Predictors of breast cancer screening in Asian and Latina University students.
Preventative screening in the form of regular breast examinations remains one of the best protections against breast cancer. The medical literature strongly recommends that women conduct monthly self-examinations and undergo a clinical breast examination every one to three years beginning at age 20; beginning at age 40, women are encouraged to continue monthly self-examinations, obtain an annual clinical breast examination, and undergo a mammogram every one to two years (American Cancer Society, 2005b).
Despite the benefits that regular examinations confer, survey data indicate that sizeable numbers of women in the United States do not obtain clinical breast examinations on a regular basis (American Cancer Society, 2005a). Ethnic minority women, particularly Asians and Latinas, are especially unlikely to undergo regular screening for breast cancer, and may suffer increased mortality and morbidity as a result of delayed detection and treatment (Durvasula, Regan, Ureno, & Howell, 2006; National Center for Health Statistics, 1996).
Given the clear utility of clinical breast examinations in the early detection of breast cancer, and the low usage rates of this screening procedure by Asian and Latina women, researchers increasingly have sought to identify the factors implicated in cancer screening behavior among these groups of women. Several putative predictors have been examined, including (lower) acculturation, (lack of) knowledge about breast cancer, (limited) access to health care, (lower) socioeconomic status, and a number of health-related beliefs. Among Asian women, for example, research generally finds that higher rates of breast cancer screening are associated with various indices of acculturation, including English language proficiency (Lee, Lee, & Stewart, 1996), duration of residency in the United States (Yu, Hong, & Seeto, 2003), and education level (Yu, Kim, Chen, & Brintnall, 2001). A similar association between acculturation and breast cancer screening behavior has been observed among Latina women (Abraido-Lanza, Chao, & Gates, 2005; Gorin & Heck, 2005). In addition, some investigators have suggested that a lack of accurate knowledge about breast cancer may explain low rates of screening behavior (see Schettino, Hernandez-Valero, Moguel, Hajek, & Jones, 2006). Indeed, one recent study employing a focus group methodology identified several misconceptions about breast cancer (e.g., the belief that breast cancer is incurable, that mastectomy is the only available treatment for breast cancer, that screening tests are needed only after the development of observable symptoms such as lumps or pain) that were commonly endorsed by Latina women of diverse nationalities (Borrayo, Buki, & Feigal, 2005; also see Simon, 2006). Still other researchers have targeted health beliefs as important predictors of screening behavior among ethnic minority women. In general, the results suggest that perceptions of cancer severity, cancer susceptibility, screening benefits, barriers to obtaining screening, and other health-related cognitions may prevent women from seeking adequate preventative care (Ahmad & Stewart, 2004; Consedine, Magai, & Neugut, 2004).
The present study was designed to expand earlier work by examining the significance of these and other predictors of breast cancer screening (clinical breast examinations) among a sample of Latina and Asian university women. In keeping with previous investigations, we included among our predictors measures of acculturation, socioeconomic status, access to health care, cancer knowledge, and health beliefs, as well as demographic and personal history variables with a known association with cancer risk (i.e., age, family history of cancer; Rothemund, Paepke, & Flor, 2001). In addition, we included a variable that only recently has come under scrutiny as a potential predictor of cancer screening behavior--sexual intercourse experience. Some researchers have found an association between sexual experience and screening behavior for other reproductive cancers (e.g., cervical cancer; see Byrd, Peterson, Chavez, & Hecker; 2004; Durvasula, Regan, Ureno, & Howell, in press). Consequently, we hypothesized that sexual intercourse experience, along with acculturation, health beliefs, and the other variables included in this investigation, would significantly predict breast cancer screening among our sample of university women.
Participants and Procedure
Participants were Asian or Latina women enrolled in an urban university in the Los Angeles Metropolitan area (N = 240; 165 Latina, 75 Asian). All were 18 years of age or older (M age = 20.15 years, SD = 3.91 years). Following the protocol approved by the university Institutional Review Board, potential participants were recruited via fliers soliciting volunteers for a large-scale study of college women's attitudes and behavior. Those who agreed to participate signed a consent form, completed a series of measures, and received compensation for their time in the form of class credit or a token fee. Participants completed the measures individually and their responses were anonymous (i.e., questionnaires were identified by numbers only and signed consent forms were collected and stored separately from completed questionnaires).
