Introduction

Breast cancer is the most common cancer among African American women. In 2007, there were an estimated 19,000 new cases of breast cancer and nearly 6,000 deaths in the African American population (Jemal et al. 2007). Although the overall incidence of breast cancer is about 12% lower in African American women than in Caucasian women, the incidence is higher in African American women under the age of 40. In addition, African American women experience poorer outcomes when diagnosed with breast cancer. Between 2000 and 2003, African American women had a 36% higher death rate from breast cancer than Caucasian women in part due to more advanced disease at diagnosis (American Cancer Society [ACS], 2007).

Delay in seeking medical care has been found to be a significant factor for late-stage of breast cancer at diagnosis in African American women (Bibb 2001; Burgess et al. 1998). Burgess et al. (2001) identified several factors that contribute to delay in seeking care and prediction of late-stage breast cancer at diagnosis. They include failure to disclose the presence of the breast symptom to others, discovery of a symptom that is not a lump, and lower socioeconomic status. Other factors include being African American, never having been married, having no private health insurance, lacking transportation, and lacking money (Lannin et al. 1998).

Predictors of delay in seeking medical care also include nonadherence to breast-screening guidelines (breast self-examination, clinical breast examination, and mammography); fear; denial; reliance on alternative therapies; other priorities; low perceived risk; belief that the symptom(s) will go away; and religious, spiritual, and cancer fatalism beliefs (Caplan et al. 1996; Facione et al. 1997, 2002; Mohamed et al. 2005, Powe 1995).

Religiosity, Spirituality, and Cancer Fatalism Beliefs Among African American Women

Both religiosity and spirituality have been identified as important factors in the health care decisions of African American women (Bourjolly 1998; Dessio et al. 2004; Kinney et al. 2002; Mitchell et al. 2002; Underwood and Powell 2006; Wallston et al. 1999). In many areas of the United States, African Americans are more religious than their Caucasian counterparts (Newport 2004; Taylor 2003). To date, however, most studies of the influence of religion have focused on the relationship between involvement in the social and structural aspects of religion and mortality among African Americans (Bourjolly 1998; Musick et al. 2004; Taylor et al. 2004). Although religiosity and spirituality are potential factors in health care decisions, they have been little studied, particularly among African American women (Dessio et al. 2004; Facione et al. 2002). The relationship between fatalism, a related construct, and delay in seeking medical care for self-detected breast symptoms has also been understudied (Mayo et al. 2001).

Yet, African American women hold strong religious and spiritual beliefs upon which they often base decisions regarding health and illness (Bourjolly 1998; Dessio et al. 2004; Underwood and Powell 2006). For example, Mitchell et al. (2002) concluded that belief in the superiority of spiritual or religious intervention by a higher power through prayer or laying on of hands compared to medical treatment may help explain why African American women delay presentation of palpable breast lumps, contributing to advanced-stage breast cancer at diagnosis.

Similarly, in a national survey, Dessio et al. (2004) found that religion and spirituality were positively associated with health-care seeking by African American women for minor illnesses but were often barriers to seeking health care for serious conditions such as cancer.

Delay in seeking medical care has been found not only in individuals who believe that because God controls their outcomes, religious intervention is preferable to medical intervention (Barroso et al. 2000; Holt et al. 2003; Johnson et al. 2005; Kinney et al. 2002; Underwood and Powell 2006) but, also in individuals with a fatalistic view of cancer diagnosis, that is, in individuals who believe that death is inevitable if cancer is present (Powe 1997). Across numerous studies, the majority of the participants who held fatalistic views were African American, less educated, of lower socioeconomic status, and over the age of 50 (Facione 2002; Mayo et al. 2001; Powe 2001; Powe et al. 2005; Powe and Johnson 1995). Studies of cancer fatalism have also found that fatalism scores are higher for African American women than for Caucasian women or men of either race.

No studies, however, have explored the influence of religiosity and spirituality on the time to seek medical care for self-detected breast changes in African American women. No studies examined the influence of cancer fatalism beliefs on delay in seeking medical care for self-detected breast changes. Therefore, this study examined the influence of religiosity, spirituality, and cancer fatalism on time to seek medical care and stage of breast cancer in African American women. Specifically, the study examined the relationships between religiosity, spirituality, breast cancer fatalism, disclosure of symptoms, and marital status, and time to seek medical care and breast cancer stage in African American women with self-detected breast changes.

