administered, computer-assisted interviews. At baseline and after the receipt of test results (1 mo, 4 mos, and 12 mos), general psychologic distress (i.e., anxiety and depression) and cancer-specific distress were measured. Statistical analyses were performed using linear mixed-model approaches for longitudinal data.
The hypothesis that mutation carriers, particularly women who had no personal history of breast carcinoma, were expected to report greater distress than noncarriers was not supported. After controlling for socioeconomic status and personal history of breast/ovarian carcinoma, noncarriers reported significant declines in the distress measures (depressive symptoms, anxiety and cancer-related worries), whereas distress was not altered markedly in carriers after genetic risk notification.
The current findings suggested that individuals receiving BRCA1 test results who learn that they are not carriers of a deleterious mutation may experience psychologic benefits. Furthermore, those who learned that they were mutation carriers did not appear to have adverse, clinically meaningful psychologic outcomes. Cancer 2005. © 2005 American Cancer Society.
In recent years, numerous studies have examined short and long-term psychologic responses to genetic testing for breast/ovarian carcinoma susceptibility. To our knowledge, virtually all of those studies focused on non-Latino whites. However, diffusion of clinical genetic testing for cancer is taking place in the context of an increasingly diverse U.S. population. A major limitation of research on the psychologic sequelae of BRCA testing is the under-representation of racial and ethnic minorities in the vast majority of studies.1–3 Because genetic testing for cancer susceptibility has potentially important mental and physical health benefits as well as the potential for adverse psychologic consequences in some individuals, greater documentation of the behavioral and psychosocial outcomes of genetic counseling and testing in African Americans and other understudied populations is needed. Such information is critical in designing interventions and providing culturally sensitive familial cancer clinic services.
Between 5% and 10% of breast carcinomas and at least 10% of ovarian carcinomas can be attributed to genetic predisposition. Most cases of hereditary breast and ovarian carcinoma are attributed to BRCA1 and BRCA2 mutations.4, 5 The estimated average lifetime risk in female BRCA1/BRCA2 mutation carriers for the development of breast carcinoma is 45–82%.6, 7 Detection of BRCA1/BRCA2 mutations in both clinical and research settings, as well as studies of penetrance, survival, and health service utilization among carriers, have been conducted primarily with non-Latino whites. To our knowledge, differences in the prevalence of BRCA1/BRCA2 between blacks and non-Latino whites have not been observed.8–10
There are limited data supporting the notion that notifying individuals of a genetic risk of cancer leads to adverse psychologic effects.11–13 In fact, there is a growing body of evidence suggesting psychologic benefits for individuals who learn that they are noncarriers of deleterious BRCA gene mutations and no significant increases in distress among mutation carriers.11–13 Several studies have shown that, compared with mutation carriers or with individuals who do not undergo testing, noncarriers have clinically meaningful reductions in both short-term and long-term psychologic distress in research and clinical samples.14–20 Because these studies focused primarily on non-Latino whites, it is important to validate these findings in members of racial/ethnic minority subgroups in addition to considering intervention strategies.
The current study's design was guided by a prior needs assessment,21, 22 a community advisory board that consisted of African Americans, and a stress and coping model that was specific to genetic testing.23 The conceptual model proposed by Baum et al.23 purports that individuals who perceive that they are at risk for a life-threatening disease may carry a substantial burden of stress because of the threat of disease and uncertainty of risk. This threat may include the possible transmission of a deleterious gene mutation to offspring. Genetic testing may help reduce distress for some but may bring about various levels of distress in others. According to this model, heightened psychologic distress may result in the event of a positive test (e.g., confirmation of the presence of a deleterious BRCA1 mutation). In the current study, we evaluated the effect of receiving genetic test results on general and cancer-specific psychologic distress among African Americans at high risk for carrying a deleterious BRCA1 mutation.21 We assessed general and cancer-specific levels of distress before and after individuals learned their BRCA1 carrier status. We also examined for evidence of differential response to BRCA1 test results.