The following measures were administered in randomized order to the participants.
Demographics. We assessed a variety of demographic variables including age, ethnicity, education level, occupation, marital status, and annual and household income. Demographic variables of interest in this study included age, ethnicity, and annual household income (which was used as a general index of socioeconomic status).
Health-Related Beliefs. Health-related beliefs were assessed with a modified version of the Adherence Determination Questionnaire (DiMatteo et al., 1993). This is a 38-item self-report inventory designed to measure nine constellations of health beliefs, including perceptions of severity, susceptibility, intentions, barriers to and supports for screening, interpersonal aspects of care, utility, subjective norms, and efficacy. The 4-item Interpersonal-communication subscale and 4-item Interpersonal-rapport subscale assess interpersonal aspects of health care related to communication and rapport with health professionals (e.g., "Doctors and other health professionals listen carefully to what I have to say"). The Utility-cost and benefits subscale consists of four items designed to measure the perceived utility of obtaining regular cancer screening as a function of estimated costs and benefits (e.g., "Following recommendations on annual cancer screening is better for me than not following these recommendations"), and the Utility-efficacy subscale contains four items that assess beliefs about the effectiveness of cancer screening in actually preventing cancer (e.g., "I believe that regular cancer screening will help to prevent me from getting cancer and will keep me healthy"). The Perceived severity subscale includes three items that reflect beliefs and attitudes about the impact and severity of cancer (e.g., "There is little hope for people who have cancer"), whereas the 5-item Perceived susceptibility subscale assesses beliefs about one's own likelihood of contracting the disease (e.g., "The chances I may develop cancer are pretty high"). The six items comprising the Subjective norms subscale reflect beliefs about the normative expectations about cancer screening held by friends, immediate family, and relatives (e.g., "Members of my immediate family think that I should get regular cancer screening"). The Intention subscale contains four items that measure the intention to obtain cancer screening (e.g., "I have made a commitment to get regular cancer screening"). The final 4-item subscale assesses perceptions of Support and Barriers related to cancer screening, and includes such items as "Lots of things get in the way of getting regular cancer screening" and "I need more assistance in order to get regular cancer screening."
Sexual Experience. We also asked participants several questions related to their sexual history. The variable of interest concerned the participants' sexual intercourse experience; namely, whether or not the participant had ever engaged in sexual intercourse.
Cancer Knowledge. This measure consisted of four open-ended questions that focused on the participants' knowledge about breast cancer and about the purpose of and procedures associated with breast cancer screening. Two coding teams rated the items, and generated a score for each item with higher scores reflective of greater knowledge.
Family History of Cancer. To assess cancer incidence within the participants' family, the number of family members with cancer was summed for each participant.
Acculturation. The Acculturation Rating Scale for Mexican Americans (ARSMA-II; Cuellar, Arnold, & Maldonado, 1995) was administered to any participants identifying as Hispanic or Latina. The ARSMA-II is a 48-item scale that assesses acculturation by measuring cultural orientation toward the Mexican culture and the Anglo culture independently. Although the original scale was designed specifically for Mexican Americans, the scale was modified to include Central Americans as well. Participants who identified as Asian received the Suinn-Lew Asian Self-Identity Acculturation Scale (Suinn,Ahuna, & Khoo, 1992). This 26-item scale is used to assess level of acculturation for Asians living in the United States, and higher scores are reflective of greater acculturation to Western culture.
Breast Cancer Screening Behavior. Participants were asked to indicate using a yes/no response option whether they had ever been screened for breast cancer (i.e., had undergone a clinical breast examination).