Methods

Sample and Setting

Participants were African American women 30 years of age or older who had been diagnosed with breast cancer after detection of one or more breast symptoms by the woman herself or her partner within 12 months prior to study initiation. A convenience sample of 129 participants was recruited from six medical clinics in a metropolitan area of the southeast. The sample of 129 was based on three power analyses for the three constructs of spirituality, religiosity, and cancer fatalism. The participants were identified from recommendations by oncologists and oncology clinic nurses at each site and were enrolled over a 6-month period. The study was approved by the centers study IRB, and all participants signed informed consent.

Variables and their Measurement

The three variables of influence explored were religiosity, spirituality, and cancer fatalism, and the two outcome variables were time to seek medical care and breast cancer stage. Because time to seek medical care was found not to be normally distributed, it was transformed from a continuous variable into a dichotomous variable expressed as <3 months and >3 months. The cut-off period of 3 months was based on previous research which found that a delay of 3 months or greater after self-detected breast changes resulted in worse outcomes (Elmore et al. 2005; Richards et al. 1999; Facione and Giancarlo 1998). Breast cancer stage was also dichotomized into Stage I and greater than Stage I (McCarthy et al. 2000; Taplin et al. 2004).

A self-report questionnaire gathered information from participants on demographic variables, breast symptoms, family or personal history of breast cancer, time elapsed from symptom discovery to seeking medical care, and whether and to whom the women disclosed information about their breast symptom. In addition, we measured three constructs which were used to measure spirituality, religiosity, and cancer fatalism. To measure spirituality, participants completed the Religious Coping Activity Scale (RCAS), a 29-item Likert-type scale divided into six subscales. For purposes of this study only the 12-item Spiritual-Based Coping subscale was used to measure an individuals’ reliance on a living relationship with God (Pargament et al. 1990). In addition to the original 12 subscale items, the study included three new or modified items on phrases often spoken among African American women when describing their spiritual beliefs (Gullatte 2006): (a) in dealing with the problem, I was guided by God to wait; (b) in dealing with the problem, I was guided by God to take care of it immediately, not wait; (c) I trusted that my faith would see me through; and (d) when faced with a difficult problem, I just “let go and let God.” Possible responses to items on the subscale ranged from one, not at all, to four, a great deal, with higher scores indicating higher levels of spirituality. Internal consistency reliability for the original subscale, as measured by Cronbach’s alpha, 0.92. In the present study, internal consistency reliability for the revised 15-item subscale was Cronbach’s alpha of 0.85.

The Religious Problem Solving Scale (RPSS) was used to measure religiosity. The 36-item RPSS is a measure of religious problem solving designed to distinguish between the responsibility individuals assign to themselves and to God, as well as measure the level of initiative taken by individuals in problem solving (Pargament et al. 1988). The RPSS has three subscales; items on all three are answered using a 5-point Likert scale from one, never, to five, always. Higher scores on the RPSS indicate greater religiosity. Pargament et al., reported Cronbach’s alphas for the subscales as collaborative = 0.93, self-directing = 0.91, and deferring = 0.89. In this study, the Cronbach’s alpha for the total scale was 0.83.

A modified version of Powe Fatalism Inventory (Powe 1995) was used to measure fatalism. This modified Powe Fatalism Inventory (mPFI) scale was used because it is breast cancer specific (Mayo et al. 2001). The mPFI consists of 11 items with yes or no responses; a high score (>5) indicates a high level of cancer fatalism. Mayo et al. reported a Cronbach’s alpha of 0.89 for the mPFI. The internal consistency reliability was alpha of 0.72 in this study. The instruments were completed by participants during a clinical visit, and then reviewed for completion by the investigator. If any items had not been completed, the women were asked if they would like to complete them. All the measures were administered in one encounter. The time needed to complete the survey ranged from 20 to 30 min.

Results

The Sample

The mean age of participants was 54 years, with a range of 30–84 years. Most of the participants were not married; 76% were divorced, widowed, separated, or never married. Fifty-two percent reported an educational level of high school graduation/high school equivalent or less. Nearly 48% reported an annual family income of <$15,000. Not surprising the predominantly self-reported religious denomination was Protestant, self-identified as Southern Baptist. Fifty-eight percent of the participants reported having partial health coverage with 25% reported as either Medicare or Medicaid.

Seventy-two percent of the subjects reported no family history of breast cancer. A breast lump or “knot” was the most frequently detected breast change (54%), and 43% of the participants reported two or more symptoms (pain, “knot” or lump, nipple discharge, itching, or change in color or skin). Fifty-nine percent reported waiting more than 3 months from symptom detection to seeking medical care. The median time to seek medical care from symptom discovery was 4 months with a standard deviation of 4.5 months and range from less than a month to 18 months.