MATERIALS AND METHODS
Participants were members of a high-risk African-American kindred (K2099) that was identified previously with the BRCA1 M1775R mutation during a genetic linkage study and a subsequent gene isolation study.24 Because of the exploratory nature of the study, linkage study participants did not have genetic counseling through the study and were not provided with information about their individual risk of carrying a BRCA1 mutation. By using the pedigree from the original study, and after they consented to be contacted, potential participants were sent an introductory letter that invited them to participate. The updated, 5-generation K2099 pedigree includes 239 members age 18 years and older (210 living members and 29 deceased members). At enrollment, study participants were at least age 18 years, provided written informed consent, and did not know their personal BRCA1 mutation-specific status. We were able to contact 161 living and eligible kindred members. Of the living, eligible, and contactable kindred members, 101 members enrolled at the time of these analyses. At 1-year follow-up, the retention rate was 81%.
All eligible study participants were invited to participate in or decline the genetic counseling and testing sessions. Eighty-four percent elected to undergo genetic testing and learn their test results; 22% were found to be carriers (n = 19 participants), and 78% were not carriers (n = 66 participants).
After the baseline interview, kindred members participated in a face-to-face genetic counseling session that was conducted by a certified genetic professional according to an established genetic counseling protocol25 that incorporated culturally targeted genetic education materials.26 Pretest genetic education and counseling consisted of a detailed family history assessment; an assessment of targeted medical and screening history; an explanation of BRCA1 testing, including the benefits, risks, and limitations; and an overview of screening recommendations and preventive options, depending on the individual's carrier status. Posttest genetic counseling included notifying participants of their carrier status and discussing probability and age-specific cancer risk, implications for themselves and their relatives, and medical management suggestions. Subsequent interviews were conducted at 1 month, 4 months, and 12 months after either the baseline assessment or the pretest genetic counseling session.
Psychologic Distress Outcome Measures
Baseline internal consistency estimates (Cronbach α) for the current study population are presented.
Center for Epidemiologic Studies-Depression.
The 20-item Center for Epidemiologic Studies-Depression scale was used to assess depressive symptoms (α = 0.87) at each interview.27
State-Trait Anxiety Inventory.
At the baseline and at 1-month interviews, the 20-item State Anxiety scale of the State-Trait Anxiety Inventory (STAI) (Form Y) was administered (α = 0.90).28
Impact of Event Scale.
Variables considered were age, gender, educational level, household income, cancer status (personal history of breast and/or ovarian carcinoma), number of first-degree relatives with breast and/or ovarian carcinoma, sibling group, and presence of living children. BRCA1 status results were obtained from Myriad Genetics Laboratories, Inc. (Salt Lake City, UT).
Data analyses were performed using linear mixed-model approaches for longitudinal data38 and were conducted using the SAS® software package39 (version 8.0; SAS Institute Inc., Cary, NC). We evaluated the main effects of carrier status (carriers, noncarriers) and time (baseline, 1 month, 4 months, and 1 year, when data were collected) as well as the moderation effect (i.e., the interaction term between time and carrier status) on psychologic distress. Statistically significant effects were followed with post-hoc testing using the Tukey–Kramer method of multiple comparisons. The primary analyses used the SAS PROC MIXED function to conduct unbalanced, repeated-measures analyses. Potential correlations between observations among siblings or other nuclear family members were assessed by fitting a random-effects modeling “sibling factor”; in all instances, these effects were nonsignificant according to likelihood ratio testing.40, 41
Because of their associations with psychologic distress in this sample, the variables age, status of breast or ovarian carcinoma history, education, and income were retained in the models for each distress-related outcome. In addition, covariates that were associated significantly with a particular dependent variable at the P ≤ 0.20 level using a backward elimination procedure in linear regression models also were included in the fully adjusted models for each distress outcome.