Overview of Analytic Strategy
Only 39.4% of Latinas and 36.0% of Asians reported having had a clinical breast examination. Logistic regression analyses were conducted to determine whether the set of variables in our study reliably predicted breast cancer screening among the two groups of participants. Predictor variables included: (1) age; (2) socioeconomic status (i.e., annual household income); (3) sexual intercourse experience (yes/no); (4) acculturation (the Suinn-Lew for Asian participants and the ARSMA-II for Latina participants); (5) familial cancer incidence (i.e., number of family members ever diagnosed with cancer); (6) health-related cognitions (i.e., the nine subscales of the Health Attitudes Scale); and (7) breast cancer knowledge. We were specifically interested in the ability of the overall set of predictors to reliably account for cancer screening among each sub-sample of participants; thus, we present information on goodness-of-fit, strength of association, and prediction success. The importance of individual predictors is evaluated using standard procedures (Wald test and odds ratio) as well as likelihood-ratio tests.
Predictors of Breast Cancer Screening: Latina Sample
The results of the analysis revealed that the set of 15 predictors successfully distinguished between Latina women who had undergone breast cancer screening and those who had not, [chi square](15) = 52.91, p < .001. Additional goodness-of-fit statistics comparing observed with expected outcome frequencies provide further support for the reliability of the predictors (p = .33 for the Deviance criterion and p = .26 for the Pearson criterion; these non-significant differences indicate that the 15-predictor model adequately duplicated the observed frequencies at both levels of the outcome variable). The overall prediction success rate for the model also was high (77.3%); the model correctly predicted the outcome category of 86.6% of the women in the non-breast exam group and 62.0% of those in the breast exam group. Additionally, a bivariate regression of outcome scores predicted from the model on actual (observed) outcome scores indicated that the set of predictors was strongly associated with the outcome variable ([R.sub.YY], = .51, F[1,130] = 4.74, p < .001), and accounted for 25.6% of the variance in breast cancer screening.
As illustrated in Table 1, tests of the individual parameters revealed that sexual intercourse experience significantly differentiated between women who had undergone a clinical breast exam and those who had not; a woman was five times more likely to have undergone breast cancer screening if she was sexually experienced than if she was not (z = 8.84,p < .005). The importance of this predictor was corroborated by a likelihood ratio test, which indicated that the 15-predictor model was significantly degraded (less reliable) when sexual experience was removed ([chi square] = 10.10, p < .005).
Cancer-related knowledge (i.e., knowledge of the importance, purpose, and procedures of a clinical breast exam) also significantly predicted cancer screening status (z = 6.57, p < .05), with greater knowledge about the examination associated with a higher likelihood of screening. As with sexual experience, removal of this variable from the model significantly reduced predictive ability ([chi square] = 7.34, p < .01). None of the other variables successfully predicted breast cancer screening status, although there was a trend for the barriers subscale of the Health Attitudes Scale. The fewer the barriers to obtaining cancer screening that women perceived, the more likely they were to have had a clinical breast exam (z = 3.21, p = .07; removal of this predictor tended to degrade the model, [chi square] = 3.35, p = .07).
Regression Analysis: Asian Sample
The set of predictors successfully distinguished between Asian women who had undergone breast cancer screening and those who had not, [chi square](15) = 30.73, p < .05. The Deviance criterion (p = .29) also indicated that the model was reliable. The overall prediction success rate for the model was high (85.5%); the model correctly predicted the outcome category of 90.9% of the women in the non-breast exam group and 77.3% of those in the breast exam group. Furthermore, a follow-up analysis regressing predicted outcome scores on actual observed outcome scores indicated that the predictors as a set were strongly associated with the outcome variable ([R.sub.YY], = .69, F[1,53] = 49.35, p < .001), and accounted for 48.2% of the variance in breast cancer screening.
As illustrated in Table 2, tests of the individual parameters revealed three variables--age, sexual intercourse experience, and perceived barriers to cancer screening--that predicted whether or not women have undergone a clinical breast exam. Both Wald (z = 6.69, p < .05) and likelihood ratio ([chi square] = 9.91, p < .005) tests point to the importance of perceived barriers in predicting breast cancer screening; the higher the perceived barriers to obtaining a clinical breast exam, the less likely a woman was to actually have undergone one. Similarly, although the Wald test did not achieve significance (z = 3.28, p = .07) for sexual intercourse experience, the likelihood ratio test indicated that removal of this predictor degraded the model and significantly reduced its ability to predict cancer screening among Asian women ([chi square] = 3.84, p < .05). The results for age displayed the same pattern. The Wald test did not achieve significance (z = 3.06, p = .08), but the likelihood ratio test demonstrated that removal of this predictor significantly degraded the reliability of the model ([chi square] = 6.10, p < .05).