The majority of the women (60%) reported that they had told a person about their breast symptom but 28% had only told God. The majority of the women (67%) presented with greater than Stage I breast cancer. Disclosure of a breast symptom has been reported to be a positive factor in decreasing a woman’s delay in seeking medical care (Facione and Giancarlo 1998). In this study, women who had told a person (a family member or church supporter) about their breast change were significantly more likely to seek medical care earlier (OR: 0.25; 95% CI: 0.11, 0.57; P = 0.01, while women who talked only to God were more likely to delay seeking medical care. A Mann–Whitney test indicated that women who reported talking only to God were more likely to delay in seeking medical care (z = −2.297, P = 0.02).

A significant association was found between time to seek medical care and breast cancer stage at diagnosis using chi- square analysis (χ² [49.2, df = 15], P = 0.01). The longer the time from detecting breast change to seeking medical care, the more advanced the breast cancer stage at diagnosis. Logistic regression showed a significant positive association between delay of symptom presentation for medical diagnosis and stage of breast cancer (OR = 6.37, 95% CI = 2.84, 14.30, P = 0.01). For every one-unit increase in time to delay, the odds of having a more advanced stage of breast cancer increased by 6.37.

Religiosity and Time to Seek Medical Care

The minimum score possible on the RPSS was 65 and the maximum was 159. The sample’s mean score was 117. Similar numbers of participants scored below (n = 69) and above (n = 60) the mean. To determine the effect of religiosity on time to seek medical care, logistic regression analyses were conducted using the RPSS total score. No association was found between religiosity and time to seek medical care (see Table 1). However, level of income was positively associated with time to seek medical care (OR: 0.79; 95% CI: 0.68, 0.92; P = 0.01); the higher the income, the more likely women were to seek medical care soon after discovery of their breast symptom.

Table 1 Logistic regression-backward model (full) time to seek medical care/religiosity (N = 129)
Full size table

An association was also found between disclosure of the breast symptom and time to seek medical care (OR: 0.25; 95% CI: 0.11, 0.57; P = 0.01). Women who had disclosed the detection of their breast symptom to another person were likely to seek medical care sooner than those who had either told no one or only talked to God.

Spirituality and Time to Seek Medical Care

Scores on the Spiritual Based Coping (SBC) subscale of the RCAS showed little variation between participants’ spiritual beliefs. The minimum possible score on the instrument was 15 and the maximum was 60. The participants’ mean score of 54 indicated a high level of spirituality among the study participants in general.

Time to seek medical care was regressed on the SBC subscale score with the covariates of age, level of education, income, history of breast cancer, marital status, insurance status, and whom the women had told of their symptom. Once again only income (OR: 0.77; 95% CI: 0.65, 0.90; P = 0.01) and whom the women had told about their breast symptom (OR: 0.19: CI: 0.08, 0.47; P = 0.01) were significant indicators of time to seek medical care. Participants with a higher income were more likely to seek medical care after self-detection of a breast symptom. Likewise, women who shared the symptom discovery with family or friends were more likely to seek medical care sooner. Table 2 displays the full model for spirituality and time to seek medical care.

Table 2 Logistic regression-backward elimination model (full) time to seek medical care/spirituality (N  = 129)
Full size table

Breast Cancer Fatalism

The extent to which cancer fatalism measured using the mPFI was associated with time to seek medical care was examined controlling for age, level of education, income, and family history. Mayo et al. (2001) has noted that a score >5 on the mPFI (on a scale from 0 to 10) is indicative of a high level of fatalism. Responses to statements about God’s will and death from breast cancer suggested that these women did not hold fatalistic beliefs about breast cancer, and this was supported by their mean score of 2.8. Table 3 summarizes responses from the women to several statements reflecting fatalistic beliefs. The responses of the women showed very low levels of fatalism.

Table 3 Selected cancer fatalism item responses (n = 129)
Full size table

This lack of fatalistic beliefs about breast cancer may have been due to the information they received about their cancer situation and its treatability after their diagnosis. That is, perhaps this information might have dispelled fatalistic beliefs if they had any. Since the study did not measure fatalism beliefs at the time of symptom discovery, it is not known whether their responses on the fatalism instrument would have differed prior to diagnosis.

When cancer fatalism was entered into the regression model along with other predictor variables (age, income, family history, disclosure of symptom, marital status, insurance status, and level of education), no association was found between cancer fatalism and time to seek medical care. This finding was consistent with the findings of Mayo et al.’s (2001) breast-cancer fatalism study. Table 4 shows the complete model for cancer fatalism and time to seek medical care. Women with higher reported incomes and women who told a person about their discovery of a breast symptom were more likely to seek medical care when they discovered a breast symptom. Cancer fatalism was not found to be either positively or negatively correlated with either religiosity (r = 0.11, P = 0.19) or spirituality (r = 0.106, P = 0.23). However, as expected, a positive correlation was found between religiosity and spirituality (r = 0.53, P = 0.01).