Demographic information on the study population is presented in Table 1. Overall, the mean age of participants was 44.0 years (standard deviation, 13.8 yrs) at baseline. Tables 2 and 3 present descriptive data on the psychosocial variables studied. By using Mantel–Haenzel chi-square, chi-square, and Fisher exact statistics, as appropriate, no significant differences were observed between participants who completed or did not complete the 1-month or 4-month interviews with respect to the sociodemographic variables assessed.
|Variable||No. of patients (%)a|
|Total no. of patients||85 (100.0)|
|No. of female patients||57 (67.1)|
|< 40 yrs||30 (36.1)|
|40-49 yrs||34 (41.0)|
|≥ 50 yrs||19 (22.9)|
|Less than high school||10 (11.8)|
|High school graduate||21 (24.7)|
|Some college or technical school||36 (42.3)|
|College graduate||18 (21.2)|
|< $20,00||18 (21.4)|
|≥ $75, 000||17 (20.2)|
|Married/living as married||68 (80.0)|
|Health insurance (yes)||63 (74.1)|
|First-degree relative(s) with breast or ovarian ca|
|Two or more||16 (19.0)|
|Personal history of breast and/or ovarian ca|
|Affected female||5 (5.9)|
|No children||21 (28.4)|
|Son(s) but no daughters||9 (12.2)|
|One or more daughters||44 (59.5)|
|Variable||Kindred members (n = 85)||BRCA1 carriers (n = 19)||BRCA1 noncarriers (n = 66)|
|Mean (SD)||Range||Mean (SD)||Range||Mean (SD)||Range|
|Social support||19.32 (3.22)||12.00-23.75||17.84 (2.64)||12.50-23.25||19.75 (3.26)||12.00-23.75|
|Religious coping||18.33 (3.04)||9.00-25.67||18.82 (3.07)||14.33-25.67||18.19 (3.04)||9.00-23.33|
|Collaborative||22.61 (5.80)||7.00-30.00||23.37 (5.11)||13.00-30.00||22.39 (6.00)||7.00-30.00|
|Self-directing||12.80(5.04)||6.00-24.00||13.00 (4.81)||6.00-19.00||12.74 (5.14)||6.00-24.00|
|Deferring||19.59 (6.32)||6.00-30.00||20.11 (6.01)||11.00-30.00||19.44 (6.45)||6.00-30.00|
|Depression||14.67 (8.98)||3.00-47.37||14.26 (7.23)||3.00-29.00||14.78 (9.47)||4.00-47.37|
|Anxiety||54.21 (3.35)||44.00-60.00||54.14 (2.50)||50.00-58.00||54.23 (3.58)||44.00-60.00|
|Cancer worry||8.06 (2.31)||3.00-14.00||7.21 (1.84)||3.00-10.00||8.31 (2.39)||3.00-14.00|
|Variable||1 month||4 months||1 year|
|Mean (SD)||Range||Mean (SD)||Range||Mean (SD)||Range||Mean (SD)||Range||Mean (SD)||Range||Mean (SD)||Range|
|Depression||9.83 (4.90)||3-19||11.47 (9.39)||2-49||12.80 (7.49)||5-28||13.12 (9.57)||2-42||11.79 (6.64)||4-24||12.01 (9.31)||4-40|
|Anxiety||33.08 (10.09)||23-50||29.04 (7.97)||23-59||NA||NA||NA||NA||NA||NA||NA||NA|
|Cancer worry||5.08 (1.93)||3-9||6.71 (2.33)||3-12||6.23 (2.74)||3-11||6.53 (2.16)||3-12||NA||NA||NA||NA|
|Intrusive thoughts||10.18 (3.68)||7-18||11.62 (5.08)||7-26||10.80 (4.57)||7-20||10.31 (4.20)||7-22||NA||NA||NA||NA|
|Avoidant thoughts||12.82 (4.83)||8-22||13.12 (5.38)||8-26||12.93 (4.10)||8-22||11.75 (4.70)||8-24||NA||NA||NA||NA|
For the depression outcome, which was measured at baseline, 1 month, 4 months, and 1 year, the model also included marital status (F[1,80] = 1.31; P = 0.26) and the number of first-degree relatives with cancer (F[2,77] = 1.54; P = 0.22) between the baseline and 1-year interviews as covariates. These results revealed a significant main effect of time on depressive symptoms (F[3,62] = 5.02; P = 0.004; effects d = 0.30) (Fig. 1), suggesting that participants reported fewer symptoms of depression over the 4 time points. There was not a statistically significant difference noted with regard to self-reports of depression between baseline and 1 month for carriers (t = 2.33; P = 0.29), but there was a significant difference for noncarriers (t = 3.67; P = 0.01). No other contrasts between time points reached statistical significance. The decrease in reports of depressive symptoms appeared to be small over the 1-year follow-up; however, it is noteworthy that the effect size for the change in depression between baseline and the 1-month follow-up was larger (a medium effect; d = 0.43).