The goal of the present investigation was to identify predictors of breast cancer screening among Asian and Latina university women. Among Latinas, the most powerful predictor of screening behavior was sexual activity. Sexually active Latina women were five times more likely to have undergone screening for breast cancer than their sexually inexperienced counterparts. Sexual experience also emerged as a predictor of screening behavior among Asian women, with sexually active Asian women nine times more likely to have had a clinical breast examination than non-sexually experienced women. It is possible that sexually active women are more likely than their sexually inactive counterparts to visit a gynecologist for contraception or other sexual health care needs and thus receive clinical breast examinations as part of their evaluation. Unfortunately, women who view gynecologic care as unnecessary prior to the onset of sexual intercourse may miss an important opportunity for early cancer detection.
Although we did not specifically address attributions with respect to gynecologic care, we did investigate general beliefs about health and health care. Our results revealed that greater perceived barriers and lower perceived support were associated with a lower likelihood of screening in both groups of women. Other researchers also have found a relationship between these particular health-related beliefs and screening behavior. For example, perceived barriers such as embarrassment and fears about pain and discomfort have been observed in women across ethnic groups, but appear particularly common among Asian and Latina women (Ahmad & Stewart, 2004; Consedine et al., 2004; Goldman & Risica, 2004; Wu & Yu, 2003).
Two other findings warrant mention. First, among Asian women, age was associated with screening likelihood, with older women more likely to have undergone at least one clinical breast examination. Because older women in our sample also were more likely to be sexually active, this finding may be attributed at least partially to the fact that older age was associated with sexual activity and an increased likelihood of receiving gynecologic care. Other investigations examining age have yielded inconclusive results (Katapodi, Lee, Facione, & Dodd, 2004). Second, among Latina women, breast cancer knowledge was associated with screening behavior; this finding was expected and mirrors the results reported by earlier researchers (e.g., Friedman, Neff, Webb, & Latham, 1996; Pearlman, Clark, Rakowski, & Ehrich, 1999). However, the reciprocal nature of the relationship between knowledge and screening behavior must be acknowledged, as women who have experienced a clinical breast examination may possess more knowledge simply as a consequence of having been exposed to the procedure.
The lack of relationship between acculturation and screening behavior for either ethnic group was unexpected and bears further consideration. Previous studies have found an association between acculturation and likelihood of breast cancer screening in ethnic minority women (Tang, Solomon, & McCracken, 2000; Tang, Solomon, Yeh, & Worden, 1999; but see Borrayo et al., 2005). The fact that we did not may be due to the fact that our sample was comprised of university students, whose general level of acculturation is likely higher than that typically observed in community samples. For example, all women in our sample were fluent in written and spoken English, which is often cited as one of the more important elements of how acculturation impacts health behaviors. This may have limited the variability of acculturation in our sample and, consequently, its predictive power.
In sum, this study provides a cross-sectional overview of breast cancer screening behavior in Asian and Latina university women. Sexual intercourse experience was the most consistent predictor of whether or not a woman had ever been screened for breast cancer. Health-related cognitions (particularly those related to perceived barriers to cancer screening) also appear to play an important role in cancer screening, as do age (particularly for Asian women) and cancer-related knowledge (particularly among Latinas).
We acknowledge that we utilized a convenience sample that was drawn from a university population. Consequently, our results may have limited utility for understanding screening practices for other Asian or Latina populations. In addition, we relied on self-report, and while we employed anonymous questionnaires, biased reporting of screening behavior may still have occurred, particularly among women who may have some embarrassment about reporting their gynecological histories. These issues notwithstanding, our findings suggest that the university or college campus can play an important role in cancer prevention. Knowledge can be disseminated to students through multiple methods, ranging from electronic messages on registration websites, to student newspapers, to campus health campaigns. Cancer screening is a particularly important behavior to target on campuses with ethnically diverse student populations who may not fully utilize available health services.