Table 4 Logistic regression backward (full) breast cancer fatalism and time to seek medical care (n = 129)
Full size table

Discussion

This study was the first to examine the relationships of religiosity, spirituality, and cancer fatalism beliefs in African American women to delay in seeking treatment for self-detected breast symptoms later diagnosed as cancer. Consistent with the literature, the women in the sample had high levels of religious and spiritual beliefs. However, this study did not find that religiosity, spirituality, or cancer fatalism beliefs were significant predictors of time to seek medical care. The fact that the women had been recently diagnosed with breast cancer and had begun to receive active treatment may have heightened both their religious and spiritual awareness and their positive views of cancer survival, influencing their responses to the questionnaires.

The study findings were consistent with similar studies showing that after disclosing symptoms to others, women are more likely to seek medical care. Women who kept the symptoms within a strict religious context by speaking only to God about them were more likely to delay seeking medical care. These findings suggest that educational messages should urge African American women to discuss their health concerns with another person when they are confronted with symptom discovery, rather than interacting only with God. The idea is not to discourage them from evoking their religious and spiritual beliefs in coping with health concerns but to use them as a complement to seeking medical care.

Cancer fatalism beliefs were not found to be predictors of time to seek medical care. In fact, these women were more optimistic than fatalistic, perhaps in part because they were actively undergoing treatment for breast cancer. Also, fatalistic beliefs were not found to be correlated with religiosity or spirituality, consistent with the findings of Mayo et al. (2001). Delay in seeking medical care was associated with a more advanced stage of cancer at diagnosis, consistent with other research showing delay continues to be a major problem for African American women. Nearly 60% of the participants had delayed seeking medical care for 3 months or more after detecting a breast change and 67% presented with greater than Stage I breast cancer at diagnosis. Clearly, while there has been an overall decrease in breast cancer mortality, African American women continue to present with advanced disease at diagnosis. Women need to be encouraged to both adhere to breast screening guidelines and seek medical care sooner. Although religiosity and spirituality were not associated with time to seek medical care, the findings in this study regarding delay in seeking medical care and the extent to which these beliefs pose a barrier or a benefit for women needs further study.

The findings offer some insights into the consequences of disclosing breast symptoms to a person and disclosing the symptoms only to God. African American women who had shared the news of their breast symptom changes with a family member or friend sought medical care sooner. These results are consistent with other findings related to barriers to screening and likelihood to delay in seeking medical care (Facione 2002; Facione et al. 1997; Johnson et al. 2005; Taplin et al. 2004). Although the majority of the participants had some form of insurance coverage, women with higher incomes were also more likely to seek care sooner.

Study Limitations

This study was based on self-report and recall and used a convenience sample of African American women. The study also was retrospective in that participants were already in treatment and had received at least one cycle of chemotherapy. Therefore, religiosity, spirituality, and fatalism for breast cancer measuring their views prior to diagnosis might have addressed the issue of whether religious, spiritual, and fatalism beliefs alter after diagnosis. Further research is needed to capture religious, spiritual, and fatalism beliefs after symptom discovery and before diagnosis to be able to fully address the question of the influence of religiosity, spirituality, and cancer fatalism on time to seek medical care.

Conclusion

Nevertheless, this study perhaps offers insights into the issue of advanced cancer at diagnosis among African American women. The participants in this study had high levels of religious and spiritual beliefs, and they were likely to delay >3 months after discovering a breast symptom and likely to present with a later stage of breast cancer. Participants who had told a person about their breast symptom were more likely to seek medical care sooner than those who had disclosed their breast symptom only to God, who were more likely to delay seeking medical care.

Religious and spiritual beliefs were not predictive of the women’s decision to delay seeking medical care once a breast symptom had been detected. Further, the participants did not hold fatalistic beliefs about their breast cancer. Perhaps the active treatment they were receiving led them to have more positive views about the treatability of cancer.

Women’s decisions regarding whether and when to seek medical care after self-discovery of a breast symptom are complex. Religious and spiritual beliefs are important factors in their initial response to finding a symptom, but it appears that additional factors, such as their willingness to share their symptom with a person and receive support, are also important in the decision making for seeking medical care.

Further research may help to understand the role of social support in disclosure of breast symptoms to others and how it might lead to prompt diagnosis. This study found that disclosure encourages women to seek medical care sooner and thus promotes earlier diagnosis. In addition, research could explore the effects of introducing faith-based intervention and educational programs that assure women that seeking medical care when they have health concerns does not mean relinquishing their religious and spiritual beliefs.