In the model that assessed state anxiety, marital status (F[1,72] = 0.00; P = 0.96), insurance status (F[1,68] = 1.99; P = 0.16), and the number of first-degree relatives (F[2,72] = 3.51; P = 0.04) were entered as additional covariates. A significant main effect of time on state anxiety emerged (F[1,72] = 227.60; P < 0.001) (Fig. 2). Both carriers and noncarriers reported significantly fewer symptoms of anxiety at 1 month compared with baseline (carriers: t = 7.83; P < 0.001; noncarriers: t = 17.68; P < 0.0001).
We also assessed changes in cancer-specific distress. For the model that assessed cancer worry at baseline, 1 month, and 4 months, the presence of living children (none, sons only, ≥ 1 daughter; F[2,67] = 2.52; P = 0.09) was included as a covariate. Results from this model revealed a significant main effect of time (F[2,58] = 11.93; P < 0.001; d = 0.53) on cancer worry (see Fig. 3). Noncarriers reported significantly fewer cancer-related worries at 1 month compared with baseline (t = 4.30; P < 0.001), and both carriers and noncarriers reported significantly fewer cancer-related worries at 4 months compared with baseline (noncarriers: t = 5.35; P < 0.001; carriers: t = 2.41; P = 0.02). No other time contrasts reached statistical significance.
For analyses that evaluated the Avoidance Scale of the IES, living children (F[2,60] = 0.72; P = 0.49) and the number of first-degree relatives (F[2,57] = 0.87; P = 0.43) were included in the model as covariates. No main or interaction effects were statistically significant in this model, including the main time effect (F[1,48] = 3.10; P = 0.08). For the Intrusive Thoughts scale of the IES, the number of first-degree relatives (F[2,69] = 2.91; P = 0.06) was included in the fully adjusted model as a covariate. In this model, no main or interaction effects differed appreciably, including time (F[1,52] =1.55; P = 0.22).
We evaluated the impact of receiving genetic test results on general and cancer-specific distress in the subgroup of females who had no history of breast or ovarian carcinoma (n = 52 participants). A significant main effect of time on state anxiety (F[2,33] = 5.34; P = 0.01) was observed. Analogous to the larger kindred member cohort, unaffected, noncarrier females reported significantly fewer anxiety symptoms at 1 month compared with baseline (t = 4.98; P = 0.001). The difference in anxiety symptoms from baseline to 1 month was not statistically significant among carriers. No other main or interaction effects were significant among the measures of psychologic distress for the unaffected female subgroup.
In the current study, we investigated associations between BRCA1 mutation carrier status, time, and psychologic adjustment in kindred members after participation in a genetic counseling and testing intervention for a deleterious BRCA1 gene mutation. To our knowledge, the current study is among the first to focus on African Americans. We did not find evidence of short-term or long-term adverse psychologic outcomes among the total number of participants or among females without a personal history of breast/ovarian carcinoma. The hypothesis that mutation carriers, particularly women with no personal history of breast carcinoma, were expected to report greater distress than noncarriers was not supported. Instead, psychologic benefits were observed; after they received genetic test results, noncarriers reported decreases in general and cancer-related distress compared with mutation carriers. These findings are consistent with previous studies in cancer registry families11, 15, 16, 20, 38, 39 and a clinic-based sample.40 In addition, general distress levels reported by kindred members were comparable to community-based normative samples.28, 41 Although it did not reach statistical significance, in the subgroup of female kindred members without a personal history of breast/ovarian carcinoma, in which we may expect to see heightened levels of distress in response to learning their carrier status, a similar pattern of results was observed.