Abraido-Lanza, A. F., Chao, M. T., & Gates, C. Y. (2005). Acculturation and cancer screening among Latinas: Results from the National Health Interview Survey. Annals of Behavioral Medicine, 29, 22-28.
Ahmad, F., & Stewart, D. E. (2004). Predictors of clinical breast examination among South Asian immigrant women. Journal of Immigrant Health, 6, 119-126.
American Cancer Society (2005a). Cancer Facts and Figures 2005. Atlanta: American Cancer Society.
American Cancer Society (2005b). Cancer Prevention & Early Detection: Facts and Figures 2005. Atlanta: American Cancer Society.
Borrayo, E. A., Buki, L. P., & Feigal, B. M. (2005). Breast cancer detection among older Latinas: Is it worth the risk? Qualitative Health Research, 15, 1244-1263.
Byrd, T. L., Peterson, S. K., Chavez, R., & Hecker, A. (2004). Cervical cancer screening beliefs among young Hispanic women. Preventive Medicine, 38, 192-197.
Consedine, N. S., Magai, C., & Neugut, A. I. (2004). The contribution of emotional characteristics to breast cancer screening among women from six ethnic groups. Preventive Medicine: An International Journal Devoted to Practice and Theory, 38, 64-77.
Cue1lar, I., Arnold, B., & Maldonado, R. (1995). Acculturation Rating Scale for Mexican Americans-It: A revision of the original ARSMA Scale. Hispanic Journal of Behavioral Sciences, 17, 275-304.
DiMatteo, R. M., Hays, R. D., Gritz, E. R., Bastani, R., Crane, L., Elashoff, R., Ganz, P., Heber, D., McCarthy, W., & Marcus, A. (1993). Patient adherence to cancer control regimens: Scale development and initial validation. Psychological Assessment, 5, 102-112.
Durvasula, R. S., Regan, P. C., Ureno, O., & Howell, L. (2006). Frequency of cervical and breast cancer screening rates in a multi-ethnic female college sample. Psychological Reports, 99, 418-420.
Durvasula, R. S., Regan, P. C., Ureno, O., & Howell, L. (in press). Predictors of cervical cancer screening in Asian and Latina university students. College Student Journal, xx, xxx-xxx.
Friedman, L. C., Neff, N. E., Webb, J. A., & Latham, C. K. (1996). Early breast cancer detection behaviors among ethnically diverse low-income women. Psycho-Oncology, 5, 283-289.
Goldman, R. E., & Risica, P. M. (2004). Perceptions of breast and cervical cancer risk and screening among Dominicans and Puerto Ricans in Rhode Island. Ethnicity and Disease, 14, 32-42.
Gorin, S. S., & Heck, J. E. (2005). Cancer screening among Latino subgroups in the United States. Preventive Medicine, 40, 515-526.
Katapodi, M. C., Lee, K. A., Facione, N. C., & Dodd, M. J. (2004). Predictors of perceived breast cancer risk and the relation between perceived risk and breast cancer screening: A meta-analytic review. Preventive Medicine, 38, 402.
Lee, M., Lee, F. & Stewart, S. (1996). Pathways to early breast and cervical detection for Chinese American women. Health Education Quarterly, 23 (Supplement), S76-S88.
National Center for Health Statistics (1996). Health, United States, 1995. Hyattsville, MD: Public Health Service.
Pearlman, D. N., Clark, M. A., Rakowski, W., & Ehrich, B. (1999). Screening for breast and cervical cancers: The importance of knowledge and perceived cancer survivability. Women and Health, 28, 93-112.
Rothemund, Y., Paepke, S., & Flor, H. (2001). Perception of risk, anxiety, and health behaviors in women at high risk for breast cancer. International Journal of Behavioral Medicine, 8, 230-239.
Schettino, M., Hernandez-Valero, M., Moguel, R., Hajek, R., & Jones, L. (2006). Assessing breast cancer knowledge, beliefs, and misconceptions among Latinas in Houston, Texas. Journal of Cancer Education, 21 (Supplement 1), S42-S46.
Simon, C. E. (2006). Breast cancer screening: Cultural beliefs and diverse populations. Health & Social Work, 31, 36-43.
Suinn, R. M.,Ahuna, C., & Khoo, G. (1992). The Suinn-Lew Asian Self-Identity Acculturation Scale: Concurrent and factorial validation. Educational and Psychological Measurement, 52, 1041-1046.
Tang, T. S., Solomon, L. J., & McCracken, L. M. (2000). Cultural barriers to mammography, clinical breast exam, and breast self-exam among Chinese American women 60 and older. Preventive Medicine, 31, 575-583.
Tang, T. S., Solomon, L. J., Yeh, C. J., & Worden, J. K. (1999). The role of cultural variables in breast self-examination and cervical cancer screening behavior in young Asian women living in the United States. Journal of Behavioral Medicine, 22, 419-436.
Wu, T. Y., & Yu, M. Y. (2003). Reliability and validity of the mammography screening beliefs questionnaire among Chinese American women. Cancer Nursing, 26, 131-142.
Yu, E. S., Kim, K. K., Chen, E. H., & Brintnall, R. A. (2001). Breast and cervical cancer screening among Chinese American women. Cancer Nursing, 9, 81-91.
Yu, M. Y., Hong, O. S., & Seeto, A. D. (2003). Uncovering factors contributing to under-utilization of breast cancer screening by Chinese and Korean women living in the United States. Ethnicity & Disease, 13, 213-219.
PAMELA C. REGAN
RAMANI S. DURVASULA
California State University, Los Angeles
Table 1. Breast Cancer Screening Among Latina Women: Tests of Individual Predictors Parameter Wald Odds Predictor estimate [chi square] ratio Age 0.17 2.26 1.18 SES -0.31 1.34 0.73 Sexual experience 1.62 8.84 *** 5.04 Familial cancer incidence -0.40 1.26 0.67 Acculturation 0.28 0.73 1.33 Cancer knowledge 0.52 6.57 ** 1.68 Health attitudes (HAS) Interpersonal--communication -0.21 2.54 0.81 Interpersonal--rapport 0.04 0.13 1.04 Utility--costs and benefits 0.14 1.26 1.15 Utility--efficacy -0.16 2.47 0.85 Perceived severity -0.14 1.26 0.87 Perceived susceptibility 0.09 0.53 1.09 Subjective norms -0.02 0.60 0.98 Intentions 0.07 0.47 1.07 Support/barriers -0.17 3.21 * 0.84 Table 2. Breast Cancer Screening Among Asian Women: Tests of Individual Predictors Wald Parameter [chi Odds Predictor estimate square] ratio Age 0.53 3.06 * 1.69 SES 0.23 0.42 1.25 Sexual experience 2.25 3.28 * 9.48 Familial cancer incidence 0.68 0.92 1.98 Acculturation 0.91 0.84 2.49 Cancer knowledge 0.64 1.61 1.90 Health attitudes (HAS) Interpersonal--communication 0.08 0.10 1.09 Interpersonal--rapport -0.24 0.33 0.79 Utility--costs and benefits 0.08 0.11 1.08 Utility--efficacy 0.12 0.21 1.13 Perceived severity -0.09 0.07 0.91 Perceived susceptibility -0.15 0.19 0.86 Subjective norms 0.08 1.88 1.09 Intentions -0.21 0.88 0.81 Support/barriers -0.61 6.69 ** 0.54
|Printer friendly Cite/link Email Feedback|
|Author:||Regan, Pamela C.; Durvasula, Ramani S.|
|Publication:||College Student Journal|
|Date:||Dec 1, 2008|
|Previous Article:||Preliminary examination of first year female university students: smoking practices and beliefs in a city with no-smoking legislation.|
|Next Article:||Junior college experience and students with learning disabilities: implications for success at the four year university.|