There have been previous reports of subgroups experiencing increased levels of distress associated with genetic testing42; however, to our knowledge, no particular subgroups were identified that had heightened psychologic distress in the current study population. Carriers of the BRCA1 mutation, for example, did not demonstrate heightened distress compared with noncarriers. Furthermore, baseline or follow-up distress levels did not differ markedly for men or women or for unaffected females compared with the overall study population.
It is noteworthy that the intervention in this study had three components: pretest genetic education and counseling, genetic testing, and posttest counseling. The offer of testing leads individuals to deal with 1) the decision regarding whether to undergo testing, 2) their potential changed risk status, and 3) their risk and the options associated with this risk. The second component is genetic counseling, the objectives of which are to educate participants, help them incorporate the new information into their life view, aid in the decision-making process, and help them adjust to the changes engendered by genetic counseling and testing. This study was not designed to identify whether outcomes such as distress were affected by the education and counseling components of the intervention separately from the testing components. Future studies may consider this in their design and evaluate these effects separately.
Some previous studies have suggested a psychologic benefit for noncarriers after testing.3, 43 The psychologic benefits were observed in studies in which participation in testing was relatively high, similar to the current study. It is unknown whether participation in genetic testing would be as high among African Americans who are at high risk for breast/ovarian cancer susceptibility in self-referred clinical samples compared with individuals recruited from registries or from kindred studies.44 In addition, more research is needed to evaluate the impact of genetic testing in clinical and population-based samples among African Americans and particularly whether the development of culturally relevant interventions is associated with the beneficial effect observed in the current study. A direct comparison between a traditional genetic intervention and a culturally targeted intervention would be needed to address this question.
Two established, health-relevant psychosocial resources, social support and religious coping, did not moderate the correlation between carrier status and psychologic adjustment in our study population (data not shown). Although the reasons for this are unclear, these findings should be interpreted with caution because of limited statistical power due to the small sample size of the carrier and noncarrier subgroups. Some studies have suggested that African Americans may report higher levels of social support45, 46 and religiosity45, 47 and therefore may have less variability or a restricted range for testing associations and interactions between these psychosocial factors and psychologic distress among this subgroup. Collecting and evaluating additional measures may be important to clarify this correlation. For instance, some investigators have observed that psychologic distress may be moderated by a siblings' test results and a participant's responses.48 In the current sample, the factor that represented the potential for correlated responses between siblings was not found to be statistically significant in any models of psychologic distress measures. However, it may prove useful to evaluate other familial factors, such as family communication, as potential moderators of the distress response to genetic testing. For example, more adaptive or cohesive patterns of communication within the family may provide a support to individuals undergoing genetic testing that facilitates beneficial psychologic responses to testing. Furthermore, the measures of social support and religious coping were administered at only one time point, before genetic testing. Although these constructs can be viewed as individual characteristics that are stable over time, their cross-sectional assessment also may limit the potential for observing an influential role.
There are important limitations that must be acknowledged in the current study. The African-American kindred participating in the sample may not be representative of African Americans in the U.S. The kindred was identified with a BRCA1 mutation, and family members identify themselves as African-American Creole and share a unique family history in rural Louisiana. Approximately one-half of the participants (52%) took part in prior genetic research, which may have resulted in potential selection bias. Moreover, because the current study population was drawn from a pedigree of a family that participated in a prior linkage study, although they did not know their BRCA1 carrier status, many were aware of their potential genetic risk because of their family history with regard to cancer. This heightened awareness of their familial risk may have influenced their emotional responses. Therefore, the impact of receiving genetic information may have been less for carriers in our sample than would be the case in clinical or population-based samples of African Americans.
An additional limitation is that published studies of distress outcomes in the setting of BRCA mutation testing or that focus on African Americans do not include our measures of social support and religious coping for purposes of comparison. Therefore, the generalizability of the current results to African Americans or to other ethnic minority groups as well as clinical samples may be limited. We assessed for potential biases and did not observe meaningful differences between those who completed the 1-month and 4-month interviews and those who did not; however, similar to any longitudinal study, loss to follow-up may introduce bias. Finally, an important point to consider is our relatively small sample size and low proportion of mutation carriers that led to low statistical power, particularly in analyses of interactions such as differential response over time according to carrier status. There are several reasons for the relatively low percentage of carriers: there were deaths in the kindred from BRCA1-associated cancers (n = 14 participants), there was nonparticipation in the study among living kindred members with breast and/or ovarian carcinoma (n = 9 participants), and testing was offered to those who were at increased risk of carrying a BRCA mutation but did not have a first-degree relative with breast and/or ovarian carcinoma. Consequently, our results may be biased in either direction. Therefore, the findings of the current study (significant and nonsignificant) should be interpreted with caution but suggest continued study of underrepresented populations based on larger samples.
Despite these limitations, the current study has important strengths. This study fills a critical gap in the literature, because to our knowledge it is one of the first prospective studies to examine the psychologic impact of genetic counseling and testing for BRCA1 mutations in high-risk African Americans who are appropriate candidates for genetic testing. It is also unique in that many of the participants reside in the rural South, whereas other published studies from the U.S. have focused on urban or suburban non-Latino whites in the Northeast, West, or Midwest. Both shorter and longer term assessments of psychologic adjustment were evaluated across the adjustment measures. In addition, both general and cancer-specific distresses were assessed, allowing for comparison with other community-based samples. Important potential confounders also were considered in each of the models.
In conclusion, the consequences of genetic testing in our study population of African Americans did not appear to differ appreciably from those among non-Latino whites. Although it remains unknown whether the observed outcomes were related to cultural targeting of the intervention, the findings of the current study suggest that potential benefits of genetic counseling and testing may not be limited to non-Latino whites. Future research in clinical and population-based samples with more representative samples of African Americans and other understudied racial/ethnic groups will be necessary to confirm our findings. In future investigations, it will be important to assess more comprehensively other potential individual characteristics, such as coping styles (e.g., optimism and monitors/blunters) and psychosocial resources, as potential mediators and moderators. It also will be important to determine whether culturally targeted versus nontargeted genetic education and counseling make a difference in psychologic outcomes. Finally, testing the effectiveness of genetic interventions within clinical and community-based systems will be imperative to determining actual responses among high-risk African Americans.
The authors thank Dr. Jeffrey Botkin and Jean Wylie for their assistance with the design and implementation of the project and Kathy Sward for her assistance with data management and programming. They are especially grateful to the genetic counselors (Regan Challinor and Kelly Jackson), the medical geneticist (Mary K. Pelias), the interviewers (Andrea Wiley, Berneice Parker, Kendra Rockwell, Josalin Hunter, Ashley Holmes, Carolyn Ross, and Jucynthia Taylor Ford), and all of the individuals who participated in this study.
- 1Centers for Disease Control and Prevention (CDC). Self-reported frequent mental distress among adults—United States, 1993–2001. MMWR Morb Mortal Wkly Rep. 2004 l53: 963–966.
- 25BRCA1 testing: genetic counseling protocol development and counseling issues. J Genet Couns. 1997; 6: 223–244., , , et al.
- 28STAI manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press, 1971., , .
- 32Health literacy, cervical cancer risk factors, and distress in low-income African-American women seeking colposcopy. Ethn Dis. 2002; 12: 541–546., , , , .
- 34Linear mixed models for longitudinal data. New York: Springer, 2000., .
- 35SAS system for mixed models. Cary, NC: SAS Publishing, 1996., , , .
- 44Relationships among breast cancer concern, risk perceptions, and interest in genetic testing for breast cancer susceptibility among African-American women with and without a family history of breast cancer. Cancer Epidemiol Biomarkers Prev. 1999; 8: 533–539., , , .